Biochimica et Biophysica Acta, 677 (1981) 263-268 263 Elsevier/North-Holland Biomedical Press

BBA 29757

COMPARATIVE STUDY ON THE EFFECTS OF A HYPOGLYCEMIC 2-SUBSTITUTED-2-1MIDAZOLINE DERIVATIVE (DG-5128) AND TOLBUTAMIDE ON INSULIN SECRETION FROM AND INSULIN SYNTHESIS IN THE ISOLATED RAT PANCREATIC ISLETS

KIN-YA KAMEDA, ISAO KOYAMA and YASUSHI ABIKO *

Laboratory of Biochemistry, Research Institute, Daiichi Saiyaku Co., Ltd., 16-13, Kitakasai-l, Edogawa-ku, Tokyo 132 (Japan)

(Received April 6th, 1981)

Key words: Hypoglycemia; Tolbutamide; Insulin secretion; Insulin synthesis; Imidazoline derivative; (Rat pancreatic islets)

212-(45-Dihydro-lH-imidazol-2-yl)-l-phenylethyl]pyridine dihydrochloride sesquihydrate (DG-5128) was f o u n d

to stimulate the glucose-primed insulin secretion from the isolated rat pancreatic islets throughout the incubation period, unlike tolbutamide which stimulated it only in the initial phase of incubation. The effect of DG-5128 was more pronounced at a higher glucose concentration (5 mg/ml). In the islet perifusion study, DG-5128 was also found to stimulate the glucose-induced insulin secretion in both the first and the second phases of the reaction, in contrast to tolbutamide which stimulated only the first phase of insulin secretion from the perifused islets. DG- 5128 gave no significant effect on the glucose-stimulated increase in incorporation of [3H]leucine into the pro- insulin and insulin fractions, while tolbutamide significantly inhibited the incorporation especially at a low glucose concentration (1 mg/ml). These and the previous fmdings indicate that DG-5128 is a new class of hypo- glycemic agent with a unique mode of action different from the known hypoglycemics ever reported.

Introduction

2 [2-(4 ~5-Dihydro-1 H-imidazol-2-yl)-I -phenyl-eth- yl]pyridine dihydrochloride sesquihydrate (DG-5128) is a new, orally effective hypoglycemic agent struc- turally unrelated to the known hypoglycemics [1,2]. DG-5128 was found to be effective in reducing fasted blood glucose levels and in suppressing post- prandial or glucose-induced hyperglycemia in various animal species including mice, rats, dogs and monkeys as well as genetically diabetio mice of the AYKK strain [2]. The effect of this compound was demon- strated to be based on stimulation of insulin release from the pancreatic islets [3], like a sulfonylurea hy- poglycemic, tolbutamide. But some fundamental difference was observed in mode of stimulation of insulin release between DG-5128 and tolbutamide.

* To whom correspondence should be addressed.

DG,5128 was found to stimulate the glucose-primed insulin release through antagonizing at the a2-adreno- receptors on the B cells [3 ].

Further studies in this laboratory have recently revealed a more interesting difference between these compounds. The present paper describes the effects of DG-5128 on the glucose-primed insulin secretion from the perifused rat pancreatic islets and on incor- poration of [3H]leucine into proinsulin and insulin in the isolated islets in comparison with those of tolbutamide.

Materials and Methods

2 [2-(4,5-Dihydro.1 H-imidazol-2-yl)-I -phenyleth- yl] pyridine dihydrochloride sesquihydrate (DG-5128) was synthesized at this institute [1]. Reagent used in this study included tolbutamide (Shin Nihon Yaku- gyo Co., Tokyo), L-[4~5)H]leucine (47 Ci/ retool, New England Nuclear, Boston), L-aminoacids (Nihon

0304-4165/81/0000-0000/$02.50 © 1981 Elsevier/North-Holland Biomedical Press

264

Rikagaku Takuhin Co., Tokyo), Sephadex G-50 (fine) (Pharmacia Fine Chemicals AB, Uppsala), aprotinin (Trasylol, Bayer Co., Leverkussen), rat insulin (Novo Industri A/S, Copenhagen), bovine serum albumin (Daiichi Pure Chemicals Co., Tokyo) and collagenase (Type I, Worthington Biochemicals Corp., Freehold). Other chemicals used were of analytical grade.

