Insulin secretion I

POS-001-058 GLUCOSE-DEPENDENT INSULIN SECRETION FROM FISH PRINCIPAL ISLETS

C LAUE, E LIAO, J BODDIN, T STALLMAC~, H MONTEFERIN~, J BEYER, J SCHREZENMEIR

Department of Endocrinology and Inst i tute of Pathology*, University of Mainz, Mainz, FRG.

Avai labi l i ty problems of i s le t transplantation may be obviated by exploiting the special anatomi cal feature of certain fishes, where the endocrine pancreas is concentrated within a principal i s le t = Brockmann body (BB). Previous studies of our group have demonstrated the su i tab i l i t y of BB's of Osphronemus gorami for i s le t transplantation with respect to a) rapid and suff icient ava i lab i l i ty , b) tolerance of mammalian temperatures in long-term culture, c) long-term survival in hollow fibers. With respect to a therapeutic use of fish is lets, a similar regulatory mecha- nism of insulin secretion in the experimental fishes is required. Therefore we tested 1. blood glucose levels of Osphronemus, 2. glucose-dependence of insulin secretion of a. naked, freshly isolated BB's and b. encapsulated, 5 week-precultured BB's. ad i . Fasting blood glucose of Os- phronemus was found to be comparable with the human levels: 89.5C29.6 mg/dl (mea~SD, n=12), ad 2 Incubation of BB's in culture medium with different glucose concentrations showed a glucose-de- + ÷ pendent stimulation of insulin secretion, a.: insulin uU/24h: 42~ i87 at 30 mg/dl, 80i-206 at 60 ÷ + t mg/dl, 1082-i59 at 90 mg/dl, 1722-217 at 200mg/dl glucose, (mean-SEM, n=12, p<O.02 between 200 + ÷ and 90 mg/dl), b.: insulin uU/24h: 94-72 at 30 mg/dl, 1271-438 at 200 mg/dl. These data demonstrate: 1. Fasting blood glucose in fishes providing BB's comes near ~o human fasting values; 2. BB's, as well naked,freshly isolated as encapsulated, precultured release insulin glucose-dependently within human-physiological ranges, which shows a further aspect of su i tab i l i ty for i s le t transplantation.

POS-001-059 ARACHIDONIC ACID IS A COMPLETE AGONIST FOR INSULIN RELEASE: PUTATIVE EFFECTS ON Ca ++ MOBILIZATION AND PROTEIN KINASE C IN INTACT AND PERMEABILIZED RAT ISLETS

STEWART A. METZ University of Colorado and Denver VA Medical Center, Denver, CO, USA

Arachidonic acid (AA) is postulated to have a role in insulin (1) secretion. To examine this possibility directly, exogenous AA was provided to intact rat islets (in Ca+o+-free medium to eliminate the formation of insoluble Ca++-AA soaps). Under these conditions, AA led to saturable increases in 4SCa++ efflux from prelabelled islets (threshold ~ 3/~M AA) and in cytosolic free Ca ++ concentration [Ca++ ] (fura-2 technique), the latter with an efficacy similar to glucose. Concomitantly, AA induced potent I release which was prompt in onset yet sustained. AA was more potent than its hydroperoxy-, hydroxy- or epoxy-metabolites or than other unsaturated fatty acids (18:1, 18:2, 18:3, 20:3, 22:6). Its effects were not blocked by inhibitors of AA oxygenation or of its incorporation into phospholipids. AA effects were slowly reversible, inhibitable by cooling or trifluoperazine, and unaccompanied by trypan blue retention or Slchromium release, indicating that they were not due to cell toxicity. Pretreatment with ionomycin markedly impaired the subsequent 4SCa++ mobilization by AA, and vice versa; thus, AA mobilized Ca ++ from membrane-bound cellular stores. In fact, AA was shown to bind 45Ca++ and translocate it into organic solvents (toluene/butanol or chloroform) in an ionophore-like fashion which was maximal at pH 7.4-8.0 and highly selective for Ca ++ over Mg ++.

Additionally, however, dose-response comparisons suggested that AA has an additional effect unrelated to Ca ++ mobilization, since Ca ++ efflux reached saturation at a lower concentration of AA than did I release. Indeed, AA (like a phorbol ester, 12-O-tetradecanoylphorbol-13-acetate = TPA) promoted secretion in digitonin- (or staphylococcal cc toxin-) permeabilized islets in which [Ca++ ] was fixed through the addition of Ca++/EGTA buffers (basal = 244+23 /.LU/10 islets/30 rain; AA, 73/~M = 436_+44; p<.01; maximal effect of Ca ++ alone = 654_+69). This effect was saturable and totally inhibitable by cooling to 16°C. An 18-hr pretreatment with TPA (but not an inactive phorbol) depleted islets of protein kinase C (as assessed by loss of the response to mezerein, a dissimilar protein kinase C activator); this maneuver blunted the I response to AA by 40_+6% (p<.001) in intact islets and nearly totally in permeabilized islets. Thus, native (unesterified and unoxygenated) AA, in concentrations achieved endogenously in glucose- stimulated islets, is a complete, non-toxic secretagogue, stimulating both the Ca ++ mobilization limb (possibly as an endogenous ionophore) and the Ca ++ potentiation limb (protein kinase C?) of exocytosis.

