0013-7227/90/1261-0253$02.00/0 Endocrinology Copyright© 1990 by The Endocrine Society
Vol. 126, No. 1 Printed in U.S.A.
Chronic in Vivo Hyperglycemia Impairs Phosphoinositide Hydrolysis and Insulin Release in Isolated Perifused Rat Islets* WALTER S. ZAWALICH, KATHLEEN C. ZAWALICH, GERALD I. SHULMAN, LUCIANO ROSSETTIf Yale University School of Nursing (W.S.Z., K.C.Z.), New Haven, Connecticut 06536-0740; and Yale University School of Medicine (G.I.S., L.R.), New Haven, Connecticut 06510
ABSTRACT. We examined the effect of chronic hyperglycemia on phosphoinositide hydrolysis and insulin secretion in isolated perifused rat islets. Rats were infused for 44 h with 40% dextrose in order to raise and maintain the plasma glucose concentration at 350 mg/dl. Control animals were infused with equiosmolar amounts of mannitol. In vivo insulin secretion and rates of glucose disposal were monitored throughout the study. At the end of the infusion, islets were collagenase isolated, and phosphoinositide (PI) hydrolysis (assessed by measuring the increment in [3H]inositol efflux as well as labeled inositol phosphates) and insulin output in response to a 20-mM glucose challenge were quantitated. Plasma insulin concentration and in vivo glucose disposal rates decreased significantly, by 47% and 35% respectively, after 6-8 h of hyperglycemia. In islets perifused immediately after isolation, prior in vivo hyperglycemia markedly altered the pattern of insulin output in response
to 20-mM glucose challenge. Compared to mannitol infusion, 20 mM glucose stimulation resulted in an exaggerated first phase insulin secretory response (1121 ± 88 vs. 467 ± 75 pg/isletsmin) and a blunted second phase insulin secretory response (392 ± 90 vs. 1249 ± 205 pg/islet-min). In islets prelabeled with myo[2-3H]inositol for 2 h, PI hydrolysis, particularly [3H]inositol efflux in response to glucose stimulation was also reduced (0.28 ± 0.03%/min) compared to that in mannitol-infused animals (0.53 ± 0.08%/min). Two hours of preincubation in a low glucose medium (2.75 mM) were able to completely reverse the islet defect in both PI hydrolysis and insulin secretion. Our results demonstrate that chronic in vivo hyperglycemia impairs PI hydrolysis in perifused rat islets and suggest that this defect accounts in part for the abnormal pattern of glucoseinduced insulin secretion. (Endocrinology 126: 253-260, 1990)
D
insulin secretion induced by chronic glucose infusion. Islets were isolated from rats infused for 44 h with either high glucose or an osmotically equivalent mannitol load. The present results demonstrate that chronic in vivo hyperglycemia results in defective in vitro PI hydrolysis and insulin secretory responses to a glucose challenge in isolated perifused rat islets obtained from these animals.
EFECTS in glucose-induced insulin secretion are a common feature of noninsulin-dependent diabetes mellitus in man (1-7). Over the last decade, evidence has accumulated to indicate that chronic sustained hyperglycemia exerts a deleterious effect on /3-cell function (814). Furthermore, reversal of the hyperglycemia is accompanied by a correction of the insulin secretory defect (15-17). Recently, Leahy and co-workers (13, 14) have shown that 48 h of marked hyperglycemia severely impairs glucose-induced insulin secretion from the isolated perfused pancreas. However, little information exists regarding the nature of the biochemical lesion responsible for this aberrant insulin secretory response pattern. The present study was designed to examine the role of phosphoinositide (PI) hydrolysis in the alteration of
Materials and Methods Animal preparation Male Sprague-Dawley rats (Charles River Co., Wilmington, MA), weighing between 300-350 g, were used in all studies. Five to 7 days before the infusion studies rats were anesthetized with phenobarbital (50 mg/kg BW), and indwelling catheters were inserted into the internal jugular vein and carotid artery, so that the animals could be studied in the awake unstressed state (17, 18). The catheters were filled with heparin (40 U/ ml)-polyvinylpyrollidone (160 g/100 ml) solution, sealed, and tunneled sc around the side of the neck to the back of the head. The catheters were externalized through a skin incision in the back of the neck. Within 48 h all animals had recovered from surgery and were eating normally. Postoperatively, rats were
Received July 20,1989. Address requests for reprints to: Dr. Walter Zawalich, Yale University School of Medicine, 855 Howard Avenue, P.O. Box 9740, New Haven, Connecticut 06536-0740. * This work was supported by the American Heart Association and the Juvenile Diabetes Foundation. t Present address: Diabetes Section, Department of Medicine, University of Texas Health Science Center, 7703 Floyd Curl Drive, San Antonio, Texas 78284-7870. 253
254
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE
housed in an environmentally controlled room with a 12-h light, 12-h dark cycle and allowed free access to standard rat chow (Purina, Richmond, IN) and water. Experimental protocol Five to 7 days postsurgery, a solution containing 40% glucose (wt/vol) or equimolar amounts of mannitol was infused into the arterial catheter for a total of 44 h using a Harvard servocontrolled pump (Harvard Apparatus, Inc., South Natick, MA) and two minature infusion swivels (Spalding Medical Products, Arroyo Grande, CA). Throughout the study, the rats were allowed to move freely within the confines of a large cage and had free access to food and water. The venous catheter was used for blood withdrawal for determination of plasma glucose and insulin concentrations. The 40% glucose solution was administered to acutely raise the plasma glucose concentration by approximately 250 mg/dl above baseline. The plasma glucose concentration was subsequently held constant at this hyperglycemic plateau by the adjustment of a variable glucose infusion, as previously described (17,18). Isolated perifused islet studies At the end of the 44-h infusion rats were anesthetized with phenobarbital (50 mg/kg BW), and islets were isolated by collagenase digestion using Hanks' solution. After hand picking, the islets were placed in a medium identical to that used during the perifusion, except for the presence of 126 mg/dl (7 mM) glucose. Some islets were directly perifused after isolation. In other experiments the islets were incubated for 2 h in a solution supplemented with myo-[2-3H]inositol. In these studies, 10 juCi myo-[2-3H]inositol were added to 250 n\ perifusion medium supplemented with either 2.75 or 7 mM glucose. Two hundred microliters of this solution were then added to the islets. After labeling, the islets were washed with 5 ml unlabeled medium and then perifused. It should be noted that the time between the addition of collagenase to begin the pancreatic digestion and the start of the perifusion or labeling with [3H] inositol averaged less than 60 min. Islets were perifused at a flow rate of 1 ml/min for 30 min to establish stable basal secretory rates. After this stabilization period, the islets were subjected to various protocols, indicated in the figure legends and in Table 2. Effluent samples were collected and analyzed for insulin content (19), using rat insulin as standard (lot 615-D63-12-3, Eli Lilly Co., Indianapolis, IN) and, when appropriate, [3H] inositol content. The sensitivity of the RIA for insulin determination was 0.195 ng/ml; the intraassay coefficient of variance was less than 9%, and the interassay coefficient of variation was less than 18%. Previous results from this laboratory have demonstrated that approximately 90% of the effluent radioactivity from glucose-stimulated islets is from [3H]inositol (20). Inositol phosphates were extracted after the perifusion with 10% perchloric acid and separated on anion exchange columns (AG1-X8, 200-400 mesh formate form, Bio-Rad Laboratories, Richmond, CA) using methods previously described (20-22). Briefly, the islets still attached to the nylon filters were retrieved and placed in small glass vials. Perchloric acid (0.4 ml; 20%) and 0.4 ml perfusion medium
Endo • 1990 Vol 126« No 1
were immediately added. After 30 min on ice the extract was neutralized with 0.28 ml 6 N KOH and then added to the columns. These columns were prepared by adding (to achieve a length of 3 cm) anion exchange resin to Pasteur pipettes. Further additions to the column included 10 ml water and 5 ml 5 mM borax-60 mM sodium formate. Elution of the inositol phosphates was accomplished by the sequential addition of 10 ml 0.1 M formic acid-0.2 M ammonium formate (inositol 1phosphate), 0.1 M formic acid-0.4 M ammonium formate (inositol 1,4-bisphosphate), and 0.1 M formic acid-1 M ammonium formate (inositol 1,4,5-trisphosphate plus inositol 1,3,4-trisphosphate). Aliquots (0.4 ml) of the eluate were then analyzed for radioactive contents. The filters, still containing the islets, were also counted. Total [3H]inositol incorporation was calculated as the amount of effluent radioactivity collected from 2860 min of the perfusion, total label in [3H]inositol phosphates, and label present in the islets after extraction of the inositol phosphates with perchloric acid. Reagents Hanks' solution (supplemented with 7 mM glucose) was used for the islet isolation. The perifusion medium consisted of 115 mM NaCl, 5 mM KC1, 2.2 mM CaCl2, 24 mM NaHCO3, and 0.17 g/dl BSA. Other compounds were added as indicated, and the solution was gassed with a mixture of 95% O2-5% CO2. The [125I] insulin used for the insulin assay was purchased from New England Nuclear (Boston, MA), and the myo-[2-3H]inositol was from Amersham (Arlington Heights, IL). BSA (RIA grade), glucose, mannitol, formic acid, as well as the salts used to make the Hanks' solution and perifusion medium were purchased from Sigma Chemical Co. (St. Louis, MO). Ammonium formate was purchased from Fisher Scientific (Fairlawn, NJ). Statistics Statistical significance was determined using Student's t test for unpaired data and, where appropriate, analysis of variance. P < 0.05 was taken as significant. Values presented are the mean ± SE of the specified number of observations.
