THE JOURNAL OF BIOLOGICAL CHEMISTRY

Vol. 250, No. 15, Issue of August 10, pp. 6134-6140,1975 Printed in U.S.A.

Insulin Secretion INTERRELATIONSHIPS OF GLUCOSE, CYCLIC ADENOSINE 3’:5’-MONOPHOSPHATE, AND CALCIUM*

(Received for publication, July 8, 1974)

M. ARTHUR CHARLES, JANUARY LAWECKI, RAYMOND PICTET,$ AND GEROLD M. GRODSKY

From the Metabolic Research Unit and Department of Biochemistry and Biophysics, University of California, San Francisco, California 94143, and the Letterman Army Institute of Research, Presidio of San Francisco, San Francisco, California 94129

Glucose elevates both cyclic adenosine 3’:5’-monophosphate (cyclic AMP) and insulin secretion rapidly and in a parallel dose-dependent fashion in perifused rat islets. Theophylline stimulates cyclic AMP much more than glucose, yet secretion is much less. When the two agents are combined, cyclic AMP is sim- ilar to theophylline alone yet secretion is augmented synergistically. Glucose-induced cyclic AMP gen- eration and insulin secretion are dependent on extracellular calcium. Theophylline-induced insulin secretion is also extracellular calcium-dependent; however, theophylline-induced cyclic AMP elevation is independent of extracellular calcium. Thus, extracellular calcium has multiple effects on insulin secretion, some of which appear unrelated to a terminal secretory process. When glucose is combined with theophylline at physiologic levels of extracellular calcium, both the first and second phases of secretion are prominent. At extracellular calcium levels of 0.05 mM, only the second phase is prominent, whereas at 10 nM extracellular calcium (ethylene glycol bis( @-aminoethyl ether)-N,N’-tetraacetic acid) only the first phase is prominent.

A divalent cation ionophore (A23187, Eli Lilly), which transports calcium and magnesium ions across biological membranes, was used to elucidate further the role of calcium and magnesium. If the ionophore (10 PM) is perifused for 5 min at low extracellular calcium and magnesium, and physiologic calcium is then added, a sudden spike of insulin release occurs in the absence of cyclic AMP generation. Similar results were obtained with magnesium. When the ionophore is perifused for 30 min at low calcium and magnesium, insulin secretion again occurs in the absence of cyclic AMP generation. Electron microscopic examination of the B cells following perifusion with the ionophore shows no specific alterations.

These observations suggest that: (a) glucose elevates cyclic AMP, but the latter acts primarily as a posi- tive feed-forward modulator of glucose-induced insulin release; and (b) extracellular calcium has multiple effects on insulin secretion bot,h upon, and independent of, the cyclic AMP system.

Cyclic adenosine 3’:5’-monophosphate is presumed to be an during glucose-induced insulin release (2-4, 8, 9). Further important mediator of insulin secretion in the B cell of the data have implied that glucose and CAMP induce insulin endocrine pancreas, since agents that elevate or depress islet secretion by inhibiting calcium efflux and enhancing mobiliza- CAMP’ levels induce appropriate changes on glucose-induced tion of intracellular calcium, respectively (10). insulin release (l-5). However, it has been reported that A heterocyclic monocarboxylic acid antibiotic (A23187) pro- glucose does not elevate islet CAMP levels, and therefore the duced by Streptomyces chartreusensis has been observed to direct role of CAMP in glucose-induced insulin release re- be a specific membrane carrier ionophore for divalent cations mained unclear (5-7). More recently, simultaneous and direct in plasma and organelle membranes (11-17). In each case, di- measurements of CAMP and insulin release in dynamic se- valent cations are transported across membranes as directed by cretion systems have documented elevated islet CAMP levels electrochemical gradients. These observations have been seen

* These investigations were supported in part by Grant AM-01410, in several systems: e.g. isolated liver mitochondria (ll), the

National Institute of Arthritis, Metabolism, and Digestive Diseases, isolated sarcoplasmic reticulum (12), the salivary gland and by a grant from the Kroc Foundation of Santa Ynez, Calif. (13, 14), the exocrine pancreas (15), mast cells (16), and

*Recipient of Career Development Award from the National Institutes of Health.

platelets (17). Preliminary studies have shown that this com-

‘The abbreviations used are: CAMP, cyclic AMP, cyclic adenosine pound may be of value in studying the interrelationships of

3’:5’-monophosphate; EGTA, ethylene glycol bis( P-aminoethyl ether)- glucose, CAMP, and calcium upon insulin secretion (Z-20).