$106

POS-O01-060 GLYCERALDEHYDE PHOSPHATE AND SUCCINATE MONOMETHYL ESTER:

TWO "NEW" POTENT INSULIN SECRETAGOGUES

M.J. MacDONALD, L.A. FAHIEN, R.S. RANA. AND R.J. MERTZ Univers i ty of Wisconsin-Madison. Madison. WI 53706 USA

Because st imulus-secret ion coupling for glucose, the most potent physio- logic insu l in secretagogue involves aerobic g lyco lys is , we tested a number of glucose metabolites for t h e i r a b i l i t i e s to st imulate insu l in release by pan- creat ic i s l e t s . Even though metabolism-induced insu l in release is probably the complex aggregate of a number of processes, we reasoned that i f a s u f f i - c ient i n t r a c e l l u l a r concentration of a p a r t i c u l a r l y important intermediate was achieved, i t might i n i t i a t e insu l in release. Of a l l intermediates tested, glyceraldehyde phosphate (GAP) was the most potent secretagogue. GAP. at con- centrat ions of 1 mM to 4 mM was even a more potent insu l in secretagogue than glucose or glyceraldehyde. In numerous experiments, insu l in released by 4 mM GAP ranged from 50-200% of that i n i t i a t e d by 16.7 mM glucose - a near-maximal stimulus. GAP was the only intermediate that st imulated i nos i t o l phosphate formation in permeabilized i s l e t s (K n ~=25 uM). Succinate, when added to i s l e t s as e i the r i t ' s monomethyl or ~1~ethyl ester (MMS or DMS) was i nsu l i n - o t rop ic . I s le ts appear to contain esterases that hydrolyze MMS or DMS to succinate. Insul in release was maximal at I0 mM MMS and ranged from 50-100% of that st imulated by 16.7 mM glucose. Unester i f ied c i t r i c acid cycle intermediates did not st imulate insu l in release; nor did esters of fumarate or pyruvate. St imulat ion of insu l in release by MMS and DMS suggests that mitochondrial metabolism alone is s u f f i c i e n t to i n i t i a t e and support insu l in release. I nh ib i t o rs of mitochondrial resp i ra t ion inh ib i ted insu l in release and i nos i t o l phosphate formation by MMS and GAP.

POS-O01-061 INSULIN AND GLUCAGON SECRETORY RESPONSES OF BgMAN FETAL ISLET-LIKE CELL CLUb-rZKS IN TISSUE CULTURE

MAINTAINED

T OTONKOSKI, H KNIP, S ANDERSSON, 0 SIMELL Children's Hospital, University of Helsinki, Helsinki, Finland

Secretory responses of human fetal islet-like cell clusters (ICC) were studied using perifusion. ICC were obtained from 7 fetuses at 13-15 wk and 21 fetuses at 17-22 wk of gestation. After 24-72 h in culture, the ICC were challenged with glucose (20 mM), arginine (I0 mM), glucagon (1.4 uM) and theophylline (I0 mM) combined with zero (OmM), low (2 mM) or high (20 mM) glucose. At 13-15 wk, glucose and arginine stimulated insulin release weakly in some experiments, whereas glucagon was always a potent stimulus (mean response 4-fold irrespective of glucose). Theophylline caused a 4-fold increase in insulin release (again, irrespective of glucose). At 17-22 wk, both glucose (20 mM) and arginine (i0 mM with 2 mM glucose) increased insulin release slightly (l.4-1.5-fold). When arginine was combined with 20 mM glucose, the response was potentiated to 2.3-fold. In contrast, glucagon was equally effective in 2 and 20 mM glucose (2.9- and 2.6-fold response, respectively), and produced a half-maximal response even in zero (0 mM) glucose. In this age range the most potent stimulus for insulin release was clearly theophylline. The effect of theophylline was also remarkably independent of the glucose concentration of the perifusate (5.6-, 8.1-, and 8.6-fold responses at 0, 2, and 20 mM glucose, respectively). Glucagon release from the ICC of the 17-22-wk fetuses was low (mean basal glucagon release 0.1; insulin 1.5 pg/ICC/min). The release of glucagon was not affected by 20 DIM glucose, but was stimulated by arginine and theophylline. These findings suggest that in the human fetal pancreas, in contrast to the adult organ, insulin release results from elevation of intracellular cAMP concentration (by glucagon or theophylline) relatively independent of exogenous glucose.

$107

POS-001-062 BIMODAL ACTIONS OF NUTRIENT SECRETAGOGUES ON CYTOPLASMIC Ca 2+ OF PANCREATIC G-CELLS.