Results Body weights, plasma glucose, and insulin concentrations (Table 1 and Fig. 1) Body weights did not change significantly during the 2 days of the infusion in either the glucose- or mannitolinfused rats. In the mannitol-infused animals, there was no significant increment above basal in plasma glucose (138 ± 3 mg/dl) and plasma insulin concentrations (2.9 ± 0.4 ng/ml) during the infusion period. In the glucoseinfused animals, the plasma glucose concentration was 346 ± 7 mg/dl and remained constant throughout the study (Fig. 1, top panel). The mean plasma insulin concentration was 23.9 ±2.1 ng/ml in the first 6 h (0-6 h) and dropped to 12.7 ± 0.8 ng/ml in the last 34 h of the study (10-44 h; Fig. 1, top panel). The glucose infusion rate necessary to maintain the plasma glucose concen-
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE
255
TABLE 1. Body weight, plasma insulin (I), plasma glucose (G), and mannitol or glucose infusion rates (IR) during the in vivo study BW (g)
I (ng/ml)
G (mg/dl)
IR (mg/kg-min)
0 44
322 ± 19 326 ± 21
2.9 ± 0.2 3.0 ± 0.3
138 ± 4 149 ± 3
38.0
0 0-6 10-44
328 ± 21
3.0 ± 0.2 23.9 ± 2.1 12.7 ± 0.8
139 ± 3 344 ± 10 348 ± 13
51.6 ± 1.0 35.4 ± 1.5
Time (h)
Infusion Mannitol
Glucose
334 ± 18
PLASMA GLUCOSE AND INSULIN CONCENTRATION 400
1200
r
1000
islet
"e
800
ASE
2
600
GLUCOSE INFUSION RATES MEAN GLUCOSE INFUSION 0-6hr = 51.6 + 1.1 10-44hr = 33.53 i i . 4
co 400
200 36
40
44
TIME (hours) FlG. 1. Effect of chronic glucose infusion on plasma glucose and plasma insulin concentrations (top) and glucose infusion rates {bottom). Rats were infused for 44 h with a variable 40% glucose solution to raise and maintain the plasma glucose concentration at approximately 350 mg/dl. The plasma glucose (milligrams per dl) and insulin (nanograms per ml) concentration (top panel) and the glucose infusion rate (milligrams per kg/min; bottom panel) are shown. The continuous line in the bottom panel represents the mean, and the interrupted line represents the SE.
tration constant at the desired hyperglycemic plateau was 51.6 ± 1.0 mg/kg-min in the first 6 h (0-6 h) and dropped to 33.4 ± 1.5 mg/kg-min during the last 34 h (10-44 h) of the study (Fig. 1, bottom panel), displaying a behavior similar to that observed for the plasma insulin concentrations. Perifused islet studies (Figs. 2-4 and Table 2) Islets isolated from mannitol-infused rats displayed a brisk biphasic insulin secretory response when directly perifused immediately after isolation (Fig. 2). Peak first phase release averaged 467 ± 75 pg/islet-min, while
° '30 40 50 60 70 DURATION OF PERIFUSION (min) FIG. 2. Chronic hyperglycemia alters /3-cell insulin secretory response to glucose. Groups of 12-15 islets (n = 6 for each condition) were isolated from either glucose-infused or mannitol-infused rats and perifused immediately. For 30 min they were perifused with 2.75 mM glucose (G275) to establish basal stable insulin secretory rates. They were then stimulated for 40 min with 20 mM glucose. The mean ± selected SEs are given. This and subsequent figures have been corrected for the dead space in the perifusion system (~2.5 ml or 2.5 min with a flow rate of 1 ml/min).
release measured 35-40 min after the onset of stimulation averaged 1249 ± 205 pg/islet-min. In contrast to this response, the insulin secretory response of islets isolated from glucose-infused animals and immediately perifused displayed an abnormal profile. Peak first phase release was dramatically amplified, reaching an average value of 1121 ± 88 pg/islet-min. Second phase release was attenuated, reaching an average value of only 392 ± 90 pg/islet-min after 35-40 min of stimulation with 20 mM glucose (Fig. 2).
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE
256 0.8
1000 r '2.75
620 rf°h HYPERGLYCEMIC L (REVERSAL)
'2.75
/HYPERGLYCEMIC D (REVERSAL)
0.6
Endo • 1990 Vol 126 • No 1
r
'E 800 MANNITOL
< 600 0.4
•
•MANNITOL
HYPERGLYCEMIC
CO LLJ
_l UJ
0.2
HYPERGLYCEMIC
cc
30 40 50 60 DURATION OF PERIFUSION ( m i n ) FIG. 3. Chronic hyperglycemia results in altered, but reversible, patterns of [3H]inositol efflux from prelabeled islets. Groups of islets were isolated from glucose-infused or mannitol-infused rats. For 2 h they were incubated in a medium supplemented with myo-[2-3H]inositol to label their PI. The glucose level used during this 2-h period was either 7 mM (hyperglycemic or mannitol) or 2.75 mM (hyperglycemic reversal). All islets were then washed, then perifused for 30 min with 2.75 mM glucose (G2.75) and for an additional 30 min with 20 mM glucose. At least four experiments were performed under each condition. Fractional efflux rates of [3H]inositol, given as percent per min, were calculated according to the method of Borle et al. (44).
The responses depicted in Figs. 3 and 4 were obtained after a 2-h incubation period to label islet PI. In response to glucose stimulation, islets (labeled in the presence of 7 mM glucose) from mannitol-infused rats demonstrated a prompt and significant increase in the fractional efflux rate of [3H]inositol (Fig. 3). After 30 min of stimulation, this value reached 0.53 ± 0.08%/min. Efflux rates from islets maintained for 60 min with 2.75 mM glucose alone averaged 0.09 ± 0.03%/min (results not shown). In contrast to this response, islets isolated from glucose-infused animals and labeled in the presence of 7 mM glucose, demonstrated a small and attenuated increase in [3H]inositol efflux rates. After 30 min of stimulation with 20 mM glucose, the fractional efflux rate of [3H]inositol averaged only 0.28 ± 0.035%/min, a value significantly (P < 0.05) different from that in islets isolated from the mannitol-infused rats. Most interestingly, labeling of islets isolated from glucose-infused animals in the presence of 2.75 mM glucose was accompanied by an amplified [3H]inositol efflux response (Fig. 3). After 30 min with 20 mM glucose, this efflux value averaged 0.79 ± 0.11%/min. In islets isolated from mannitol-infused rats, labeling with [2-3H]inositol resulted in the incorporation of 27,175 ± 1,392 cpm/40 islets. Islets isolated from glucose-infused animals incorporated
400
200
0 L,
_ 30
40
50
60
DURATION OF PERIFUSION (min) FlG. 4. Chronic hyperglycemia results in altered, but reversible, patterns of insulin output from prelabeled islets. Groups of islets were isolated from mannitol- or glucose-infused rats, then labeled for 2 h as indicated in Fig. 3. They were perifused for 30 min to establish stable basal insulin release rates and then stimulated for 30 min with 20 mM glucose. Note the difference in insulin response characteristics in islets from hyperglycemic animals if they are perifused immediately after isolation (Fig. 2) or subjected to a 2-h labeling period.
30,215 ± 1,885 cpm/40 islets. In response to 20-mM glucose stimulation, islets isolated from mannitol-infused rats released 2,144 ± 415 cpm/40 islets over the final 20 min of stimulation. Islets isolated from glucoseinfused animals and labeled in the presence of 7 mM glucose released 1,132 ± 322 cpm/40 islets over this time. In contrast, islets isolated from glucose-infused rats but labeled in the presence of 2.75 mM glucose released 3,321 ± 469 cpm/40 islets over this time. Results similar to the [3H]inositol efflux findings were obtained when the insulin release profiles were monitored from these same islets (Fig. 4). A biphasic insulin secretory response was noted from the mannitol-infused group. An abnormal insulin secretory profile, again accompanied by a dramatic first phase secretion and a reduced second phase response, was observed in glucoseinfused islets labeled in the presence of 7 mM glucose. Islets isolated from glucose-infused animals and labeled in the presence of 2.75 mM glucose displayed normal, albeit slightly amplified, temporal patterns of insulin output. At least when glucose (2.75 or 20 mM) is used as the stimulant, average fractional rates of [3H]inositol efflux measured during the last 10 min of the perifusion correlated significantly with total insulin release during this time (r = 0.925; P < 0.02). At the termination of the perifusion, labeled inositol phosphate levels were measured (Table 2). Compared to
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE TABLE 2. Influence of chronic mannitol or glucose infusions on inositol phosphate levels in isolated rat islets Labeling protocol
Perifusion protocol (min)
IPi
IP2
IP,
Mannitol infusion 1. G 7 2. G7 Glucose infusion 3. G7 4. G7 5. G275 6. G275
G2.75 481 ± 26° 121 ± 2 3 73 ± 19 (60) G2.76 -> G20 1066 ± 124 343 ± 50 184 ± 21 (30) (30) G 275 (60) 625 ± 69 G2.75 -+ G20 802 ± 53 (30) (30) G275 564 ± 51 (60) G2.75 -> G20 1646 ± 214 (30) (30)
179 ± 1 8 99 ± 8 309 ± 25 183 ± 29 198 ± 23 135 ± 7 936 ± 91 487 ± 51
Groups of 30-50 islets were isolated from mannitol- or glucoseinfused rats. They were incubated for 2 h in a medium supplemented with [3H]inositol to label their PI. Also present was 7 mM (protocols 1-4) or 2.75 mM glucose (G; protocols 5 and 6). After washing to remove unincorporated label, the islets were perifused for 30 mn with 2.75 mM glucose to establish basal and stable insulin secretory and [3H]inositol efflux rates. Some groups (protocols 1, 3, and 5) were maintained for an additional 30 min with 2.75 mM glucose, while the others (protocols 2, 4, and 6) were stimulated for 30 min with 20 mM glucose. After this, inositol phosphates were extracted and measured as detailed in Materials and Methods. The duration of each perifusion is indicated in parentheses below each protocol. Mean values ± SEs are given for at least 4 experiments. IPi, Inositol 1-phosphate; IP2, inositol 1,4-bisphosphate; IP3, inositol trisphosphate. Statistical analysis: protocol 1 vs. 2, P < 0.05 for all inositol phosphates; protocol 3 vs. 4, P < 0.05 for all inositol phosphates; protocol 6 vs. 5, P < 0.05 for all inositol phosphates; protocol 6 vs. 2, P < 0.05 for all inositol phosphates; protocol 6 vs. 4, P < 0.05 for all inositol phosphates. 0 Counts per min/40 islets (mean ± SE).
the levels of inositol phosphates measured in the presence of 2.75 mM glucose, islets isolated from mannitolinfused rats displayed dramatic and significant increases in the levels of all inositol phosphates measured in response to 20 mM glucose (Table 2, line 2). In contrast, islets isolated from glucose-infused rats labeled in the presence of 7 mM glucose displayed attenuated responses to 20 mM glucose .(Table 2, line 4). Most dramatic, however, was the response observed in islets isolated from glucose-infused rats but labeled in the presence of 2.75 mM glucose. Similar to the [3H]inositol efflux and insulin release results, a dramatic increase in the levels of all labeled inositol phosphates measured occurred in response to 20-mM glucose stimulation (Table 2, line 6).