N,N’-tetraacetic acid. This report investigates: (a) the relationships between glucose-

6134

by guest on October 6, 2020

http://ww

w.jbc.org/

Dow

nloaded from

induced CAMP elevation and insulin secretion; (b) the multiple

effects of extracellular calcium on secretory regulation; (c) the

ability of calcium and magnesium to induce insulin secre-

tion in the presence of the divalent cation ionophore; and (d)

the ultrastructural appearance of B cells in the perifused islets

during ionophore-induced insulin release.

MATERIALS AND METHODS

Islet Isolation-Adult, fed, Long-Evans rats were used for all studies. Islets were isolated by the method of Lacy and Kostianov- sky (21). In brief, the pancreas was distended with Hanks’ solution containing 0.02% bovine serum albumin, minced, and then digested by collagenase (Worthington Biochemical Corp., Freehold, N. J.) in a mixture containing 3,000 kallikrein inhibitory units of the protease inhibitor, Trasylol (F.B.A. Pharmaceuticals, New York). Digestion was performed by agitation at 37” for 10 to 15 min. Two hundred to three hundred islets were isolated routinely, and each perifusion was per- formed with 200 islets from one rat.

Perifusion Technique-This procedure has been previously described in detail (2) and is similar to that originally described by Lacy et al. (22). The perifusate consisted of a phosphate-bicarbonate buffer (pH 7.4) containing 0.3% human serum albumin, 2 mM glucose, 1.5 rnM phosphate, 29 mM bicarbonate, 145 mM sodium, 114 mM chloride, 4 mM potassium, and 2 mM calcium and 1.2 mM magnesium unless otherwise stated. For experiments performed at low extracellular calcium or magnesium (no added calcium or magnesium), the peri- fusate total calcium and total magnesium were 0.05 and 0.03 mM, re- spectively, as determined by atomic absorption spectrometry. For experiments performed with ethylene glycol bis(P-aminoethyl ether)- N,N’-tetraacetic acid (EGTA, Sigma Chemical Co., St. Louis, MO.), the estimated free calcium was 10 nM and magnesium was 10 PM (23). All perifusions were preceded by a 35min equilibration, and for experi- ments performed at low calcium or low calcium and magnesium the equilibration was performed with the same low concentration of divalent cations in the perifusate. EGTA, when used, was dissolved in the perifusate, and at 0.2 mM no change was noted in the pH of the medium.

Both the first and second phases of insulin release, as well as the half-maximal and maximal responses for glucose-induced release (refer to “Results”), were similar to previously reported data from the isolated perfused rat pancreas (24), thus indicating that perifused rat islets behave in a like fashion.

Insulin and CAMP Determinations-Insulin in the effluent peri- fusate was estimated by radioimmunoassay using a solid state modification of a previously described method (25). Total islet CAMP was measured by modification of the Gilman method (26), and was previously described in detail (2). Islet CAMP levels were determined at the conclusion of each perifusion. Ninety-eight per cent of the radio- activity of [3H]~AMP (New England Nuclear Labs, Gardena, Calif.) eluted in the same fractions as standard CAMP on Dowex 50 (27). Authenticity of the islet CAMP was determined separately by phosphodiesterase degradation (26) and Dowex AG l-X8 formate purification (27). During the course of these studies (2 years), control values of islet CAMP steadily increased from about 1 to 9 pmol/200 islets. For this reason, the CAMP assay was monitored periodically by

6135

the phosphodiesterase and Dowex-formate chromatographic methods to insure that values obtained were in fact those of CAMP. Although the levels of CAMP in nonstimulated islets increased during this period, insulin secretion (both nonstimulated and stimulated) was un- changed (Table I). To minimize overinterpretation of these CAMP values, all data for each experiment are expressed as per cent change over their nonstimulated controls. Galactose (16.7 mM) ele- vated neither insulin secretion nor CAMP levels.

Calcium Studies-The divalent cation ionophore A23187 (a gift from Dr. R. A. Hamill of the Eli Lilly Research Laboratory) was solubilized in ethanol, and the ethanol concentration presented to the islets was 1%. Control experiments using 1%’ ethanol revealed no alterations of calcium-induced insulin secretion and no ultrastructural abnormalities of the B cell structures. The ionophore did not alter the pH of the perifusate.