E GYLFE Department of Medical Cell Biology, Uppsala University, Uppsala, Sweden

The effect of insulin secretagogues on the cytoplasmic Ca 2+ concentration (Ca2+i) of oh~oh-mouse G-cells was studied by continuously recording the 340/380 nm fluorescence excitation ratio of the intracellular indicator fura-2. When the glucose concentration was raised from substimulatory to stimulatory concentrations, there was an initial lowering of Ca2+i followed by a sustained increase. Whereas the reduction in Ca2+i was related to the extracellular glucose concentration in a hyperbolic manner, the increasing component exhibited a sigmoidal dose-response relationship, and both effects reached maximum at 15-20 mM of the sugar. Qualitatively similar bimodal Ca2+i responses were obtained with mannose, a-ketoisocaproic acid, leucine and the non-metabolizable but metabolism- -stimulating leucine analogue 2-endo-aminonorbornane-2-carboxylic acid (BCH). Fructose alone (30 mM) had virtually no effect on Ca2+i and very weak bimodal actions in the presence of 3 mM glucose. Arginine, which is practically not metabolised and stimulates insulin release by directly depolarizing the G-cells, only induced a rise of Ca2+i . The results indicate that nutrient secretagogues stimulate both the entry of Ca 2+ into the G-cells and its elimination from the cytoplasm by processes like organelle sequestration and outward transport. Consequently, the Ca 2+. l level determining insulin secretion results from the balance between two opposing actions.

POS-001-063 TPA-MEDIATED STIMULATION OF INSULIN RELEASE FROM HUMAN FETAL PANCREAS REQUIRES

EXTRACELLULAR CALCIUM

K.J. 0SGERBY, B.E. TUCH and J.R. TURTLE. Department of Medicine, University of Sydney, NSW, Australia.

Unlike the adult pancreas, human fetal pancreas releases little or no insulin in response to glucos{; However, insulin release from human fetal pancreas is stimulated by agents allowing Ca influx protein kinase C activators such as 12-O-tetradecanoyl ph~bol- 13-acetate (TPA) and agents which stimulate cyclic AMP. The role of extracellular Ca-- in the TPA-med~ated stimulation of insulin secretion was investigated.

One mm explants of human fetal pancreas (14-20 wk gestational age) were perifused for 120 min in basal medium followed by 50-min period in which various agents were added. Tissue viability was determined by response to 10 mM theophylline. Insulin values (uU/min) are the mean ~ SEM of the area under the curve, from which has been subtracted the mean basal area. The enhanced release of insulin by 1.3 uM TPA [1.18 + 0.46 (n=12); P=0"004]a~S6 inhibited when the tissue was perifused with a combination of TPA and either of the channel blockers i0 uM verapandl or 0.3 uM nifedipine [TPAlverapamil:0.59 ~ 0.21 (n=7):P=NS; TPAlnifedipine: -0.44 ~ 0.15 (n=7):P=0.002]. This inhibition was not reversed by adding 20 mM glucose [TPAlverapamillgl~ose:0.17 + 0.ii (n=7): P=NS; TPAlnifedipinelglucose:-0.27 + 0.07 (n=7): P=NS]. When Ca-- was removed from the medium (and 0.5 mM EGTA added), n~ enhancement was seen in response to TPA/glucose [0.63 + 0.28 (n=~.):P=NS].

When the tissue was perifused with 20 uM BAY-K-8644, a Ca z* channel activator (which alone did not stimulate insulin secretion [-0.22 ~ 0.28 (n=7):P=NS]) and 1.3 uM TPA, enhancement of insulin secretion was seen. This was greater than that observed for TPA alone [BAYITPA:2.66 + 1.07; TPA:0.87 + 0.38 (n=13)~=0.002]. Similar ~hancement was observed when the tissue was perifusedwith i0 mM Ca + 1.3 uM TPA. [Ca- ITPA. 24 25 +

Z÷ " " " -- 5.34; TPA:I6.27 ~ 4.86 (n=8):P=0.O01]. These data show that Ca entry into the cell is required for the stimulatory effect of TPA on insulin secretion from human fetal pancreas.

$108

POS-001-064 HYPERINSULINI~MIA IN A PATIENT WITH INSULIN-AUTOANTIBODIES

H VON SCHENCK, H ARNQVIST Departments of Clinical Chemistry and Internal Medicine, Link6ping University, Sweden