The Langerhans islets were isolated from the pancreas of male Wistar : STD rats weighing 200- 300 g according to the method of Lacy et al. [4] with a slight modification. Collagenase digestion of the pancreatic tissue was carded out in the presence of aprotinin (250 KIE */ml) for 10-15 min. The iso- lated islets were preincubated in Krebs-Ringer bicar- bonate buffer (pH 7.4) containing glucose (0.5 mg/ ml) at 37°C for 30 min. The islets were transferred into Hanks' solution, then used for the experiments.

The incubation mixture for the insulin release reaction contained 3-5 islets each in 1 ml of a me- dium in a siliconized test tube. The medium consisted of Krebs-Ringer bicarbonate buffer (pH 7.4) contain- ing bovine serum albumin (3 mg/ml) and glucose (2 or 5 mg/ml). The reaction was carried out by shaking at 150 strokes/min at 37°C under gas phase of 95% 02/5% C02. The compounds to be tested were dis- solved in the incubation medium.

Perifusion of the isolated islets was carried out according to the method of Lacy et al. [4] with a slight modification. The perifusion medium was Krebs-Ringer bicarbonate buffer (pH 7.4) containing bovine serum albumin (3 mg/ml), glucose (0.5 mg/ ml) and aprotinin (500 KIE/ml). The medium was bubbled with 95% 02/5% CO2. The experiment was performed by a paired chamber system consisting of 2 plastic chambers of the same size (2 ml). Both chambers contained 30-40 islets suspended in approx. 0.8 ml of the medium. The islets in both chambers were perifused with the medium simul- taneously at 37°C at a flow rate of 0.8 ml/min by a peristaltic pump (Mitsumi Scientific Industry, Tokyo) After 30 rain, the perifusion medium was exchanged to that containing glucose at 5 mg/ml with or without a test compound for further perifusion for 2 h. The pefifusates were collected continuously every minute for determination of immunoreactive insulin.

Biosynthesis of insulin was studied by measuring

* Kallikrein inhibitor units.

incorporation of [aH]leucine into proinsulin and insulin in the isolated islets according to the reported method [5,6]. The isolated pancreatic islets (10-15 islets each) were incubated in the presence of 50 #Ci of [3H]leucine in 1 ml of Krebs-Ringer bicarbonate buffer (pH 7.4) supplemented with bovine serum albumin (3 mg/ml), glucose (1 or 3 mg/ml) and 17 aminoacids including glycine, alanine, valine, iso- leucine, serine, threonine, proline, cysteine, methio- nine, phenylalanine, tyrosine, tryptophane, aspartic acid, glutamic acid, lysine, histidine and arginine (20 /~g/ml). After incubation by shaking at 150 strokes/ min at 37°C for 1 h under gas phase of 95% 02/5% CO2, aliquots of the medium were taken for deter- mination of immunoreactive insulin. The islets were washed three times with cold Hanks' solution, and 8-10 washed islets were transferred into 1 ml of 1 M acetic acid. The islets were disrupted by ultra- sonication for 45 s (20 kHz, 150 W, Ohtake Seisa- kusho Co., Tokyo) and centrifuged at 10000 ×g for 15 min. Proinsulin and insulin were separated from the supernatant solution by gel f'dtration on a Sepha- dex G-50 column (1.2×50 cm) which had been equilibrated with 1 M acetic acid. Radioactivity and insulin-related antigen were determined in each frac- tion. Insulin-related antigen was radioimmunoassayed by the use of 'Insulin Riakit' (Dainabot RI Institute, Tokyo) after removal of acetic acid by lyophilyzation of the eluates.

Data are presented by mean -+S.E. The results were analyzed statistically by the method of Welch [7].