Discussion A reduced or abolished insulin secretory response to glucose characterizes the /3-cell of type II diabetic subjects (1-7). Since the response to other secretagogues, such as arginine, is preserved, it has been postulated that
257
chronic exposure of the /3-cell to an elevated ambient glucose concentration causes specific desensitization to the glucose stimulus (8-14). Studies in animals support this notion, since insensitivity to an acute glucose stimulus has been demonstrated to occur in various different animal models of diabetes in vivo (17) and in vitro (814). This defect in insulin secretion can be completely reversed by correcting the chronic hyperglycemia (1417). Recently, Bolaffi and co-workers (23) have shown that prolonged in vitro exposure of different /3-cell preparations to high glucose results in a reduced pattern of the glucose-induced insulin output. Interestingly, the phenomenon seems to be reproduced by our chronic glucose infusion in vivo. After 6-10 h of sustained hyperglycemia a brisk decline of the insulin output was observed, despite constant hyperglycemic stimulation. However the biochemical basis for this desensitization process has not been defined. A recent report by Weir et al. (24) indicates that, unlike adipose cells (25), the defect caused by hyperglycemia does not appear to reside in altered transport of the hexose into the cell. Our recent in vitro studies with islets incubated in high (250 mg/dl) glucose or in low (50 mg/dl) glucose plus the gut hormone cholecystokinin have indicated that the defective insulin secretory response noted after a 2-h exposure to these compounds is paralleled by impaired PI hydrolysis (26). Because of the strategic location of these phospholipids and because of the plethora of second messenger molecules generated by their hydrolysis (27, 28), we suggested that a defect in the hydrolysis of these phospholipids might contribute to the altered insulin secretory response to a subsequent 20-mM glucose challenge. In the present report we investigated whether abnormal PI responses play any role in the aberrant patterns of insulin secretion noted after chronic in vivo exposure to high glucose, a model previously studied extensively by Leahy and coworkers (13, 14). The results suggest that this manipulation impairs, in a parallel fashion, glucose-stimulated PI hydrolysis and insulin secretion. Before discussing our findings in detail, a few comments concerning the relationship between PI hydrolysis and insulin release in islets are appropriate. Previous studies by Axen et al. (29), Clements and Rhoten (30), Best et al. (31-33), Blachier et al. (34), and Zawalich et al. (22,35,36), have demonstrated that islets [3H]inositol labeled in low (2.7-5.6 mM) glucose are responsive, in terms of increases in PI hydrolysis, to subsequent glucose stimulation. These findings stand in contrast to the studies of Rana et al. (37), who found it necessary to [3H] inositol label islets in high (16.7 mM) glucose in order to demonstrate subsequent glucose-induced increases in PI hydrolysis. In a most recent in vitro study (26), we have been unable to confirm these results. In fact, labeling in high glucose is actually detrimental to the islet, in terms
258
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE
of both PI hydrolysis and insulin release (26). The in vivo findings reported here further support the concept that hyperglycemia adversely influences both of these processes. Finally, although our results suggest that a positive correlation exists between increases in PI hydrolysis (when [3H]inositol efflux is used as the parameter of PI activation) and insulin release, this appears to be the case only when glucose is used as the stimulant. For example, if PI hydrolysis is stimulated by the enteric hormone cholecystokinin (38) or by cholinergic agonists (34, 39, 40), significant changes in [3H]inositol efflux and [3H]inositol phosphate accumulation are observed whether the glucose level is low (2.75 mM) or high (7.0 mM). However, insulin release significantly increases only at the higher glucose level. Clearly, it is possible with these other agonists to dissociate increases in PI hydrolysis from increases in insulin release. Unfortunately, the nature of the signal supplied by glucose which permits the coupling of PI hydrolysis to release under these conditions is not known, although it does appear to be metabolic in nature. The data presented in Table 1 indicate that neither the mannitol-infused nor the glucose-infused rats displayed increases in body weight despite free access to food and water during the 44 h of the study. Consequently, the impact of nutritional deprivation as a contributory factor in the observed defects has to be considered. However, a similar variation in the nutritional state seems to occur in hyperglycemic and euglycemic rats. The effects of the 44-h hyperglycemic clamp on insulin secretion and glucose disposal are displayed in Fig. 1. After acute elevation of the plasma glucose concentration to approximately 350 mg/dl, a brisk and sustained elevation of the plasma insulin concentration was observed in all rats, with a mean plasma concentration of 23.9 ng/ ml in the first 6 h of hyperglycemia. In a period between 6 and 10 h after the glucose infusion a significant decline in plasma insulin concentrations was observed, and a new plateau level was achieved at approximately 55% of the first plateau. This second steady state was then maintained throughout the study. The glucose disposal profile closely resembles the variation in the plasma insulin concentration. In fact, while approximately 52 mg/kg-min were necessary to maintain the plasma glucose concentration at 350 mg/dl in the first 6 h of hyperglycemia, thereafter the glucose infusion rate had to be reduced to 34 mg/kg-min, and it remained stable throughout the 44 h of the study. The in vivo downregulation of glucose-induced insulin secretion we observed in the present study seems to closely resemble the third phase of insulin secretion, described in vitro by Bolaffi et al. (23). Since hyperosmolarity per se may affect insulin secretion and PI hydrolysis (41) in isolated cell systems,
Endo • 1990 Vol 126 • No 1
control rats were infused for 44 h with mannitol in order to approximate the osmolar load administered during the hyperglycemic clamp study. Plasma insulin and glucose concentrations remained stable during the mannitol infusion. Islets isolated from mannitol-infused rats and immediately perifused responded with a brisk biphasic insulin secretory response during perifusion with 20 mM glucose. In contrast, islets isolated from glucose-infused rats displayed an altered secretory response. Peak first phase release was approximately 3-fold greater than that observed in the mannitol group, while second phase release was significantly reduced 60-70% compared to that in this group of control rats. Tentatively at least, the heightened first phase insulin secretory response might be ascribed to the induction, in vivo, of time-dependent potentiation or memory. The existence of this phenomenon is well established (22, 42) and is known to persist for at least 40 min after lowering the glucose level. Less clear is the mechanism responsible for the reduced second phase insulin secretory response from these freshly isolated perifused islets. We previously reported that exposure of the isolated islets in vitro to high (250 mg/dl) glucose or low (50 mg/dl) glucose plus cholecystokinin for 2 h resulted in an abnormal insulin secretory response to 20 mM glucose, an effect that appeared to result from impaired phosphoinositide hydrolysis (26). This reduced second phase insulin secretory response could be restored by inclusion of the diacylglycerol kinase inhibitor monooleoyglycerol (26) together with 20 mM glucose. In the present experiments, islets isolated from glucose-infused rats displayed reduced [3H]inositol efflux and inositol phosphate responses to 20 mM glucose. If labelling with [2-3H]inositol were carried out in 50 mg/ dl (2.75 mM) rather than 126 mg/dl (7 mM) glucose, an amplified [3H]inositol efflux response accompanied by a dramatic increase in inositol phosphate levels and insulin secretion were noted. In other words, these defective PI and insulin secretory responses reversed rapidly, a process regulated at least in part by the ambient glucose level. In fact, peak first phase output from islets perifused immediately after isolation was approximately 1100 pg/ islet-min, a value that fell to 600 (7 mM glucose) or 400 (2.75 mM) pg/islet-min depending on the glucose level used during the 2-h labelling procedure; in contrast, as peak first phase release rates fell, second phase release rates rose in a reciprocal fashion. This observation is consistent with recent reports by Weir and co-workers (24) and Grill and co-workers (43). At this time, one can only speculate on the nature of the biochemical lesion responsible for impaired PI hydrolysis and the ensuing altered insulin secretory response. Uptake of label into the islet does not seem to be a major problem. Rather, it would appear that factors
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE
normally responsible for the phospholipase-C-mediated cleavage of islet phosphoinositides in response to high glucose are desensitized in some manner. While the nature of the lesion remains obscure, the fact that it reverses so quickly in response to lowering the glucose level would appear to limit the number of possible candidates for further study. In summary, our findings demonstrate that chronic (44-h) hyperglycemia results in a reduction in 1) in vivo insulin secretion and glucose disposal, 2) second phase of glucose induced-insulin release, and 3) PI hydrolysis in perifused isolated islets. These impairments cannot be attributed to the effect of hyperosmolarity, since the infusion of an equiosmolar amount of mannitol cannot reproduce the phenomena. These data suggest that excessive in vivo stimulation of 0-cell PI hydrolysis desensitizes this biochemical pathway (and insulin release as well) to glucose stimulation. Although our finding does not prove a casual relationship between impaired PI hydrolysis and altered insulin secretion, the fact that the glucose level bathing the islet determines the longevity and reversibility of the impairment suggests to us that this lesion may account for the toxicity of high glucose on /3-cell insulin secretory response patterns. Additional studies focusing on the involvement of various Pi-derived second messenger molecules and the enzymes activated by them will be required to further characterize the defect in PI hydrolysis.