Electron Microscopy-At the end of representative perifusions, the islets were immediately fixed at room temperature by replacing the perifusate with fixation medium (44 glutaraldehyde in 0.044 M so- dium phosphate buffer, pH 7.4). After at least 60 min, the islets were collected, washed in 0.132 M sodium phosphate buffer (pH 7.4) and fixed in 2% osmium tetroxide in the same buffer at 0” for 60 min. Subsequently, the islets were dehydrated in ethanol and embedded in Epon according to the method described by Luft (28). Thin sections were treated with uranyl acetate and lead citrate and examined in a Phillips 300 electron microscope.

Statistical Analyses-All probability values were determined from the unpaired Student’s t test.

RESULTS

Kinetics of Glucose-induced Insulin Secretion and CAMP Eleuation-Fig. 1 illustrates the pattern of the elevation of CAMP in response to 16.7 mM glucose. By 30 s, CAMP is increased; after 5 min, the levels are maximal and appear to decline slowly during the next 15 min. Since CAMP levels are relatively stable following the initial elevation by glucose, measurements were routinely made 20 min after stimulation. Fig. 2 shows that increasing concentrations of glucose from 2 to 55 mM induce a dose-dependent parallel elevation of in- sulin secretion and CAMP. Half-maximal (9 mM) and maximal (16 mM) responses of both occur at similar glucose concentra- tions.

Effect of Extracellular Calcium on Insulin Secretion and CAMP Elevation-Fig. 3 shows the calcium dependence of glucose-induced insulin secretion and CAMP elevation. Peri- fusion with 27 mM glucose in the presence of physiologic levels of extracellular calcium (2 DIM) elicits the typical biphasic pattern of insulin secretion accompanied by an elevation of CAMP. In contrast, neither perifusion with 27 mM glucose at low calcium (0.05 mM) nor perifusion with 2 mM glucose at physiologic calcium concentrations induces insulin secretion or CAMP accumulation. Further, basal secretion appears inde-

TABLE I

Date -

Control* January to June 1972

October to December 1972 September to December 1973

Stimulation”

January to June 1972

October to December 1972

September to December 1973 I (n) Insulin secretion p value CAMP”

(f-3 (4) (7)

(8)

(5)

(3)

ng/20 minl200 islets

15.1 * 2.1 14.8 + 2.1

20.7 i 2.7

144.9 * 16.5 109.4 * 32.9

131.7 * 13.3

NS’ NS

NS

NS

1.27 + 0.14 3.18 * 0.59

9.29 * 0.71

3.52 zt 0.54

4.74 * 0.21

15.9 * 1.59

p value

0.005 <0.005

NS 0.005

nCyclic AMP levels were reduced to zero by phosphodiesterase (10 mM) combined with glucose (16 mM) gave similar results.

treatment (25) during January to June 1972 and September to b Performed with 2 mM glucose (control) or maximal levels of glucose December 1973; during the latter period, Dowex-formate column (16 to 27 mM) (see Fig. 2).

chromatography was also used (26). Other studies using theophylline c NS, not significant.

by guest on October 6, 2020

http://ww

w.jbc.org/

Dow

nloaded from

6136

400

CYCLIC AMP

(%of control)

TIME (min]

FIG. 1. Time course of glucose-induced CAMP elevation. Islets were perifused with 2 rnrvr glucose (control) or 16.7 rnM glucose, as described under “Materials and Methods.” At 30 s (n = 3), 2 min (n = 7), 5 min (n = 3), 10 min (n = 5), and 20 min (n = lo), the perifusion was discontinued, and 200 islets in each experiment were examined for CAMP content. Control CAMP values over a period of months ranged from 0.98 to 2.5 pmol/200 islets, and all points (mean * S.E.) were statistically different (p < 0.01) from their control values.

300

CYCLIC AMP ,% of Control,

200 0

FIG. 2. Glucose dose-response curves for insulin release and CAMP elevation. At each level of glucose indicated, values (mean * S.E.) were determined from several groups of 200 islets each. After stimulation for 20 min, CAMP was determined. 0, total insulin secreted over the control level of the prestimulatory period; 0, CAMP levels over the control value (1.5 pmo1/200 islets) measured at 2 mM glucose, n values ranged from 4 to 11, and all results are statistically different from the value at 2 rnM glucose.

pendent, of extracellular calcium.