Results

Effect o f DG-5128 and tolbutamide on insulin secre- tion from the isolated pancreatic islets

DG-5128 was found to stimulate the glucose- primed insulin secretion from the islets especially at a high glucose concentration (5 mg/ml), as shown in Fig. 1. On the other hand, tolbutamide showed a slight tendency to stimulate the insulin secretion only in the early phase of the reaction (Fig. I). In another experiment with a larger number of the islets, the effect of tolbutamide was significant at a low concen- tration of glucose (see Table I).

Figs. 2 and 3 show the insulin secretion profdes during perifusion of the isolated islets with glucose and the effect of DG-5128 and tolbutamide on the

265

A

~° I T/a~

o 6'0 9'0 Incubation time (min)

B

/ " 100

O 30 60 90 120 ImcLJbotion t#me (rain)

F~. 1. Effect of DG-S128 and tolhutamide on the glucose- primed insulin secretion from the isolated rat pancreatic islets. The isolated islets (3-5 islets) were incubated in I ml of Krebs-Ringer bicarbonate buffer (pH 7.4) in the presence of glucose at 2 mg/ml (A), or 5 rag/m1 (B), with or without 0.1 mM DG-5128 or 1 mM tolbutamide at 37*(3 for 2 h..At every 30 rain, aliquots of the incubation medium were assayed for immunoreactive insulin (IRI). Each point represents mean ±S.E. of three observations. (o o), con- trol; (e :), DG-5128 and (A A) tolhutamide. ** P < 0.01 versus control.

4 ~ Glucose- Glucose 5rng/m[ QSmgml . . .

r ~ z

E 5 +DG 5128 r i ~ (0.1 raM) ~/" ,e,, , , e - ~ ' : ' , ,

. T ,, i ", .d" i b-'O

. coot,o,

-20 0 20 40 60 80 100 120 Perifusion time (rain)

-Glucose~ Glucose 5 mg/ml 0.5 mg,'ml ~ ~

~-TOLBUTAMIDE T (lmM) [ ,~,,~ 2 = f ± 1 " ~

~ / 1 z Control y,'f" 7

" ~ ' " ~ ' ,~¢;~ "~' , , , • , l : , $ . E I ' $.E. , ~ ,

-20 0 20 40 60 80 100 120 Perifusion time (rain)

Fig. 3. Effect of tolbutamide on the glucose4ndueed insulin secretion from the isolated mt pancreatic islets during perifu- sion. The experimental conditions were the same as in Fig. 2, except for 1 mM tolbutamide in place of DG-5128. Each point represents mean ±S.E. of six experiments. ( . . . . . 9, control; ( ), tolbutamide. Open circles on the solid line indicate significant difference (P<0.05) from the corre- sponding control values.

secretion profile. Insulin was secreted in 2 phases with the first small peak followed by the second sus- tained large peak of secretion. When DG-5128 was

added to the perifusion medium at 0.1 mM, the glucose-induced insulin secretion was markedly increased in both phases (Fig. 2). But, when tolbuta- mide was added to the medium at 1 mM, the insulin secretion was stimulated in the first phase but not in

the second phase of secretion (Fig. 3).

Effect o f DG-5128 and tolbutamide on biosynthesis of proinsulin in the isolated pancreatic islets

Biosynthesis of proinsulin was examined by mea- suring incorporation of [3H]leucine hito the pro-

Fig. 2. Effect of DG-5128 on the glucose-induced insulin secretion from the isolated rat pancreatic islets during perifu- sion. The isolated islets (30-40 islets) were perifused with Krebs-Ringer bicarbonate buffer (pH 7.4) containing 0.5 rag/ ml glucose for 30 rain. The perifusion medium was then exchanged to the medium containing 5 mg/ml glucose with or without 0.1 mM DG-5128. The perifusates were collected every minute at a flow rate of 0.8 ml/min for determination of immunoreactive insulin (IRI). Each point represents mean ±S.E. of six experiments. Standard error was omitted from the points in the initial phase of the reaction. ( . . . . . -), con- trol; ( ), DG-5128. Open circles on the solid line indi- cate significant difference (P < 0.05) from the corresponding control values.