Acknowledgments The excellent technical assistance of Eveline Klein-Robbenhaar and the expert secretarial assistance of Ms. Rhonda Wolfe and Debbie Jung are gratefully acknowledged.
References 1. Robertson RP, Porte Jr D 1973 The Glucose receptor. A defective mechanism in diabetes mellitus distinct from beta adrenergic receptor. J Clin Invest 52:870 2. Cerasi E, Luft R, Efendic S 1972 Decreased sensitivity of the pancreatic beta cells to glucose in prediabetic and diabetic subjects. Diabetes 21:224 3. Brunzell JD, Robertson RP, Lerner RL, Hazzard WR, Ensinck JW, Bierman EL, Porte Jr D 1976 Relationship between fasting plasma glucose levels and insulin secretion during intravenous glucose tolerance tests. J Clin Endocrinol Metab 46:222 4. Dimitriadis G, Cryer P, Gerich J 1985 Prolonged hyperglycemia during infusion of glucose and somatostatin impairs pancreatic Aand B-cell responses to decrements in plasma glucose in normal man: evidence for induction of altered sensitivity to glucose. Diabetologia 28:63 5. Ward WK, Bolgiano DC, McKnight B, Halter J, Porte D 1984 Diminished beta cell secretory capacity in patients with noninsulin dependent diabetes mellitus. J Clin Invest 74:1318 6. Halter JB, Gray RJ, Porte D 1979 Potentiation of insulin secretion responses by plasma glucose levels in man: evidence that hyperglycemia in diabetes compensates for impaired glucose potentiation. J Clin Endocrinol Metab 48:948 7. DeFronzo RA, Ferrannini E, Koivisto V 1983 New concepts in the pathogenesis and treatment of non-insulin-dependent diabetes
259
mellitus. Am J Med 75:52 8. Bonner-Weir S, Trent DF, Weir GC 1983 Partial pancreatectomy in the rat and subsequent defect in glucose-induced insulin release. J Clin Invest 71:1544 9. Weir GC, Clore ET, Zmachinski CJ, Bonner-Weir S 1981 Islet secretion in a new experimental model for non-insulin dependent diabetes. Diabetes 30:590 10. Giroix MH, Portha B, Kergoat M, Bailbe D, Picon L 1983 Glucose insensitivity and amino-acid hypersensitivity of insulin release in rats with non-insulin dependent diabetes. Diabetes 32:445 11. Grill V, Rundfeldt M 1986 Abnormalities of insulin responses after ambient and previous exposure to glucose in streptozocin-diabetic and dexamethasone-treated rats. Diabetes 35:44 12. Leahy JL, Bonner-Weir S, Weir GC 1984 Abnormal glucose regulation of insulin secretion in models of reduced B-cell mass. Diabetes 33:667 13. Leahy JL, Cooper HE, Deal DA, Weir GC 1986 Chronic hyperglycemia is associated with impaired glucose influence on insulin secretion: a study in normal rats using chronic in vivo glucose infusion. J Clin Invest 77:908 14. Leahy JL, Weir GC 1988 Evolution of abnormal insulin secretory responses during 48 hr of in vivo hyperglycemia. Diabetes 37:217 15. Kosaka K, Kuzuya T, Akanuma Y, Hagura R 1980 Increase in insulin response after treatment of overt maturity onset diabetes mellitus is independent of the mode of treatment. Diabetologia 18:23 16. Stanik S, Marcus R 1980 Insulin secretion improves following dietary control of plasma glucose in severely hyperglycemic obese patients. Metabolism 29:346 17. Rossetti L, Shulman GI, Zawalich W, DeFronzo RA 1987 Effect of chronic hyperglycemia on in vivo insulin secretion in partially pancreatectomized rats. J Clin Invest 80:1037 18. Rossetti L, Smith D, Shulman GI, Papachristou D, DeFronzo RA 1987 Correction of hyperglycemia with phlorizin normalizes tissue sensitivity to insulin in diabetic rats. J Clin Invest 79:1510 19. Albano JDM, Ekins RR, Maritz G, Turner RC 1972 A sensitive, precise radioimmunoassay of serum insulin relying on charcoal separation of bound and free hormone moieties. Acta Endocrinol (Copenh) 70:487 20. Zawalich WS, Zawalich KC 1988 Phosphoinositide hydrolysis and insulin release from isolated perifused rat islets: studies with glucose. Diabetes 37:1294 21. Berridge MJ, Dawson RMC, Downes CP, Heslop JP, Irvine RF 1983 Changes in the level of inositol phosphates after agonistdependent hydrolysis of membrane phosphoinositides. Biochem J 212:473 22. Zawalich WS, Diaz VA, Zawalich KC 1988 Role of phosphoinositide metabolism in the induction of memory in isolated rat islets. Am J Physiol 254:E609 23. Bolaffi JL, Heldt A, Lewis LD, Grodsky GM 1986 The third phase of in vitro insulin secretion. Evidence for glucose insensitivity. Diabetes 35:370 24. Weir GC, Leahy JL, Wilson J, Matschinsky FM 1988 Defective insulin secretion induced by hyperglycemia is not due to altered glucose transport in the beta cell. Diabetes [Suppl 1] 37:110A 25. Kahn BB, Cushman SW, Shulman GI, DeFronzo RA, Rossetti L 1987 Reversal of insulin resistant glucose transport in adipose cells from diabetic rats by normalization of plasma glucose without insulin therapy. Clin Res 35:507A (Abstract) 26. Zawalich WS, Zawalich KC, Rasmussen H 1989 The conditions under which rat islets are labeled with (3H)-inositol alter the subsequent responses of these islets to a high glucose concentration. Biochem J 259:743 27. Best L, Malaisse WJ 1983 Phospholipids and islet function. Diabetologia 25:299 28. Zawalich WS 1988 Modulation of insulin secretion from B-cells by phosphoinositide-derived second messenger molecules. Diabetes 37:137 29. Axen KV, Schubart UK, Blake AD, Fleisher N 1983 Role of Ca2+ in secretagogue-stimulated breadkown of phosphatidylinositol in rat pancreatic islets. J Clin Invest 72:13 30. Clements RS, Rhoten WB 1976 Phosphoinositide metabolism and
260
31. 32. 33. 34. 35. 36. 37.
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE
insulin secretion from isolated rat pancreatic islets. J Clin Invest 57:684 Best L, Malaisse WJ 1983 Stimulation of phosphoinositide breakdown in rat pancreatic islets by glucose and carbanylcholine. Biochem Biophys Res Commun 116:9 Best L, Malaisse WJ 1984 Nutrient and hormone-neurotransmitter stimuli induce hydrolysis of polyphosphoinositides in rat pancreatic islets. Endocrinology 115:1814 Best L, Tomlinson S, Hawkins PT, Downes CP 1987 Production of inositol trisphosphates and inositol tetrakisphosphate in stimulated pancreatic islets. Biochim Biophys Acta 927:112 Blachier F, Segura MC, Malaisse WJ 1985 Unresponsiveness of phospholipase C to the regulatory proteins Ns and Ni in pancreatic islets. Res Commun Chem Pathol Pharmacol 55:335 Zawalich WS, Zawalich KC 1989 Interleukin 1 is a potent stimulator of islet insulin secretion and phosphoinositide hydrolysis. Am J Physiol 256:E19 Zawalich WS, Diaz VA, Zawalich KC 1989 Influence of cAMP and calcium on [3H]-inositol efflux, inositol phosphate accumulation, and insulin release from isolated rat islets. Diabetes 37:1478 Rana RS, Mertz RJ, Kowluru A, Dixon JF, Hokin LE, MacDonald MJ 1985 Evidence for glucose-responsive and -unresponsive pools
Endo • 1990 Vol 126* No 1
of phospholipid in pancreatic islets. J Biol Chem 260:7861 38. Zawalich WS, Takuwa N, Takuwa Y, Diaz VA, Rasmussen H 1987 Interactions of cholecystokinin and glucose in rat pancreatic islets. Diabetes 36:426 39. Garcia MC, Hermans MP, Henquin JC 1988 Glucose-, calciumand concentration-dependence of acetylcholine stimulation of insulin release and ionic fluxes in mouse islets. Biochem J 254:211 40. Zawalich WS, Zawalich KC, Rasmussen H 1989 Cholinergic stimulation primes the beta cell to glucose stimulation. Endocrinology 125:2400 41. Einspahr KJ, Peeler TC, Thompson Jr GA 1988 Rapid changes in polyphosphoinositide metabolism associated with the response of Dunaliella salina to hypoosmotic shock. J Biol Chem 263:5773 42. Grill V, Adamson V, Cerasi E 1988 Immediate and time-dependent effects of glucose on insulin release from rat pancreatic tissue. Evidence for different mechanism of action. J Clin Invest 61:1034 43. Grill V, Westberg M, Ostenson CG 1987 B Cell insensitivity in a rat model of non-insulin-dependent diabetes. Evidence for a rapidly reversible effect of previous hyperglycemia. J Clin Invest 80:664 44. Borle AB, Uchikawa T, Anderson JA 1982 Computer simulation and interpretation of 45Ca efflux profile patterns. J Membr Biol 68:36
Vol. 126, No. 1 Printed in U.S.A.