Perifusion with theophylline and a nonstimulatory concen-

tration of glucose (2 mM) yields distinctly different results

for both insulin secretion and CAMP elevation (Fig. 4). At 2

mM calcium, the amount of insulin secretion is lower for both

phases than that observed with the optimal level of glucose. The

theophylline-induced secretion is probably not related to the

low level of glucose in the basal perifusate, since similar

secretion results with theophylline have been obtained in the

absence of glucose in both the isolated perfused rat and

hamster pancreases (29, 30) and in perifused mouse islets using

isomethylbutylxanthine (4). At low calcium concentrations

(0.05 mM), there is an initial small amount of secretion lasting

3 to 4 min. Thereafter, no release above the control level

is noted. In contrast to the calcium dependence of glucose-

induced CAMP elevation, theophylline-induced CAMP eleva-

tion was observed at both low and normal extracellular

FIG. 3. Role of extracellular calcium on glucose-induced insulin se- cretion and CAMP elevation. Groups of 200 islets were perifused in the presence of 27 mM glucose at 0.05 rnM (0 (mean), n = 4) and 2 rnM (0 (mean), n = 8) calcium. Difference between the two curves was significant (p < 0.001) at all periods after zero time. At 20 min, islets were collected for CAMP measurements (mean * S.E.) at both 0.05 mM (open bar) and 2 mM (closed bar) calcium (p < 0.01). The control CAMP level was determined from islets perifused with 0.05 or 2 rnM calcium at 2 rnM glucose, since both gave similar results (8.7 versus 8.4 pmol/200 islets (not significant)).

‘r

I Ii FIG. 4. Role of extracellular calcium on theophylline-induced insu-

lin secretion and CAMP elevation. Groups of 200 islets were perifused in the presence of 10 rnkr theophylline and 2 mM glucose at 0.05 mM (0, n = 4) and 2 rnM (0, n = 6) calcium. Circles represent mean insulin secretion per min. (Differences between the two curves are significant at all periods after 3 min (p < O.Ol).) At 20 min, islets were collected for CAMP measurements (mean * S.E.) at both 0.05 mM (open bar) and 2 rnM (closed bar) calcium; the values were not statistically different. The CAMP level of control islets not perifused with theophylline and collected at the same time was 8.7 + 2 pmo1/200 islets (n = 3).

calcium concentrations. As we have previously observed at

2 mM calcium, when both glucose-induced CAMP levels and

insulin secretion are apparently maximal both are further aug-

mented by the addition of theophylline (Fig. 5, Ref. 2).

Although theophylline alone stimulates CAMP to much higher

levels (Fig. 4) than does glucose, insulin secretion is of much

less magnitude.

Regulation of Expression of Insulin Secretion by Extra- cellular Calcium-As seen in Fig. 5, stimulatory levels of

glucose (27 mM) combined with theophylline (10 mM) at

various levels of extracellular calcium induce both insulin

secretion and CAMP elevation. At the three concentrations of

extracellular calcium, the patterns of insulin secretion are

clearly different. At physiological levels (2 mM calcium), both

the first and second phases of secretion are prominent. At

0.05 mM calcium, only the second phase is prominent and a

by guest on October 6, 2020

http://ww

w.jbc.org/

Dow

nloaded from

I I

~0

too

100

CYCLIC AMP ,% of control)

zoo

00

FIG. 5. Role of extracellular calcium on glucose/theophylline-stim- mated insulin secretion. Groups of 200 islets were perifused separately at three different extracellular calcium concentrations (10 nM, 0.05 mM, and 2 mM) in the presence of 27 mM glucose and 10 mM theophylline. All values represent mean insulin secretion per min; total insulin secreted during the stimulation period was statistically different for each experiment (p < 0.01). 0-O (n = 7), 2 mM extracellular calcium; O-O (n = 5), no added calcium (or 0.05 mM calcium); and 0.. .O (n = 6), no added calcium plus EGTA (or 10 nM calcium). At 20 min, CAMP values (mean * S.E.) were elevated over the control value (p < 0.001) (see legend to Fig. 3). Closed bar, 2 mM calcium; open bar, 0.05 mM calcium; and shaded bar repre- sents 10 nM extracellular calcium.