266

TABLE I

EFFECT OF DG-5128 AND TOLBUTAMIDE ON THE INCORPORATION OF [3H]LEUCINE INTO PROINSULIN AND INSULIN IN THE ISOLATED RAT PANCREATIC ISLETS

The isolated rat pancreatic islets (10-15 islets) were incubated with [3H]leucine (50 #Ci) in 1 ml of Krebs-Ringer bicarbonate buffer (pH 7.4) containing other 17 naturally occurring aminoacids (20 rig each/ml), 3 mg/ml albumin and 1 or 3 mg/ml glucose. 0.1 mM DG-5128 or 1 mM tolbutamide was added to the incubation medium as indicated. After incubation at 37°C for 1 h, pro- insulin and insulin were extracted from the islets and separated by gel filtration shown in Fig. 4. Radioactivity of the pooled pro- insulin and insulin fractions was measured. Immunoreactive insulin released into the medium and in the insulin fraction from the islets was also measured, n, number of observations. The figures are mean + S.E.

Glucose Addition N Immunoreactive [ 3 H ] Leucine incorporated Immunoreactive insulin released (dpm/islet per h) into: insulin content 0zU/islet per h) in insulin fraction

Proinsulin Insulin (~U/islet)

1 mg/ml None 13 DG-5128 7 Tolbutamide 6

3 mg/ml None 7 DG-5128 4 Tolbutamide 4

69_+10 590_+135 245_ + 57 1720_+ 73 2 2 6 + 2 9 b 548-+166 190-+ 43 1763_4- 89 154-+16b 200-+ 94a 79 _4- 14a 1739_+147 267 -+ 15 2 142 _+ 350 765 _+ 121 1 682 _4- 150 364 _+ 18 b 2 762 _4- 423 1 057 -+ 251 1 671 _4- 153 313 -+ 25 1 734 _+ 321 626 _+ 153 1 610 _+ 221

a p < 0.05 and b p < 0.01 versus no addition.

L300

.~ 200

& ~I00

Blue dextren Rot insulin A I { Leucine

Proinsulin

Insulin /

30 40 50 60 Fr'(]ct ion No

70

400

L ~ o 6

E

30 - ~

Proinsulin

_ ~ L e u c i n e

50 60 70 Frection No.

Fig. 4. Effect of DG-5128 and tolbutamide on the incorpora- tion of [3 H ] leucine into proinsulin and insulin in the isolated rat pancreatic islets. The isolated islets (10-15 islets) were incubated with 50 /ICi of [3H]leucine in 1 ml of Krebs-

insulin and insulin fractions after incubation of the isolated islets with labelled leucine and other 17 aminoacids. Proinsulin and insulin were extracted from the islets and separated by t'titration on Sepha- dex G-50 gel. As shown in Figs. 4A and B a greater incorporation of [3H]leucine was observed at a higher concentration of glucose (3 mg/ml) than at a low glucose concentration (1 mg/ml) in the incubation medium. When tolbutamide (1 mM) was added to the medium during incubation of the islets, incorporation of [3H]leucine into proinsulin and insulin was inhibited, especially at a low concentration of glucose

Ringer bicarbonate buffer (pH 7.4) containing another 17 naturally occurring aminoacids (20 $ig each/ml), 3 mg/ml albumin and glucose at 1 mg/ml (A) or 3 mg/ml (B). 0.1 mM DG-5128 or 1 mM tolbutamide was added to the incubation medium. After incubation at 37°C for 1 h, proinsulin and insulin were extracted from the islets and separated by gel filtration on a Sephadex G-50 column (1.2)<50 cm). The filtrates were collected in fractions of 1 ml each at a flow rate of 10 ml/h, and radioactivity was measured. Proinsulin and insulin peaks were identified by insulin-related antigen and elution volume of rat insulin in a separate gel filtration as indicated by an arrow at the top. Each point represents mean of 13 (control, o o), 7 (DG-5128, • •) and 6 (tolbutamide, • 1) in (A), and 7 (control, o o), 4 (DG-5128, • e) and 4 (tolbutamide, • A) in (B).

(1 mg/ml) in the medium. On the other hand, DG- 5128 hardly affected the incorporation of [3H]- leucine, although it rather slightly stimulated the incorporation of [3H]leucine at a high glucose con- centration (3 mg/ml) in the medium.