Chronic in Vivo Hyperglycemia Impairs Phosphoinositide Hydrolysis and Insulin Release in Isolated Perifused Rat Islets* WALTER S. ZAWALICH, KATHLEEN C. ZAWALICH, GERALD I. SHULMAN, LUCIANO ROSSETTIf Yale University School of Nursing (W.S.Z., K.C.Z.), New Haven, Connecticut 06536-0740; and Yale University School of Medicine (G.I.S., L.R.), New Haven, Connecticut 06510
ABSTRACT. We examined the effect of chronic hyperglycemia on phosphoinositide hydrolysis and insulin secretion in isolated perifused rat islets. Rats were infused for 44 h with 40% dextrose in order to raise and maintain the plasma glucose concentration at 350 mg/dl. Control animals were infused with equiosmolar amounts of mannitol. In vivo insulin secretion and rates of glucose disposal were monitored throughout the study. At the end of the infusion, islets were collagenase isolated, and phosphoinositide (PI) hydrolysis (assessed by measuring the increment in [3H]inositol efflux as well as labeled inositol phosphates) and insulin output in response to a 20-mM glucose challenge were quantitated. Plasma insulin concentration and in vivo glucose disposal rates decreased significantly, by 47% and 35% respectively, after 6-8 h of hyperglycemia. In islets perifused immediately after isolation, prior in vivo hyperglycemia markedly altered the pattern of insulin output in response
to 20-mM glucose challenge. Compared to mannitol infusion, 20 mM glucose stimulation resulted in an exaggerated first phase insulin secretory response (1121 ± 88 vs. 467 ± 75 pg/isletsmin) and a blunted second phase insulin secretory response (392 ± 90 vs. 1249 ± 205 pg/islet-min). In islets prelabeled with myo[2-3H]inositol for 2 h, PI hydrolysis, particularly [3H]inositol efflux in response to glucose stimulation was also reduced (0.28 ± 0.03%/min) compared to that in mannitol-infused animals (0.53 ± 0.08%/min). Two hours of preincubation in a low glucose medium (2.75 mM) were able to completely reverse the islet defect in both PI hydrolysis and insulin secretion. Our results demonstrate that chronic in vivo hyperglycemia impairs PI hydrolysis in perifused rat islets and suggest that this defect accounts in part for the abnormal pattern of glucoseinduced insulin secretion. (Endocrinology 126: 253-260, 1990)
D
insulin secretion induced by chronic glucose infusion. Islets were isolated from rats infused for 44 h with either high glucose or an osmotically equivalent mannitol load. The present results demonstrate that chronic in vivo hyperglycemia results in defective in vitro PI hydrolysis and insulin secretory responses to a glucose challenge in isolated perifused rat islets obtained from these animals.
EFECTS in glucose-induced insulin secretion are a common feature of noninsulin-dependent diabetes mellitus in man (1-7). Over the last decade, evidence has accumulated to indicate that chronic sustained hyperglycemia exerts a deleterious effect on /3-cell function (814). Furthermore, reversal of the hyperglycemia is accompanied by a correction of the insulin secretory defect (15-17). Recently, Leahy and co-workers (13, 14) have shown that 48 h of marked hyperglycemia severely impairs glucose-induced insulin secretion from the isolated perfused pancreas. However, little information exists regarding the nature of the biochemical lesion responsible for this aberrant insulin secretory response pattern. The present study was designed to examine the role of phosphoinositide (PI) hydrolysis in the alteration of
Materials and Methods Animal preparation Male Sprague-Dawley rats (Charles River Co., Wilmington, MA), weighing between 300-350 g, were used in all studies. Five to 7 days before the infusion studies rats were anesthetized with phenobarbital (50 mg/kg BW), and indwelling catheters were inserted into the internal jugular vein and carotid artery, so that the animals could be studied in the awake unstressed state (17, 18). The catheters were filled with heparin (40 U/ ml)-polyvinylpyrollidone (160 g/100 ml) solution, sealed, and tunneled sc around the side of the neck to the back of the head. The catheters were externalized through a skin incision in the back of the neck. Within 48 h all animals had recovered from surgery and were eating normally. Postoperatively, rats were
Received July 20,1989. Address requests for reprints to: Dr. Walter Zawalich, Yale University School of Medicine, 855 Howard Avenue, P.O. Box 9740, New Haven, Connecticut 06536-0740. * This work was supported by the American Heart Association and the Juvenile Diabetes Foundation. t Present address: Diabetes Section, Department of Medicine, University of Texas Health Science Center, 7703 Floyd Curl Drive, San Antonio, Texas 78284-7870. 253
254
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE
housed in an environmentally controlled room with a 12-h light, 12-h dark cycle and allowed free access to standard rat chow (Purina, Richmond, IN) and water. Experimental protocol Five to 7 days postsurgery, a solution containing 40% glucose (wt/vol) or equimolar amounts of mannitol was infused into the arterial catheter for a total of 44 h using a Harvard servocontrolled pump (Harvard Apparatus, Inc., South Natick, MA) and two minature infusion swivels (Spalding Medical Products, Arroyo Grande, CA). Throughout the study, the rats were allowed to move freely within the confines of a large cage and had free access to food and water. The venous catheter was used for blood withdrawal for determination of plasma glucose and insulin concentrations. The 40% glucose solution was administered to acutely raise the plasma glucose concentration by approximately 250 mg/dl above baseline. The plasma glucose concentration was subsequently held constant at this hyperglycemic plateau by the adjustment of a variable glucose infusion, as previously described (17,18). Isolated perifused islet studies At the end of the 44-h infusion rats were anesthetized with phenobarbital (50 mg/kg BW), and islets were isolated by collagenase digestion using Hanks' solution. After hand picking, the islets were placed in a medium identical to that used during the perifusion, except for the presence of 126 mg/dl (7 mM) glucose. Some islets were directly perifused after isolation. In other experiments the islets were incubated for 2 h in a solution supplemented with myo-[2-3H]inositol. In these studies, 10 juCi myo-[2-3H]inositol were added to 250 n\ perifusion medium supplemented with either 2.75 or 7 mM glucose. Two hundred microliters of this solution were then added to the islets. After labeling, the islets were washed with 5 ml unlabeled medium and then perifused. It should be noted that the time between the addition of collagenase to begin the pancreatic digestion and the start of the perifusion or labeling with [3H] inositol averaged less than 60 min. Islets were perifused at a flow rate of 1 ml/min for 30 min to establish stable basal secretory rates. After this stabilization period, the islets were subjected to various protocols, indicated in the figure legends and in Table 2. Effluent samples were collected and analyzed for insulin content (19), using rat insulin as standard (lot 615-D63-12-3, Eli Lilly Co., Indianapolis, IN) and, when appropriate, [3H] inositol content. The sensitivity of the RIA for insulin determination was 0.195 ng/ml; the intraassay coefficient of variance was less than 9%, and the interassay coefficient of variation was less than 18%. Previous results from this laboratory have demonstrated that approximately 90% of the effluent radioactivity from glucose-stimulated islets is from [3H]inositol (20). Inositol phosphates were extracted after the perifusion with 10% perchloric acid and separated on anion exchange columns (AG1-X8, 200-400 mesh formate form, Bio-Rad Laboratories, Richmond, CA) using methods previously described (20-22). Briefly, the islets still attached to the nylon filters were retrieved and placed in small glass vials. Perchloric acid (0.4 ml; 20%) and 0.4 ml perfusion medium
Endo • 1990 Vol 126« No 1
were immediately added. After 30 min on ice the extract was neutralized with 0.28 ml 6 N KOH and then added to the columns. These columns were prepared by adding (to achieve a length of 3 cm) anion exchange resin to Pasteur pipettes. Further additions to the column included 10 ml water and 5 ml 5 mM borax-60 mM sodium formate. Elution of the inositol phosphates was accomplished by the sequential addition of 10 ml 0.1 M formic acid-0.2 M ammonium formate (inositol 1phosphate), 0.1 M formic acid-0.4 M ammonium formate (inositol 1,4-bisphosphate), and 0.1 M formic acid-1 M ammonium formate (inositol 1,4,5-trisphosphate plus inositol 1,3,4-trisphosphate). Aliquots (0.4 ml) of the eluate were then analyzed for radioactive contents. The filters, still containing the islets, were also counted. Total [3H]inositol incorporation was calculated as the amount of effluent radioactivity collected from 2860 min of the perfusion, total label in [3H]inositol phosphates, and label present in the islets after extraction of the inositol phosphates with perchloric acid. Reagents Hanks' solution (supplemented with 7 mM glucose) was used for the islet isolation. The perifusion medium consisted of 115 mM NaCl, 5 mM KC1, 2.2 mM CaCl2, 24 mM NaHCO3, and 0.17 g/dl BSA. Other compounds were added as indicated, and the solution was gassed with a mixture of 95% O2-5% CO2. The [125I] insulin used for the insulin assay was purchased from New England Nuclear (Boston, MA), and the myo-[2-3H]inositol was from Amersham (Arlington Heights, IL). BSA (RIA grade), glucose, mannitol, formic acid, as well as the salts used to make the Hanks' solution and perifusion medium were purchased from Sigma Chemical Co. (St. Louis, MO). Ammonium formate was purchased from Fisher Scientific (Fairlawn, NJ). Statistics Statistical significance was determined using Student's t test for unpaired data and, where appropriate, analysis of variance. P < 0.05 was taken as significant. Values presented are the mean ± SE of the specified number of observations.