Case history_: A 49 year old woman was admitted to our clinics to investigate a suspected insu- linoma after episodes of postprandial hypoglycemia and an acute period of unconsciousness with registered low bloodglucose. Laboratory investigations: The traditional insulin RIA showed values of 180-600 mU/L indicative of insulinoma, however high blanks urged for more careful studies. In the prolonged fasting state her bloodglucose fell once to 2.7 mmol/L. Otherwise she had normoglycemia. Repeated de- termination of free and total insulin in the fasting state resulted in free insulin values between 12 and 36 mU/L (ref. values: <20 mU/L), whereas total insulin values ranged between 36 and 130 mU/L, thus suggesting the varying presence of insulin antibodies. An oral glucose load (50 g) induced a rise in blood glucose from 4.1 to 14.9 mmol/L at 60 min followed by a drop to 5.6 at 120 and 1.3 at 180 min, at which time we discontinued measurements and injected glucose. During the load, free insulin and C-peptide rose while glucagon decreased adequately. The pres- ence of insulin antibodies was verified by 40% binding of 125I-insulin to the patients serum with full displacement using 10 kU/L human insulin. Gelchromatography of Protein A stripped serum demonstrated insulinlike immunoreactivity in the size of proinsulin. Ultrasound investi- gation and arteriography of the pancreatic gland as well as laparatomy did not reveal any ab- normalities. Discussion: This case demonstrates that laboratory investigation of a suspected insulinoma may give quite unexpected results. Even though the patient never was treated with insulin, insulin autoantibodies developed. We think that those antibodies led to the formation of a large circu- lating pool of insulin liberated pestprandially then producing hypoglycemia. We cannot rule out that the formation of insulinantibodies may be due to improper endogenous proinsulin secretion.

POS-001-065 2-CYCLOHEXENE-I-ONE INHIBITS GLUCOSE-INDUCED INSULIN RELEASE BY INACTIVATING GLUCOKINASE

I MIWA, S MITSUYAMA, J OKUDA Department of Clinical Biochemistry, Faculty of Pharmacy, Meijo University, Nagoya, Japan

2-Cyclohexene-l-one (CHX) was reported to inhibit glucose-stimulated insulin release by decreasing reduced glutathione levels in pancreatic islets (A. Sener, et al., Biochem. Pharmacol. 35, 3701, 1986). We examined the effect of CHX on glucokinase (GK) activity, since GK has been suggested to play a key role in the regulation of glucose-induced insulin release.

Partially purified rat liver GK which has quite similar properties to islet GK was inhibited time-dependently by CHX: GK inhibition by 5 mM CHX after incubation for 5, i0, 20, and 30 min at 37°C was 44, 59, 74, and 84%, respectively. CHX also produced concentration-dependent inhibition of GK with half-maximal inhibition at about 1.7 mM when incubated for 30 min at 37°C. Inhibition of GK by CHX was not released by dialysis, indicating that the inhibition is of irreversible type. Inhibition of GK by CHX (2 mM) was markedly blocked by the substrates (20 mM) glucose and mannose but not by galactose which is not a substrate of GK. CHX (2 mM) inhibited GK in supernatants of rat liver hemogenates by 48% but little affected hexokinase activity when incubated for 30 min at 37°C. The agent (2 mM) inhibited both glyceraldehyde-3-phosphate dehydrogenase and 6-phosphofructokinase partially purified from rabbit muscle by only 15%. Preincubation of pancreatic islets with CHX (2 mM) for 30 min at 37°C resulted in severe inhibition of subsequent glucose-induced insulin release. The inhibition of insulin release was also blocked by glucose and mannose but not by galactose.

The result suggests that CHX interacts covalently with the substrate-binding site of GK and affects glucose-stimulated insulin secretion thereby.

$109

POS-001-066 ALLOXAN ACUTELY INCREASES INSULIN RELEASE, CALCIUM UPTAKE, CALCIUM MOBILIZATION AND INOSITOL TRISPHOSPHATE CONCENTRATION OF ISOLATED MOUSE PANCREATIC ISLETS

L.A.H. BORG, A. HALLBERG Department of Medical Cell Biology, University of Uppsala, Uppsala, Sweden

Studies of the selective effects of alloxan on the pancreatic islets may advance our knowledge on diabetogenic mechanisms. Earlier investigations have shown that the drug induces an irreversible inhibition of glucose-stimulated insulin release after an initial monophasic insulin response. The present study aimed at an ex- planation of this early insulin response to alloxan. By perifusion of isolated mouse islets prelabelled with SICr and pulsed (5 min) with 2 mmol/l alloxan it was found that the alloxan-dependent insulin release was not an effect of an acute B-cell destruction. Both islet net uptake of ~SCa and ~SCa efflux from prelabelled islets were increased concomitantly with the insulin response to 2 mmol/l alloxan. Verapamil (5 ~mol/l), which blocks voltage-dependent calcium channels, inhibited completely alloxan-stimulated net uptake of ~SCa but had only a limited effect on ~SCa efflux and alloxan-stimulated insulin release. Dantro- lene (25 ~mol/l), which may block intracellular mobilization of calcium, inhib- ited both net uptake of ~SCa and efflux of ~SCa and insulin release in response to alloxan. When prelabelled islets were perifused with a calcium-free medium, alloxan caused an acute increase of ~SCa efflux as well as an insulin response. Alloxan-treatment (2 mmol/l) for 5 min of isolated islets prelabelled with [3H]inositol induced a significantly increased concentration of [~H]inositol trisphosphate in the islets. The present data suggest that pancreatic islet B- cells are equipped with a stimulus-secretion coupling mechanism, which can acutely be activated by alloxan, and which includes an increased influx and in- tracellular mobilization of calcium.