delay in secretion is observed. At approximately 10 nM calcium (obtained by adding 0.2 mM EGTA to the perifusate), only the first phase is prominent. In control experiments, 0.2 mM EGTA did not alter the insulin secretory response obtained by glucose in 2 mM calcium even when the EGTA was pre-infused at low extracellular calcium. To determine whether theophylline or glucose induces the prominent first phase of secretion at 10 nM extracellular calcium, an experiment was performed in the absence of a stimulatory concentration of glucose (Fig. 6). As is seen, theophylline induces a first phase of secretion of the same amplitude as that induced by glu- cose combined with theophylline. Further, 27 mM glucose in the absence of theophylline at 10 nM calcium did not induce secretion (data not shown).

Zonophore-mediated Insulin Secretion-As shown in Fig. 7, A

and B, perifusion with 10 HM ionophore at low extracellular calcium (0.05 mM) and magnesium (0.03 mM) for 5 min does not induce insulin release. The addition of 2 mM calcium to the perifusate results in a transient increase of insulin secretion (Fig. 7A). The secretory response is the same whether the ionophore infusion is continued or not. A smaller spike-like release is also noted when calcium is increased from 0.05 to 2 mM in the absence of the ionophore. Cyclic AMP levels measured 2 min after the 2 InM calcium infusion at the time of maximum secretion were not elevated. The addition of 1.2 mM magnesium in the place of calcium led to similar results (Fig. 7B). However, insulin release under these conditions requires the presence of the ionophore (Fig; 7B). Fig. 8 shows that, if the ionophore is infused for longer periods of time at low extracellular calcium and magnesium, insulin release is slowly progressive following an initial lag. Again CAMP is not elevated (7.8 * 3 pmo1/200 islets uersus 9.3 * 4 for control) when insulin release is maximal. Similar effects on secretion

6137

EGTA (0.2 mM1

THEOPHYLLfNE (IO mM)

a

fNSULlN RELEASE (ng/min/2OO ir~etr~ A

2

y I I I I I 5 0 5 IO 15 20

TIME (min)

FIG. 6. Effect of theophylline in the presence of 2 mM glucose on insulin secretion in the virtual absence of extracellular calcium. Before stimulation, the islets were perifused for 35 min with low calcium perifusate containing 0.2 rnM EGTA. Circles represent mean insulin secreted during each minute (n = 3). For minutes 1, 2, and 3, secretion is higher than base-line (p < 0.01).

LI o / - . . _L I

5 10 I5 5 IO 15 20

TIME (min)

FIG. 7. Insulin secretion induced by extracellular calcium and mag- nesium. Groups of 200 islets were perifused at 0.05 mM calcium and 0.03 mM magnesium and 2 mM glucose in the presence (0) or absence (0) of 10 PM ionophore (A23187). At 5 min, the islets were stimu- lated by 2 mM calcium (A) or 1.2 mM magnesium (B). Circles represent insulin secreted per min from a typical experiment. Each of ten ex- periments for calcium and seven for magnesium showed the same qualitative pattern of increased secretion. Statistical evaluation of the change from the minute prior to divalent cation stimulation to the peak stimulatory level when compared with the control changes was significant (p < 0.05; unpaired Student’s t-test).

are obtained if EGTA (0.2 mM) is pre-infused for 35 min and infused simultaneously with the ionophore (results not shown;.

Electron Microscopy-After perifusion for 30 min with 10 ELM ionophore at low extracellular calcium and magnesium, all organelles were identified clearly by low power electron microscopy (x 8600) and appear similar to organelles in islets fixed immediately after collagenase digestion; both B and A cells were observed (data not shown). Since the release of insulin in the presence of the ionophore at low extracellular calcium could result from an increase of cytoplasmic calcium by translocation from organelles, special interest is focused on their appearance. Higher resolution electron microscopy of

by guest on October 6, 2020

http://ww

w.jbc.org/

Dow

nloaded from

6138

INSULIN RELEASE

(“g/ml”/200 4Il.bl

TIME (mm,

FIG. 8. Insulin release induced by the ionophore A23187. Circles (n = 4) represent mean insulin secretion per min for groups of 200 islets stimulated by the ionophore in the absence of added calcium (0.05 mM) and magnesium (0.03 mM). All points from minute 10 were greater than basal (p < 0.01).

islets is shown in Fig. 9. Panel A shows intact plasma and vesicle membranes following exposure to the ionophore. Panel B illustrates mitochondria following low divalent cation perifusion without the ionophore, and Panel C shows the mito- chondria after exposure to the ionophore (no differences are noted). Panel D shows the normal appearance of microfila- ments and microtubules following ionophore perifusion.