Results Body weights, plasma glucose, and insulin concentrations (Table 1 and Fig. 1) Body weights did not change significantly during the 2 days of the infusion in either the glucose- or mannitolinfused rats. In the mannitol-infused animals, there was no significant increment above basal in plasma glucose (138 ± 3 mg/dl) and plasma insulin concentrations (2.9 ± 0.4 ng/ml) during the infusion period. In the glucoseinfused animals, the plasma glucose concentration was 346 ± 7 mg/dl and remained constant throughout the study (Fig. 1, top panel). The mean plasma insulin concentration was 23.9 ±2.1 ng/ml in the first 6 h (0-6 h) and dropped to 12.7 ± 0.8 ng/ml in the last 34 h of the study (10-44 h; Fig. 1, top panel). The glucose infusion rate necessary to maintain the plasma glucose concen-
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE
255
TABLE 1. Body weight, plasma insulin (I), plasma glucose (G), and mannitol or glucose infusion rates (IR) during the in vivo study BW (g)
I (ng/ml)
G (mg/dl)
IR (mg/kg-min)
0 44
322 ± 19 326 ± 21
2.9 ± 0.2 3.0 ± 0.3
138 ± 4 149 ± 3
38.0
0 0-6 10-44
328 ± 21
3.0 ± 0.2 23.9 ± 2.1 12.7 ± 0.8
139 ± 3 344 ± 10 348 ± 13
51.6 ± 1.0 35.4 ± 1.5
Time (h)
Infusion Mannitol
Glucose
334 ± 18
PLASMA GLUCOSE AND INSULIN CONCENTRATION 400
1200
r
1000
islet
"e
800
ASE
2
600
GLUCOSE INFUSION RATES MEAN GLUCOSE INFUSION 0-6hr = 51.6 + 1.1 10-44hr = 33.53 i i . 4
co 400
200 36
40
44
TIME (hours) FlG. 1. Effect of chronic glucose infusion on plasma glucose and plasma insulin concentrations (top) and glucose infusion rates {bottom). Rats were infused for 44 h with a variable 40% glucose solution to raise and maintain the plasma glucose concentration at approximately 350 mg/dl. The plasma glucose (milligrams per dl) and insulin (nanograms per ml) concentration (top panel) and the glucose infusion rate (milligrams per kg/min; bottom panel) are shown. The continuous line in the bottom panel represents the mean, and the interrupted line represents the SE.
tration constant at the desired hyperglycemic plateau was 51.6 ± 1.0 mg/kg-min in the first 6 h (0-6 h) and dropped to 33.4 ± 1.5 mg/kg-min during the last 34 h (10-44 h) of the study (Fig. 1, bottom panel), displaying a behavior similar to that observed for the plasma insulin concentrations. Perifused islet studies (Figs. 2-4 and Table 2) Islets isolated from mannitol-infused rats displayed a brisk biphasic insulin secretory response when directly perifused immediately after isolation (Fig. 2). Peak first phase release averaged 467 ± 75 pg/islet-min, while
° '30 40 50 60 70 DURATION OF PERIFUSION (min) FIG. 2. Chronic hyperglycemia alters /3-cell insulin secretory response to glucose. Groups of 12-15 islets (n = 6 for each condition) were isolated from either glucose-infused or mannitol-infused rats and perifused immediately. For 30 min they were perifused with 2.75 mM glucose (G275) to establish basal stable insulin secretory rates. They were then stimulated for 40 min with 20 mM glucose. The mean ± selected SEs are given. This and subsequent figures have been corrected for the dead space in the perifusion system (~2.5 ml or 2.5 min with a flow rate of 1 ml/min).
release measured 35-40 min after the onset of stimulation averaged 1249 ± 205 pg/islet-min. In contrast to this response, the insulin secretory response of islets isolated from glucose-infused animals and immediately perifused displayed an abnormal profile. Peak first phase release was dramatically amplified, reaching an average value of 1121 ± 88 pg/islet-min. Second phase release was attenuated, reaching an average value of only 392 ± 90 pg/islet-min after 35-40 min of stimulation with 20 mM glucose (Fig. 2).
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE
256 0.8
1000 r '2.75
620 rf°h HYPERGLYCEMIC L (REVERSAL)
'2.75
/HYPERGLYCEMIC D (REVERSAL)
0.6
Endo • 1990 Vol 126 • No 1
r
'E 800 MANNITOL
< 600 0.4
•
•MANNITOL
HYPERGLYCEMIC
CO LLJ
_l UJ
0.2
HYPERGLYCEMIC
cc
30 40 50 60 DURATION OF PERIFUSION ( m i n ) FIG. 3. Chronic hyperglycemia results in altered, but reversible, patterns of [3H]inositol efflux from prelabeled islets. Groups of islets were isolated from glucose-infused or mannitol-infused rats. For 2 h they were incubated in a medium supplemented with myo-[2-3H]inositol to label their PI. The glucose level used during this 2-h period was either 7 mM (hyperglycemic or mannitol) or 2.75 mM (hyperglycemic reversal). All islets were then washed, then perifused for 30 min with 2.75 mM glucose (G2.75) and for an additional 30 min with 20 mM glucose. At least four experiments were performed under each condition. Fractional efflux rates of [3H]inositol, given as percent per min, were calculated according to the method of Borle et al. (44).
The responses depicted in Figs. 3 and 4 were obtained after a 2-h incubation period to label islet PI. In response to glucose stimulation, islets (labeled in the presence of 7 mM glucose) from mannitol-infused rats demonstrated a prompt and significant increase in the fractional efflux rate of [3H]inositol (Fig. 3). After 30 min of stimulation, this value reached 0.53 ± 0.08%/min. Efflux rates from islets maintained for 60 min with 2.75 mM glucose alone averaged 0.09 ± 0.03%/min (results not shown). In contrast to this response, islets isolated from glucose-infused animals and labeled in the presence of 7 mM glucose, demonstrated a small and attenuated increase in [3H]inositol efflux rates. After 30 min of stimulation with 20 mM glucose, the fractional efflux rate of [3H]inositol averaged only 0.28 ± 0.035%/min, a value significantly (P < 0.05) different from that in islets isolated from the mannitol-infused rats. Most interestingly, labeling of islets isolated from glucose-infused animals in the presence of 2.75 mM glucose was accompanied by an amplified [3H]inositol efflux response (Fig. 3). After 30 min with 20 mM glucose, this efflux value averaged 0.79 ± 0.11%/min. In islets isolated from mannitol-infused rats, labeling with [2-3H]inositol resulted in the incorporation of 27,175 ± 1,392 cpm/40 islets. Islets isolated from glucose-infused animals incorporated
400
200
0 L,
_ 30
40
50
60
DURATION OF PERIFUSION (min) FlG. 4. Chronic hyperglycemia results in altered, but reversible, patterns of insulin output from prelabeled islets. Groups of islets were isolated from mannitol- or glucose-infused rats, then labeled for 2 h as indicated in Fig. 3. They were perifused for 30 min to establish stable basal insulin release rates and then stimulated for 30 min with 20 mM glucose. Note the difference in insulin response characteristics in islets from hyperglycemic animals if they are perifused immediately after isolation (Fig. 2) or subjected to a 2-h labeling period.
30,215 ± 1,885 cpm/40 islets. In response to 20-mM glucose stimulation, islets isolated from mannitol-infused rats released 2,144 ± 415 cpm/40 islets over the final 20 min of stimulation. Islets isolated from glucoseinfused animals and labeled in the presence of 7 mM glucose released 1,132 ± 322 cpm/40 islets over this time. In contrast, islets isolated from glucose-infused rats but labeled in the presence of 2.75 mM glucose released 3,321 ± 469 cpm/40 islets over this time. Results similar to the [3H]inositol efflux findings were obtained when the insulin release profiles were monitored from these same islets (Fig. 4). A biphasic insulin secretory response was noted from the mannitol-infused group. An abnormal insulin secretory profile, again accompanied by a dramatic first phase secretion and a reduced second phase response, was observed in glucoseinfused islets labeled in the presence of 7 mM glucose. Islets isolated from glucose-infused animals and labeled in the presence of 2.75 mM glucose displayed normal, albeit slightly amplified, temporal patterns of insulin output. At least when glucose (2.75 or 20 mM) is used as the stimulant, average fractional rates of [3H]inositol efflux measured during the last 10 min of the perifusion correlated significantly with total insulin release during this time (r = 0.925; P < 0.02). At the termination of the perifusion, labeled inositol phosphate levels were measured (Table 2). Compared to
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE TABLE 2. Influence of chronic mannitol or glucose infusions on inositol phosphate levels in isolated rat islets Labeling protocol
Perifusion protocol (min)
IPi
IP2
IP,
Mannitol infusion 1. G 7 2. G7 Glucose infusion 3. G7 4. G7 5. G275 6. G275
G2.75 481 ± 26° 121 ± 2 3 73 ± 19 (60) G2.76 -> G20 1066 ± 124 343 ± 50 184 ± 21 (30) (30) G 275 (60) 625 ± 69 G2.75 -+ G20 802 ± 53 (30) (30) G275 564 ± 51 (60) G2.75 -> G20 1646 ± 214 (30) (30)
179 ± 1 8 99 ± 8 309 ± 25 183 ± 29 198 ± 23 135 ± 7 936 ± 91 487 ± 51
Groups of 30-50 islets were isolated from mannitol- or glucoseinfused rats. They were incubated for 2 h in a medium supplemented with [3H]inositol to label their PI. Also present was 7 mM (protocols 1-4) or 2.75 mM glucose (G; protocols 5 and 6). After washing to remove unincorporated label, the islets were perifused for 30 mn with 2.75 mM glucose to establish basal and stable insulin secretory and [3H]inositol efflux rates. Some groups (protocols 1, 3, and 5) were maintained for an additional 30 min with 2.75 mM glucose, while the others (protocols 2, 4, and 6) were stimulated for 30 min with 20 mM glucose. After this, inositol phosphates were extracted and measured as detailed in Materials and Methods. The duration of each perifusion is indicated in parentheses below each protocol. Mean values ± SEs are given for at least 4 experiments. IPi, Inositol 1-phosphate; IP2, inositol 1,4-bisphosphate; IP3, inositol trisphosphate. Statistical analysis: protocol 1 vs. 2, P < 0.05 for all inositol phosphates; protocol 3 vs. 4, P < 0.05 for all inositol phosphates; protocol 6 vs. 5, P < 0.05 for all inositol phosphates; protocol 6 vs. 2, P < 0.05 for all inositol phosphates; protocol 6 vs. 4, P < 0.05 for all inositol phosphates. 0 Counts per min/40 islets (mean ± SE).
the levels of inositol phosphates measured in the presence of 2.75 mM glucose, islets isolated from mannitolinfused rats displayed dramatic and significant increases in the levels of all inositol phosphates measured in response to 20 mM glucose (Table 2, line 2). In contrast, islets isolated from glucose-infused rats labeled in the presence of 7 mM glucose displayed attenuated responses to 20 mM glucose .(Table 2, line 4). Most dramatic, however, was the response observed in islets isolated from glucose-infused rats but labeled in the presence of 2.75 mM glucose. Similar to the [3H]inositol efflux and insulin release results, a dramatic increase in the levels of all labeled inositol phosphates measured occurred in response to 20-mM glucose stimulation (Table 2, line 6).