POS-001-067 PERSISTENT IMPAIRMENT OF THE INSULIN RESPONSE TO GLUCOSE AFTER STREPTOZOTOCIN EXPOSURE: STUDIES WITH GRAFTED ISLETS AND ISLETS MAINTAINED IN CULTURE.

DL EIZIRIK, S SANDLER, O KORSGREN, L JANSSON, A ANDERSSON Department of Medical Cell Biology, Uppsala University, Uppsala, Sweden.

An understanding of the molecular mechanisms behind the functional responses of the B-cells after cytotoxic damage may give valuable insights into the pathogenesis and early treatment of type i diabetes mellitus. We have recently observed a preferenlial reduction of insulin productior in mouse pancreatic islets maintained in culture after in vitro exposure to streptozotocin (SZ). In order to evaluate the in vivo relevance of these findings, two set of experiments have been performed. First, mouse pancreatic islets were exposed in vitro to 2.2 mM SZ or vehicle ~lone, cultured for 6 days and fina]ly grafted under the renal capsule of normoglycemic nude mice. Two weeks after transplantation there was no difference in the DNA content between the two groups of grafted islets, but the insulin content was decreased by 40% after SZ. The insulin release of the grafts, during perfusion of the kidney with 16.7 mM glucose, was decreased by 70% in the SZ group, whilst perfusion with 16.7 mM gluco~.e plus 5 mM theophylline was able to partly restore this inhibition. In the second set of experiments, NMRI mice were injected intravenously with 160 mg/kg SZ or vehicle alone and their islets were isolated 15 min after the injections. After 6 days in culture, there was no decrease in DNA content, but the insulin content was decreased by 30% in the SZ-exposed islets. These islets also showed a 60% decrease in the insulin response to 16.7 mM glucose, partly counteracted by incubation with 16.7 mM glucose plus 5 mM theophylline. The present in vivo and in vitro observations suggest that after a cytotoxic injury there remains a population of partially damaged B-cells, with a severly impaired ability to reccgnize glucose as an insulin secretagogue. Furthermore, this defective insulin release is not related to chronic exposure of the islets to high glucose environments.

$110

POS-001-068 GROWTH INHIBITION AND STIMULATION OF HORMONE SECRETION IN POLYAMINE-DEPLETED RAT INSULINOMA CELLS.

A SJOHOLM, N WELSH, C HELLERSTR~M Department of Medical Cell Biology, Uppsala University, Uppsala, Sweden.

In view of the well established close linkage between the naturally occurring polyamines putrescine, spermidine and spermine and rapid cell proliferation we have investigated the functional significance of polyamines for replication and hormone secretion in clonal rat insulinoma cells (RINm5F). For this purpose cells were cultured with 1 mM DL-e-difluoromethylornithine (DFMO), a specific and irreversible inhibitor of ornithine decarboxylase (ODC), the rate limiting enzyme in the polyamine biosynthetic pathway. Exposure of cells to DFMO for 4 days resulted in a profound reduction of ODC activity (>90%) and polyamine contents, the decreases in putrescine and spermidine being the most prominent, whereas the addition of small amounts of putrescine (50 ~M) during culture elevated the intracellular content of putrescine 5-7-fold and suppressed ODC activity by 50%. DFMO-exposure did not alter the endogenous activity of the Ca2+-phospholipid-dependent protein kinase C as assessed by immunoprecipitation of a 32p-labelled specific substrate for the C-kinase. Neither were any differences in the endogenous activity of the polyamine-dependent casein kinase II detected between control cells and DFMO-treated cultures when assayed similarly as the C-kinase. Polyamine depletion resulted in a pronounced inhibition of the cellular proliferative activity (>90%) resulting in a 3-fold increase in the population doubling time. Moreover, it was found that the intracellular ATP content was increased 2-fold in DFMO-treated cells. The inhibitor did not affect the insulin mRNA content but evoked a 2-3-fold increase in the cellular insulin content and increased the amount of insulin released by approximately 70% in a glucose-insensitive manner. The finding of a constant insulin mRNA content may reflect the dual effect exerted by DFMO on the contents of polyamines and ATP. Putrescine, when added together with DFMO, completely counteracted the effects elicited by DFMO on all parameters studied, thus confirming the specificity of action of the inhibitor. We conclude that polyamines regulate the rapid proliferation of RINm5F cells and that inhibition of polyamine biosynthesis results in an increased insulin content and -release, without affecting the protein kinase C activity, in parallel with an increased ATP content thus supporting a role for ATP in insulin secretion of these cells.