DISCUSSION

Gradual changes in base-line levels of CAMP over prolonged periods of time have been described previously (31). It is unclear why this phenomenon occurs, but it is unlikely that it reflects altered biologically effective CAMP in rat islets since the sensitivity of insulin secretion to glucose was not affected during this interval. Our initial observation that glucose increases islet CAMP levels (2) was in contrast to other observations (5-7); however, subsequently, groups using dif- ferent methods have recently confirmed and extended this finding (3, 4, 8, 9). In this study we have further demonstrated that the time courses of the CAMP elevation and secretion are similar, with an increase as early as 30 s, and that both CAMP and insulin secretion have identical dose-response relationships to physiologic glucose concentrations.

Since many regulatory systems involving CAMP have been shown to be involved intimately with calcium (32), we have evaluated its relationship with CAMP elevation and the regula- tion of insulin secretion. The ability of glucose to stimulate both CAMP and insulin secretion was found to be dependent upon near-physiological levels of extracellular calcium, whereas the action of theophylline was dependent upon extracellular calcium only for secretion but not CAMP elevation. Two im- plications are apparent: (a) a different mechanism must exist for the elevation of CAMP induced by either glucose or theo- phylline; and (b) extracellular calcium is involved in at least two sites of glucose-stimulated insulin secretion, one involving the formation of CAMP and the other regulating a process ap- parently distal to CAMP action. The two phases of glucose- stimulated insulin secretion are dependent on extracellular calcium. In contrast, theophylline stimulates secretion similar to the first phase at nanomolar levels of extracellular calcium. Thus, first phase secretion can be produced by a mechanism

apparently unrelated to the extracellular calcium concentra- tion. It should be noted that extracellular calcium refers only to the bulk phase calcium and not to the extracellular calcium ad- jacent to the cell surfaces (33). Finally, the stimulation by the simultaneous presence of glucose and theophylline induces markedly different secretory patterns depending upon the level of extracellular calcium, further indicating that the require- ment for extracellular calcium varies according to the type of stimulus present.

The secretion of insulin by theophylline and glucose at low levels of extracellular calcium has been postulated previously to result from mobilization of intracellularly bound calcium mediated by CAMP (10). Consistent with the proposed model, our data directly demonstrate that CAMP is in fact elevated at these low levels of extracellular calcium. In addition, glucose may cause an inhibition of the cellular calcium efflux (34). Both mechanisms may lead to an increase of intracellular cal- cium, which in turn may induce insulin secretion. If this is true, direct elevation of intracellular calcium should stimulate insulin secretion without a glucose or CAMP requirement. The properties of the divalent cation ionophore A23187 have been successfully used in a variety of cells to investigate the role of calcium (11-20). In B cells, under appropriate conditions, the ionophore induces insulin secretion. An early spike-like secre- tion occurs upon addition of calcium to the extracellular milieu. As has been shown in several systems (12, 13, 15, 16), this response is likely to result from the rapid entry of calcium into the cells. The delayed continuous secretion, following the ionophore infusion at low extracellular calcium, may originate from a secondary action of the ionophore on intracellular mem- branes involved in calcium storage, with a resulting release of stored calcium into the cytoplasm. The perfect preservation of the plasma membrane and the various intracellular organelles militates against the trivial explanation of ionophore-induced plasma membrane or organelle disruption. Extracellular mag- nesium inhibits glucose-induced secretion at normal levels of extracellular calcium (35). Contrary to observations using an- other cell system (36), magnesium stimulates secretion in the presence of the ionophore. Since the ionophore A23187 is known to increase transport of both cations, it is possible that there may be naturally occurring ionophores (37) whose action in membranes is specific for calcium and blocked by other cations. Alternatively, the divalent cation-dependent mechanisms af- fecting insulin secretion may not be specific for calcium but may respond to magnesium if the latter ion can be transported to the proper cellular site. Since the insulin release induced in the presence of the ionophore is not accompanied by in- creased levels of CAMP, this calcium effect seems to be distal to the CAMP-sensitive step of the secretory mechanism.