Discussion A reduced or abolished insulin secretory response to glucose characterizes the /3-cell of type II diabetic subjects (1-7). Since the response to other secretagogues, such as arginine, is preserved, it has been postulated that
257
chronic exposure of the /3-cell to an elevated ambient glucose concentration causes specific desensitization to the glucose stimulus (8-14). Studies in animals support this notion, since insensitivity to an acute glucose stimulus has been demonstrated to occur in various different animal models of diabetes in vivo (17) and in vitro (814). This defect in insulin secretion can be completely reversed by correcting the chronic hyperglycemia (1417). Recently, Bolaffi and co-workers (23) have shown that prolonged in vitro exposure of different /3-cell preparations to high glucose results in a reduced pattern of the glucose-induced insulin output. Interestingly, the phenomenon seems to be reproduced by our chronic glucose infusion in vivo. After 6-10 h of sustained hyperglycemia a brisk decline of the insulin output was observed, despite constant hyperglycemic stimulation. However the biochemical basis for this desensitization process has not been defined. A recent report by Weir et al. (24) indicates that, unlike adipose cells (25), the defect caused by hyperglycemia does not appear to reside in altered transport of the hexose into the cell. Our recent in vitro studies with islets incubated in high (250 mg/dl) glucose or in low (50 mg/dl) glucose plus the gut hormone cholecystokinin have indicated that the defective insulin secretory response noted after a 2-h exposure to these compounds is paralleled by impaired PI hydrolysis (26). Because of the strategic location of these phospholipids and because of the plethora of second messenger molecules generated by their hydrolysis (27, 28), we suggested that a defect in the hydrolysis of these phospholipids might contribute to the altered insulin secretory response to a subsequent 20-mM glucose challenge. In the present report we investigated whether abnormal PI responses play any role in the aberrant patterns of insulin secretion noted after chronic in vivo exposure to high glucose, a model previously studied extensively by Leahy and coworkers (13, 14). The results suggest that this manipulation impairs, in a parallel fashion, glucose-stimulated PI hydrolysis and insulin secretion. Before discussing our findings in detail, a few comments concerning the relationship between PI hydrolysis and insulin release in islets are appropriate. Previous studies by Axen et al. (29), Clements and Rhoten (30), Best et al. (31-33), Blachier et al. (34), and Zawalich et al. (22,35,36), have demonstrated that islets [3H]inositol labeled in low (2.7-5.6 mM) glucose are responsive, in terms of increases in PI hydrolysis, to subsequent glucose stimulation. These findings stand in contrast to the studies of Rana et al. (37), who found it necessary to [3H] inositol label islets in high (16.7 mM) glucose in order to demonstrate subsequent glucose-induced increases in PI hydrolysis. In a most recent in vitro study (26), we have been unable to confirm these results. In fact, labeling in high glucose is actually detrimental to the islet, in terms
258
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE
of both PI hydrolysis and insulin release (26). The in vivo findings reported here further support the concept that hyperglycemia adversely influences both of these processes. Finally, although our results suggest that a positive correlation exists between increases in PI hydrolysis (when [3H]inositol efflux is used as the parameter of PI activation) and insulin release, this appears to be the case only when glucose is used as the stimulant. For example, if PI hydrolysis is stimulated by the enteric hormone cholecystokinin (38) or by cholinergic agonists (34, 39, 40), significant changes in [3H]inositol efflux and [3H]inositol phosphate accumulation are observed whether the glucose level is low (2.75 mM) or high (7.0 mM). However, insulin release significantly increases only at the higher glucose level. Clearly, it is possible with these other agonists to dissociate increases in PI hydrolysis from increases in insulin release. Unfortunately, the nature of the signal supplied by glucose which permits the coupling of PI hydrolysis to release under these conditions is not known, although it does appear to be metabolic in nature. The data presented in Table 1 indicate that neither the mannitol-infused nor the glucose-infused rats displayed increases in body weight despite free access to food and water during the 44 h of the study. Consequently, the impact of nutritional deprivation as a contributory factor in the observed defects has to be considered. However, a similar variation in the nutritional state seems to occur in hyperglycemic and euglycemic rats. The effects of the 44-h hyperglycemic clamp on insulin secretion and glucose disposal are displayed in Fig. 1. After acute elevation of the plasma glucose concentration to approximately 350 mg/dl, a brisk and sustained elevation of the plasma insulin concentration was observed in all rats, with a mean plasma concentration of 23.9 ng/ ml in the first 6 h of hyperglycemia. In a period between 6 and 10 h after the glucose infusion a significant decline in plasma insulin concentrations was observed, and a new plateau level was achieved at approximately 55% of the first plateau. This second steady state was then maintained throughout the study. The glucose disposal profile closely resembles the variation in the plasma insulin concentration. In fact, while approximately 52 mg/kg-min were necessary to maintain the plasma glucose concentration at 350 mg/dl in the first 6 h of hyperglycemia, thereafter the glucose infusion rate had to be reduced to 34 mg/kg-min, and it remained stable throughout the 44 h of the study. The in vivo downregulation of glucose-induced insulin secretion we observed in the present study seems to closely resemble the third phase of insulin secretion, described in vitro by Bolaffi et al. (23). Since hyperosmolarity per se may affect insulin secretion and PI hydrolysis (41) in isolated cell systems,
Endo • 1990 Vol 126 • No 1
control rats were infused for 44 h with mannitol in order to approximate the osmolar load administered during the hyperglycemic clamp study. Plasma insulin and glucose concentrations remained stable during the mannitol infusion. Islets isolated from mannitol-infused rats and immediately perifused responded with a brisk biphasic insulin secretory response during perifusion with 20 mM glucose. In contrast, islets isolated from glucose-infused rats displayed an altered secretory response. Peak first phase release was approximately 3-fold greater than that observed in the mannitol group, while second phase release was significantly reduced 60-70% compared to that in this group of control rats. Tentatively at least, the heightened first phase insulin secretory response might be ascribed to the induction, in vivo, of time-dependent potentiation or memory. The existence of this phenomenon is well established (22, 42) and is known to persist for at least 40 min after lowering the glucose level. Less clear is the mechanism responsible for the reduced second phase insulin secretory response from these freshly isolated perifused islets. We previously reported that exposure of the isolated islets in vitro to high (250 mg/dl) glucose or low (50 mg/dl) glucose plus cholecystokinin for 2 h resulted in an abnormal insulin secretory response to 20 mM glucose, an effect that appeared to result from impaired phosphoinositide hydrolysis (26). This reduced second phase insulin secretory response could be restored by inclusion of the diacylglycerol kinase inhibitor monooleoyglycerol (26) together with 20 mM glucose. In the present experiments, islets isolated from glucose-infused rats displayed reduced [3H]inositol efflux and inositol phosphate responses to 20 mM glucose. If labelling with [2-3H]inositol were carried out in 50 mg/ dl (2.75 mM) rather than 126 mg/dl (7 mM) glucose, an amplified [3H]inositol efflux response accompanied by a dramatic increase in inositol phosphate levels and insulin secretion were noted. In other words, these defective PI and insulin secretory responses reversed rapidly, a process regulated at least in part by the ambient glucose level. In fact, peak first phase output from islets perifused immediately after isolation was approximately 1100 pg/ islet-min, a value that fell to 600 (7 mM glucose) or 400 (2.75 mM) pg/islet-min depending on the glucose level used during the 2-h labelling procedure; in contrast, as peak first phase release rates fell, second phase release rates rose in a reciprocal fashion. This observation is consistent with recent reports by Weir and co-workers (24) and Grill and co-workers (43). At this time, one can only speculate on the nature of the biochemical lesion responsible for impaired PI hydrolysis and the ensuing altered insulin secretory response. Uptake of label into the islet does not seem to be a major problem. Rather, it would appear that factors
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE
normally responsible for the phospholipase-C-mediated cleavage of islet phosphoinositides in response to high glucose are desensitized in some manner. While the nature of the lesion remains obscure, the fact that it reverses so quickly in response to lowering the glucose level would appear to limit the number of possible candidates for further study. In summary, our findings demonstrate that chronic (44-h) hyperglycemia results in a reduction in 1) in vivo insulin secretion and glucose disposal, 2) second phase of glucose induced-insulin release, and 3) PI hydrolysis in perifused isolated islets. These impairments cannot be attributed to the effect of hyperosmolarity, since the infusion of an equiosmolar amount of mannitol cannot reproduce the phenomena. These data suggest that excessive in vivo stimulation of 0-cell PI hydrolysis desensitizes this biochemical pathway (and insulin release as well) to glucose stimulation. Although our finding does not prove a casual relationship between impaired PI hydrolysis and altered insulin secretion, the fact that the glucose level bathing the islet determines the longevity and reversibility of the impairment suggests to us that this lesion may account for the toxicity of high glucose on /3-cell insulin secretory response patterns. Additional studies focusing on the involvement of various Pi-derived second messenger molecules and the enzymes activated by them will be required to further characterize the defect in PI hydrolysis.