POS-001-069 STIMULATION OF C1

SECRETION COUPLING

PERMEABILITY BY GABA TRIGGERS B-CELL STIMULUS-

J SEHLIN, P-E SANDSTROM

Department of Histology and Cell Biology, University of Ume&, Ume&, Sweden

Previous studies have shown that D-glucose increases Cl- permeability and have suggested that change in C1 permeability may be a factor in the regulation of voltage-controlled calcium uptake and insulin release. Gamma-aminobutyric acid (GABA) is an important effector in the central nervous system by gating chloride channels We therefore tested the effect

36 - 45 2+ ~ • . of GABA on C1 efflux and Ca uptake in mxcrodlssec~d B-cell-rich islets from ob/ob-mice. GABA (I-100 ;JM) increased basal Cl- efflux (0-3 mM D-glucose)in a dose-dependent manner with apparent maximum at 50 ~M. Studies on the interaction between the stimulations by GABA (50 ~LM) and D-glucose (6 or 20 mM) showed a complete overlap, suggesting that the two compounds may affect the same C1 permeabilitY3~echanism. The GABA receptor agonist N-glycylmuscimol also stimulated CI- efflux but with an apparent ~ximum effect at 10 ~JM. GABA (50 ~LM) potentiated the glucose- induced =~Ca~- uptake in the presence of 10 or 20 mM D-glucose but had no effect on the uptake at 0-6 mM sugar. The data provide evidence that the pancreatic B-cells possess GABAergic receptors and GABA-gated chloride channels, which may participate in controlling calcium uptake.

S111

POS-O01-070 EFFECT OF GLYBURIDE (GLB) ON DESENSITIZATION (DSZ) OF INSULIN (INS) SECRETION FBOM RAT ISLETS.

J. L. BOLAFFI, L. BRUNO, G. G. BODD, N. M. BOSTICK, G. M. GBODSKY. Metabolic Research Unit, University of California, San Francisco, CA USA.

Constant glucose (7-22 r~) stimulates islet INS secretion through a rising phase which peaks at 135-160 min then declines by 6-8 hr to a new desensitized level at 20% of peak secretion. We investigated the effect of the sulfonylurea, GLB, on DSZ of glucose-induced INS secretion using computer-controlled flow-through perifusion or batch incubation of freshly isolated rat islets. Perifusion for 20 hr at 5 mM glu- cose + 5 ng/ml GLB stimulated peak INS secretion to 6.2~0.7 ng/islet/hr and total insulin secretion 7-fold to 58.9+_11.0 ng/islet/20 hr vs glucose alone. Secretion desensitized to 1.9+_0.1 ng/islet/hr, a level sustained many-fold higher than with 5 mM glucose alone. However, in contrast to all stimulatory glucose levels (5, 7, ii, and 20 mM), time to peak secretion was prolonged by 95 ndn and time to 50% DSZ post- peak by 94 min. INS response to a 20 mM glucose pulse at hr 20 was reduced by 50% in GLB-perifused islets. Batch-incubated islets showed similar time course and amounts of INS secreted, with little difference seen at 5-2500 ng/ml GLB (0.25 ng/ml was non- stimulatory). As an index of absolute INS synthesis, total INS recovery was calcu- lated from total INS secreted and in islets before or after incubation. Although an effective potentiator of INS secretion, GLB had no effect on INS synthesis in 5 mM glucose with a total INS recovery of 102+6.5% of initial content vs 120 166% with 7 22 mM glucose. Fractional secretion rates fell off less over time in GLB-treated islets (5.9+0.5%/hr at hr 3 to 3.0+0.4% at h 22) than the 80% reduction found at stimulatory glucose levels. Thus i) GLB potentiated and sustained INS secretion at low glucose for 24 hrs, 2) delayed onset of DSZ, 3) sustained more efficient frac- tional secretion despite no effect on insulin synthesis, but 4) did not prevent even- tual DSZ of secretion

POS-O01-071 RECIPROCAL RELATIONSHIPS BETWEh~N THE ADENYLA~ CYCLASE SYSTEM AND SODIUM/PCq'ASSIUM DEPENDENT ADENOSINE ~q~IPHOSPHATASE IN ISOLATED RAT ISLETS

P q~/NG, SR Lh~VIN, DG JOHNSON, G PAI. Wadsworth VA Medical Center and University of California-Los Angeles, c2~, USA

Enhancement of adenyiate cyclase (AC) activity accompanies insulin secretion. Our previous experiments (J Clin Invest 62:692, 1978, Diabetes 36:1448, 1987) have suggested that suppressior of islet Na,K-ATPase activi~ is also involved in ~is process. This suggests that a recip- rocal relationship might be demonstrated between these two phosphoprotein syst~ns (Na,K-ATPase and AC) in rat pancreatic islets. Initially, we examined the effect of theophylline (Theo) upon islet membrane Na,K-ATPase (pellet from a 35,000xg - 30 min. ultracentrifugate and subsequent I0 min. incubation with 3 ,~. ATP at 37oc) . Theo caused a dose-dependent decrease in enzyme activity with half-maximal inhibition at 0.5 raM. (n=5; Na,K-ATPase=umol Pi/mg prot/hr) (*=significant vs 0 Theo) : Theo (,M) 0 0.5 2.0 5.0 N'a,K-ATPase 5.1+ 0.2 2.~+0.2 * 1.8+0.1 * 0.0+0.~ ~