All these data are consistent with CAMP functioning to mo- bilize the intracellularly bound calcium, producing a secondary effect on insulin release. Further, CAMP may be required but is not sufficient to stimulate glucose-induced insulin secre- tion, since theophylline induces much higher levels of CAMP but much less insulin secretion than does glucose. Neverthe- less, CAMP may potentiate secretion, as shown by the fact that the maximum insulin secretion and CAMP elevation in- duced by glucose can both be further increased by theophyl- line. These observations are compatible with a dual effect of glucose in the mechanism of secretion: (a) stimulation of insulin secretion by a mechanism in which CAMP is not a direct inter- mediary but in which calcium appears to be directly involved;

by guest on October 6, 2020

http://ww

w.jbc.org/

Dow

nloaded from

6139

FIG. 9. A, high magnification of a B cell plasma membrane after 30. min perifusion in the presence of 10 PM of the ionophore A23187. The continuous bilaminar appearance is preserved (arrows) in the plasma membrane as well as in the secretory vesicle (V) membrane. The cell boundaries appear smudged where sections of the membrane are tan- gential (x 450,000). B and C, mitochondria after 30-min incubation with low divalent cations in the absence (B) or presence (C) of 10 PM

and (b) an increase of CAMP, creating a positive feed-forward on the secretory process. Thus, a specific secretory mechanism may be established by glucose. A further elevation of glucose increases proportionally the level of CAMP, which in turn mod- ulates the effect of this mechanism in a positive direction by a factor proportional to the CAMP concentration. The resulting effects upon the secretory process appear to be mediated by calcium. Under these conditions, glucose-induced CAMP eleva- tion and insulin secretion have, as has been observed, an identical dose-response relationship. The lower secretion ob- tained with theophylline results from the absence of stimula- tion of the glucose-sensitive mechanism. On the other hand, the potentiation of glucose-induced secretion by theophylline is ex- plained by the further increase in CAMP.

of the ionophore A23187. The mitochondria have a similar appearance in both situations (x 37,500). D, microtubules and microfilaments in a B cell after 30 min of perifusion in the presence of 10 PM iono- phore A23187. Both microfilaments (m) and microtubules (arrows) are present in cells after perifusion, and their ultrastructural appearance is not altered by the presence of the ionophore or the level of extra- cellular calcium (x 41,000).

Acknowledgments-We wish to thank Doctors J. Goodsby and J. Lundblad of Cutter Laboratories, Berkeley, Calif., and James B. Lesh of The Armour Pharmaceutical Co., Kankakee, Ill. for supplying the albumin. We also thank Mr. Rudy Fan- ska, Mrs. Mary Ann Jones, and the Misses Florence Schmid and Michele Manning for excellent technical assistance, and Miss Patricia Phelps for electron microscopy.

REFERENCES

1. Turtle, J. R., and Kipnis, D. M. (1967) Biochem. Biophys. Res. Commun. 28, 797-802

2. Charles, M. A., Fanska, R., Schmid, F. G., Forsham, P. H., and Grodsky, G. M. (1973) Science 179,569-571

3. Grill, V., and Cerasi, E. (1974) J. Biol. Chem. 249, 4196-4201 4. Hellman, B., Idahl, L.-A., Lernmark, A., and Tiiljedal, I.-B. (1974)

by guest on October 6, 2020

http://ww

w.jbc.org/

Dow

nloaded from

6140

5. 6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17. 18.

19.

20.

21.

Proc. Natl. Acad. Sci. U. S. A. 71, 3405-3409 Montague, W., and Cook, J. R. (1971) Biochem. J. 122, 115-120 Miller, E. A., Wright, P. H., and Allen, D. 0. (1972) Endocrinology

91,1117-1119

22.

23.

Cooper, R. H., Ashcroft, S. J. H., and Randle, P. ,J. (1973) 24. Biochem. J. 134, 599-605 25.