Acknowledgments The excellent technical assistance of Eveline Klein-Robbenhaar and the expert secretarial assistance of Ms. Rhonda Wolfe and Debbie Jung are gratefully acknowledged.
References 1. Robertson RP, Porte Jr D 1973 The Glucose receptor. A defective mechanism in diabetes mellitus distinct from beta adrenergic receptor. J Clin Invest 52:870 2. Cerasi E, Luft R, Efendic S 1972 Decreased sensitivity of the pancreatic beta cells to glucose in prediabetic and diabetic subjects. Diabetes 21:224 3. Brunzell JD, Robertson RP, Lerner RL, Hazzard WR, Ensinck JW, Bierman EL, Porte Jr D 1976 Relationship between fasting plasma glucose levels and insulin secretion during intravenous glucose tolerance tests. J Clin Endocrinol Metab 46:222 4. Dimitriadis G, Cryer P, Gerich J 1985 Prolonged hyperglycemia during infusion of glucose and somatostatin impairs pancreatic Aand B-cell responses to decrements in plasma glucose in normal man: evidence for induction of altered sensitivity to glucose. Diabetologia 28:63 5. Ward WK, Bolgiano DC, McKnight B, Halter J, Porte D 1984 Diminished beta cell secretory capacity in patients with noninsulin dependent diabetes mellitus. J Clin Invest 74:1318 6. Halter JB, Gray RJ, Porte D 1979 Potentiation of insulin secretion responses by plasma glucose levels in man: evidence that hyperglycemia in diabetes compensates for impaired glucose potentiation. J Clin Endocrinol Metab 48:948 7. DeFronzo RA, Ferrannini E, Koivisto V 1983 New concepts in the pathogenesis and treatment of non-insulin-dependent diabetes
259
mellitus. Am J Med 75:52 8. Bonner-Weir S, Trent DF, Weir GC 1983 Partial pancreatectomy in the rat and subsequent defect in glucose-induced insulin release. J Clin Invest 71:1544 9. Weir GC, Clore ET, Zmachinski CJ, Bonner-Weir S 1981 Islet secretion in a new experimental model for non-insulin dependent diabetes. Diabetes 30:590 10. Giroix MH, Portha B, Kergoat M, Bailbe D, Picon L 1983 Glucose insensitivity and amino-acid hypersensitivity of insulin release in rats with non-insulin dependent diabetes. Diabetes 32:445 11. Grill V, Rundfeldt M 1986 Abnormalities of insulin responses after ambient and previous exposure to glucose in streptozocin-diabetic and dexamethasone-treated rats. Diabetes 35:44 12. Leahy JL, Bonner-Weir S, Weir GC 1984 Abnormal glucose regulation of insulin secretion in models of reduced B-cell mass. Diabetes 33:667 13. Leahy JL, Cooper HE, Deal DA, Weir GC 1986 Chronic hyperglycemia is associated with impaired glucose influence on insulin secretion: a study in normal rats using chronic in vivo glucose infusion. J Clin Invest 77:908 14. Leahy JL, Weir GC 1988 Evolution of abnormal insulin secretory responses during 48 hr of in vivo hyperglycemia. Diabetes 37:217 15. Kosaka K, Kuzuya T, Akanuma Y, Hagura R 1980 Increase in insulin response after treatment of overt maturity onset diabetes mellitus is independent of the mode of treatment. Diabetologia 18:23 16. Stanik S, Marcus R 1980 Insulin secretion improves following dietary control of plasma glucose in severely hyperglycemic obese patients. Metabolism 29:346 17. Rossetti L, Shulman GI, Zawalich W, DeFronzo RA 1987 Effect of chronic hyperglycemia on in vivo insulin secretion in partially pancreatectomized rats. J Clin Invest 80:1037 18. Rossetti L, Smith D, Shulman GI, Papachristou D, DeFronzo RA 1987 Correction of hyperglycemia with phlorizin normalizes tissue sensitivity to insulin in diabetic rats. J Clin Invest 79:1510 19. Albano JDM, Ekins RR, Maritz G, Turner RC 1972 A sensitive, precise radioimmunoassay of serum insulin relying on charcoal separation of bound and free hormone moieties. Acta Endocrinol (Copenh) 70:487 20. Zawalich WS, Zawalich KC 1988 Phosphoinositide hydrolysis and insulin release from isolated perifused rat islets: studies with glucose. Diabetes 37:1294 21. Berridge MJ, Dawson RMC, Downes CP, Heslop JP, Irvine RF 1983 Changes in the level of inositol phosphates after agonistdependent hydrolysis of membrane phosphoinositides. Biochem J 212:473 22. Zawalich WS, Diaz VA, Zawalich KC 1988 Role of phosphoinositide metabolism in the induction of memory in isolated rat islets. Am J Physiol 254:E609 23. Bolaffi JL, Heldt A, Lewis LD, Grodsky GM 1986 The third phase of in vitro insulin secretion. Evidence for glucose insensitivity. Diabetes 35:370 24. Weir GC, Leahy JL, Wilson J, Matschinsky FM 1988 Defective insulin secretion induced by hyperglycemia is not due to altered glucose transport in the beta cell. Diabetes [Suppl 1] 37:110A 25. Kahn BB, Cushman SW, Shulman GI, DeFronzo RA, Rossetti L 1987 Reversal of insulin resistant glucose transport in adipose cells from diabetic rats by normalization of plasma glucose without insulin therapy. Clin Res 35:507A (Abstract) 26. Zawalich WS, Zawalich KC, Rasmussen H 1989 The conditions under which rat islets are labeled with (3H)-inositol alter the subsequent responses of these islets to a high glucose concentration. Biochem J 259:743 27. Best L, Malaisse WJ 1983 Phospholipids and islet function. Diabetologia 25:299 28. Zawalich WS 1988 Modulation of insulin secretion from B-cells by phosphoinositide-derived second messenger molecules. Diabetes 37:137 29. Axen KV, Schubart UK, Blake AD, Fleisher N 1983 Role of Ca2+ in secretagogue-stimulated breadkown of phosphatidylinositol in rat pancreatic islets. J Clin Invest 72:13 30. Clements RS, Rhoten WB 1976 Phosphoinositide metabolism and
260
31. 32. 33. 34. 35. 36. 37.
HYPERGLYCEMIA, PI HYDROLYSIS, AND INSULIN RELEASE
insulin secretion from isolated rat pancreatic islets. J Clin Invest 57:684 Best L, Malaisse WJ 1983 Stimulation of phosphoinositide breakdown in rat pancreatic islets by glucose and carbanylcholine. Biochem Biophys Res Commun 116:9 Best L, Malaisse WJ 1984 Nutrient and hormone-neurotransmitter stimuli induce hydrolysis of polyphosphoinositides in rat pancreatic islets. Endocrinology 115:1814 Best L, Tomlinson S, Hawkins PT, Downes CP 1987 Production of inositol trisphosphates and inositol tetrakisphosphate in stimulated pancreatic islets. Biochim Biophys Acta 927:112 Blachier F, Segura MC, Malaisse WJ 1985 Unresponsiveness of phospholipase C to the regulatory proteins Ns and Ni in pancreatic islets. Res Commun Chem Pathol Pharmacol 55:335 Zawalich WS, Zawalich KC 1989 Interleukin 1 is a potent stimulator of islet insulin secretion and phosphoinositide hydrolysis. Am J Physiol 256:E19 Zawalich WS, Diaz VA, Zawalich KC 1989 Influence of cAMP and calcium on [3H]-inositol efflux, inositol phosphate accumulation, and insulin release from isolated rat islets. Diabetes 37:1478 Rana RS, Mertz RJ, Kowluru A, Dixon JF, Hokin LE, MacDonald MJ 1985 Evidence for glucose-responsive and -unresponsive pools
Endo • 1990 Vol 126* No 1
of phospholipid in pancreatic islets. J Biol Chem 260:7861 38. Zawalich WS, Takuwa N, Takuwa Y, Diaz VA, Rasmussen H 1987 Interactions of cholecystokinin and glucose in rat pancreatic islets. Diabetes 36:426 39. Garcia MC, Hermans MP, Henquin JC 1988 Glucose-, calciumand concentration-dependence of acetylcholine stimulation of insulin release and ionic fluxes in mouse islets. Biochem J 254:211 40. Zawalich WS, Zawalich KC, Rasmussen H 1989 Cholinergic stimulation primes the beta cell to glucose stimulation. Endocrinology 125:2400 41. Einspahr KJ, Peeler TC, Thompson Jr GA 1988 Rapid changes in polyphosphoinositide metabolism associated with the response of Dunaliella salina to hypoosmotic shock. J Biol Chem 263:5773 42. Grill V, Adamson V, Cerasi E 1988 Immediate and time-dependent effects of glucose on insulin release from rat pancreatic tissue. Evidence for different mechanism of action. J Clin Invest 61:1034 43. Grill V, Westberg M, Ostenson CG 1987 B Cell insensitivity in a rat model of non-insulin-dependent diabetes. Evidence for a rapidly reversible effect of previous hyperglycemia. J Clin Invest 80:664 44. Borle AB, Uchikawa T, Anderson JA 1982 Computer simulation and interpretation of 45Ca efflux profile patterns. J Membr Biol 68:36