Dibutyryl-cAMP (D,I n~) and caffeine (3,1 n~4) also inhibited Na,K-ATPase activity. (*=significant vs buffer):

A~dition (n) Buffer (6) D (6) C (7) Na,K-ATPase 5.05+0.26 2.52+0.38 * 2.36+0.55 *

Kinetic data showed noncompetitive inhibition by 1 mM Theo, with an apparent Km (ATP) of 370 uM. However, we observed competitive inhibition of Na,K-ATPase activity by 1 n#4 D: apparent Km (ATP) increased from 322 uM without D to 770 uM with D. Thus, phosphodiesterase inhibitors and a cAMP analogue decrease islet Na,K-ATPase activity, indicating reciprocity between AC and Na,K-ATPase with regard to ATP disposition within the cell.

Sl12

POS-001-072 ROLE OF PHOSPHOPROTEIN PHOSPHATASES IN THE b~_f~JLNISM OF INSULIN RELEASE IN RIN-m 5F CELL LINE.

P MASr~.TO I , L ROTONDABO 2 , AA DE PAOLI-ROACH 2 I. Istituto di Patologia Generale, University of Pisa, Pisa, Italy 2. Department of Biochemistry, Indiana University School Medicine, Indianapolis, In., U.S.A.

The involvement of protein phosphorylation in the regulation of insulin secretion has been sug- gested and there is evidence of stimulation of a number of protein kinase activities by some in- sulin secretagogues. In centrast, protein phosphatase activities have not yet been studied in insulin-secreting cells. Therefore, we decided to investigate the role of protein phosphatases and their heat-stable inhibitors I and 2, in RIN-m 5F cells during stimulation of insulin re- lease. Depolarizing concentrations of KCI or fuel molecules, such as glyceraldehyde, i-ketoiso- caproate and glutamine, in combination with the phosphodiesterase inhibitor isobutylmethylxan- thine (IBMX), elicited a good secretory response from these cells. Although no significant change in protein phosphatase activity was observed in stimulated cells, however the activity of inhib- itor I (I-I) increased by 25% following IBMX treatment and decreased by 30% during incubation with 30 n~M KCI, in respect to controls. These effects on I-I activity are comparable to those re- ported in muscle in correlation with changes in cAMP levels. Recently, we have isolated cDNA clones encoding inhibitor 2 (I-2). A cDNA fragment containing all the coding region for I-2 was inserted in an expression vector, ~Mt-neo, in both the 5'-->3' and 3'-->5' orientations. The I-2-pMt-neo was introduced into RIN-m 5F cells by the calcium phosphate transfection method. Stable transformants were selected in the presence of geneticin (G418). Analysis of genomic DNA by Southern hybridization revealed that multiple copies of the vector were integrated in the cellular [INA of several transformants. In addition, I-2 transcripts were detected by dot-blot analysis of cytoplasmic extracts and appeared to be enhanced by induction with zinc chloride. Insulin content and secretory abilities of uninduced transfected cells were similar to those of control cells. Studies are in progress to determine whether changes in the insulin responsive- ness occur in transformants that over-express I-2 or in transformants where expression of endo- genous I-2 is blocked.

POS-001-073 INSULIN RESPONSES TO THE ANOMERS OF HEXOSES IN A RAT MODEL OF NON-INSULIN-DEPENDENT DIABETES.

H NIKI, A NIKI, T HASHIOKA, I NIKI* Department o f In terna l Medicine, Aichigakuin Un ivers i t y and *Third Department of In terna l Medicine, Un ive rs i t y o f Nagoya, Nagoya, Japan.

The ~ anomer of D-glucose or D-mannose is known to be more potent than the B anomer in s t imu la t - ing i nsu l i n release from the pancreas of normal animals. We have examined i f the pancreat ic B ce l l in non- insul in-dependent diabetes (NIDD) d iscr iminates the ~ and B anomers of D-glucose or D-mannose, using the i so la ted perfused ra t pancreas.

A ra t model of NIDD was induced by s t rep tozo toc in i n j e c t i o n (90 mg/kg) at 2 days of age. Oral glucose to lerance (2g/kg body weight) of the s t rep tozo toc in - t rea ted rats at 8 - I0 weeks of age was m i l d l y d i abe t i c , the mean maximal blood glucose leve l reaching approximately 16.7 mmol/ l . Perfusion experiments were car r ied out in the fed s ta te w i th in 2 weeks a f t e r the glucose to lerance tes ts . The ~ or B anomer of hexoses was rap id l y d issolved in ice-co ld basal medium j us t before use, and the so lu t ion was warmed up to 37°C immediately before reaching the pancreas in a heat ing co i l immersed in a water bath. Insu l in release from the pancreas of the d iabe t i c rats in response to I0 mmol/l ~-D-glucose was markedly impaired, whi le i nsu l i n response to I0 mmol/l B-D-glucose in the d iabe t i c pancreas was only s l i g h t l y reduced as compared to that in the cont ro l pancreas. S imi la r resu l ts were obtained when the pancreas was perfused wi th the ~ or 8 anomer of D-mannose (16.7 mmol/ l ) . Insu l in response to to lbutamide (0.15 mg/ml) in the d iabe t ic pancreas was not impaired comparing to tha t in the contro l pancreas. Add i t ion of to lbutamide to the perfusates did not improve impaired d i sc r im ina t ion of two anomers of D-glucose in the d iabe t i c pancreas.