Zawalich, W., Ferrendelli, J., and Matschinsky, F. M. (1973) Dia- betes 22, Suppl. 1, 331

Selawry, H., Marcks, C., Fink, G., Lavine, R., Cresto, J., and Recant, L. (1973) Diabetes 22, Suppl. 1, 295

Brisson, G. R., Malaisse-Lagae, F., and Malaisse, W. J. (1972) J. Clin. Inuest. 51, 232-241

26. Gilman, A. G. (1970) Proc. Natl. Acad. Sci. U. S. A. 67, 305-312 27. Murad. F.. Manganiello. V.. and Vaughan. M. (1971) Proc. Natl.

28. 29.

Reed, P. W., and Lardy, H. A. (1972) J. Biol. Chem. 247, 6970- 6977 30.

Scarpa, A., Baldassare, J., and Inesi, G. (1972) J. Gen. Physiol. 60, 735-749

Prince, W. T., Rasmussen, H., and Berridge, M. J. (1973) Biochim. Biophys. Acta 329, 98-107

31.

Selinger, Z., Eimerl, S., and Schram, M. (1974) Proc. Natl. Acad. 32. Sci. U. S. A. 71, 128-131

Williams. J.. and Lee. M. (1974) Biochem. Biohvs. Res. Com- mun. 60, ‘542-548 ’

‘ ”

Foreman, J. C., Mongar, J. L., and Gomperts, B. D. (1973) Nature 245, 249-251

Yuen, M., and Macey, R. (1974) Fed. Proc. 33,271 Charles, M. A., Lawecki, J., Manning, M., and Grodsky, G. M.

(1974) Clin. Res. 22, 128a Charles, M. A., Lawecki, J., Pictet, R., Manning, M., Fanska, R.,

and Grodsky, G. M. (1974) Diabetes 23, Suppl. 1, 337 Zawalich, W. S., Karl, R. C., Ferrendelli, J., and Matschinsky,

F. M. (1974) Diabetes 23, Suppl. 1, 337 Lacy, P. E., and Kostianovsky, M. (1967) Diabetes 16,35-39

33.

34.

35.

36.

37.

Lacy, P. E., Walker, M. M., and Fink, C. J. (1972) Diabetes 21, 987-998

Hubbard, J. I., Jones, S. F., and Landau, E. M. (1968) J. Physiol. 194, 355-380

Grodsky, G. M. (1972) J. Clin. Inuest. 51, 2047-2059 Grodsky, G. M., Bennett, L. L., Smith, D. F., and Schmid, F. G.

(1967) Metabolism 16, 222-233

Acad. Sci. U. s. A. 68, 736-739 I Luft, J. H. (1961) J. Biophys. Biochem. Cytol. 9,409-414 Landgraf, R., Kotler-Brajtburg, J., and Matschinsky, F. M.

(1971) Proc. Natl. Acad. Sci. U. S. A. 68,536-540 Frankel, B. J., Gerich, J. E., Hagura, R., Fanska, R. E., Ger-

ritsen, G. C., and Grodsky, G. M. (1974) J. Clin. Inuest. 53, 1637-1646

Gilman, A. G., and Minna, J. D. (1973) J. Biol. Chem. 248, 6610-6625

Rasmussen, H., Goodman, D. B. P., and Tenenhouse, A. (1972) CRC Crit. Reu. Biochem. 1,95-150

Muller, R. U., and Finkelstein, A. (1974) Proc. Natl. Acad. Sci. U. S. A. 71, 923-926

Malaisse, W. J., Brisson, G. R., and Baird, L. E. (1973) Am. J. Physiol. 224, 389-394

Bennett, L. L., Curry, D. L., and Grodsky, G. M. (1969) En- docrinology 85, 594-596

Cochrane, D. E., and Douglas, W. W. (1974) Proc. Nat’. Acad. Sci. U. S. A. 71, 408-412

Warren, G. B., Toon, P. A., Birdsall, N. J. M., Lee, A. G., and Metcalfe, J. C. (1974) Proc. Natl. Acad. Sci. U. S. A. 71, 622- 626

by guest on October 6, 2020

http://ww

w.jbc.org/

Dow

nloaded from

M A Charles, J Lawecki, R Pictet and G M Grodsky3:5-monophosphate, and calcium.

Insulin secretion. Interrelationships of glucose, cyclic adenosine

1975, 250:6134-6140.J. Biol. Chem. 

  http://www.jbc.org/content/250/15/6134Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/250/15/6134.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on October 6, 2020

http://ww

w.jbc.org/

Dow

nloaded from