These f ind ings suggest tha t the pancreat ic B-cel l dysfunct ion in the NIDD ra t model is re la ted to the mechanism by which the B ce l l d iscr iminates the ~ and 8 anomers of the hexoses as a s t imu la to r of i nsu l i n re lease.

S l 1 3

POS-001-074 PREGNANCY AND INSULIN RELEASE: INVOLVEMENT OF cAMP AND PROTEIN KINASE C ~N THE FUNCTION OF B-CELLS OF PREGNANT RAT PANCREAS

K.TANIGAWA, S.OHGUNI, Y.MASAKI, S.TSUCHYAMA, M.KAWAGUCHI and Y.KATO, First Division, Department of Medicine, Shimane Medical University, Shimane,Japan

In order to elucidate intracellular mechanisms of exaggerated insulin release(It) in gestation, the isolated pancreses in pregnant rats (P;21st day of gestation) and age-matched nonpregnant rats(N) were perfused and Ir as well as insulin content (Ic) of the perfused pancreas was determined. In the presence of 16.7 mM glucose, forskolin(FSK,luM), an activator of adenylate cyclase, potentiated Ir in both states. However, percent changes in Ir were not different between two groups (200 ± 12 vs. 195 ± 16%). When the concentrations of FSK was increased to 10uM, PSK showed--a greater effect on B-cells in N rats than in P rats. The magnitude of Ic increase caused by FSK(IuM) was 5-fold in N rats and 2.5-fold in P rats(Table). Thus, the cAMP system might be fully activated in B-cells. In contrast, the effect of phorbol ester TPA, an activator of protein kinase G, was less potent in N rats than in P rats, indicating that the activity of protein kinase C in B-cells might not be fully elevated in late gestation. These data suggest that cAMP may play more pivotal role than protein kinase C in exaggerated fuel-mediated Ir from pregnant rat pancreas.

n Forskolin(uM) TPA(nM) Ir(mU) Ic(mU) Nonpregnant 5 0 0 1 9 ± 0.3 a 235.0 ± 27.0

4 1 0 3 7± 0.3 1164.4 ± 88.5 3 0 20 2 8 ± 0.6 --

Pregnant 5 0 0 5 7± 1.0" 615.5± 65.2** 4 i 0 ii 4± 0.7** 1671.5± 158.3 4 0 20 29 4± 4.8 --

*P< 0.05, **P < 0.005 vs. corresponding values in nonpregnant, a :mean ± SE.

POS-001-075 EFFECT OF CELLULAR cAMP LEVELS ON INSULIN SECRETION DUE TO GLUCOSE, POTASSIUM AND A 23187: A PERIFUSION STUDY

S KAGAWA, K YAMASHITA, S WAKABAYASHI, A MATSUOKA Department of Clinical Pathology and Clinical Laboratory, Hyogo College of Medicine, Hyogo, Japan With a view to understanding the mechanism by which glucose stimulates insulin secretion, perifusion was conducted on pancreatic monolayer cultures of the neo- natal rats. The response of B cells to 16.7 mM glucose was menophasic and was not sustained during a continued glucose stimulus. The stimulus of 16.7 mM glu- cose plus imM 3-isobutyl-l-methylxanthine (IBMX) resulted in a biphasic response. In perifusions performed in the absence of glucose, B cells responded in a mono- phasic fashion to stimulus of either 25 mM K + or I0 uM A 23187 alone. In the presence of IBMX and high K + or A 23187, a remarkable increase in secretion in the second phase was induced, the magnitude in response being as high as that of B cells stimulated with glucose plus IBMX. Furthermore, when B cells were changed to medium with high K ÷ or A 23187 after an initial 14-min stimulation with high K + plus IBMX, insulin secretion in the second phase increased to levels observed in cells that had been stimulated with high K ÷ plus IBMX throughout. This increase was abolished by addition of 1 uM epinephrine or 200 uM verapamil. In addition, when exposed to a linear concentration gradient of 6-30 mM K + or 0-20 uM A 23187, B cells responded in a dose-dependent manner to secretagogues together with IBMX but not in its absence. After exposure for 3 days to medium with 16.7 mM glucose plus i0 uM iodoacetic acid to kill fibro- blasts selectively, cAMP in cells and Ca influx were determined;addition of IBMX increased the cAMP content (10.6 to 15.3-fold), and Ca influx, measured by 45Ca, was increased in cells stimulated with high K ÷ or A 231S7 (2.4 to 3.8- fold). These results suggest that cAMP may potentiate insulin secretion induced by a rise in the cellular calcium concentrations.

S114