Table I summarizes the effect of DG-5128 and tolbutamide on the amount of incorporated [3H]- leucine into the pooled fractions of proinsulin and insulin at the 2 levels of glucose concentration. DG- 5128 was found to exert no significant effect on the incorporation of labelled leucine into proinsulin and insulin, although it tended to stimulate the incorpora- tion at a higher glucose concentration. Tolbutamide significantly inhibited the incorporation of [3H]- leucine into both proinsulin and insulin at a low glucose concentration. Insulin content in the incu- bated islets was not affected by these compounds. DG-5128 and tolbutamide stimulated insulin secre- tion into the medium during incubation at a low glucose concentration. DG-5128 also stimulated it even at a high glucose concentration.

Discussion

The present study has demonstrated a significant difference in the effect on insulin secretion and bio- synthesis between tolbutamide and a new hypogly- cemic imidazoline derivative, DG-5128. DG-5128 secreted more insulin after longer incubation of the islets especially at a high glucose concentration (5 mg/ml), while tolbutamide stimulated insulin release only in the earlier phase of incubation and at lower glucose concentrations (Fig. 1 and Table I). Tolbuta- mide was reported to stimulate the glucose-primed insulin secretion from the pancreatic islets only in the first phase of the reaction [4,8] and to inhibit the glucose-induced biosynthesis of proinsulin in the islets at low glucose concentrations [5,9-12]. These effects of tolbutamide were reproduced in the present study. In contrast to tolbutamide, DG-5128 was found to stimulate the glucose-induced insulin secretion in both of the first and the second phases of the reaction (Fig. 2) and to exert no significant effect with a slight stimulatory tendency on the incorpora- tion of [3H]leucine into the proinsulin and insulin fractions in the islets (Fig. 4 and Table I).

The biphasic insulin secretion induced by glucose is known to be contributed to by intracellular cal-

267

cium store and calcium influx in B cells of the islets [13,14]. Monophasic insulin secretion by tolbuta- mide was reported to be due to a rapid and transient monophasic calcium entry into the cells followed by the delayed reduction in calcium permeability to the cells [15]. In the previous study [3], we demon- strated the a2-adrenergic antagonist character of DG- 5128 in stimulation of the glucose-primed insulin secretion from the islets. This finding suggested a similarity in the mode of stimulation of insulin secre- tion between this compound and 'islet-activating pro- tein' [16] which was reported to reverse t~-adrenergic inhibition of insulin release from the islets [17] and to active natural calcium ionophore on the B cell mem- branes [18]. Our recent preliminary experiment has indicated that DG-5128 effectively prevents the vera- pamil-induced inhibition of insulin secretion from the islets in the presence of glucose at 3 mg/ml (control, 51.7 + 1.3 /.tU insulin released/islet; verapamil, 0.1 raM, 16.8 + 4.0, P < 0.05; verapamil plus DG-5128, 0.1 raM, 50.8 +_ 20.1; n = 3). These and the present findings may suggest a possible involvement of cal- cium control in stimulation of insulin secretion by DG-5128, although a functional relationship between ~-adrenergic regulation (adenylate cyclase) and cal- cium flux in the cells is not clear at the present time. Further studies are now under way along this line to elucidate the mode of action of DG-5128 in stimula- tion of insulin secretion.

The inhibitory action of tolbutamide on biosyn- thesis of proinsulin in the islets has been discussed in relation to decreased oxidation of glucose [10,11 ] or decreased incorporation of radioactive uridine into RNA [5] in the islets. DG-5128 was found to be free of such unfavorable effects on insulin biosynthesis in the islets.

References

1 Ishikawa, F. (1980) Chem. Pharm. Bull. (Tokyo) 28, 1 394-1 402

2 Kameda, K., One, S. and Abiko, Y. (1981) Arzneim.- Forsch., in the press

3 Kameda, K., One, S., Koyama, I. and Abiko, Y. (1981) Acta Endocrinol., in the pxess

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

5 Schatz, H., Maier, V., Hinz, M., Nierle, C. and Pfefffer, E.F. (1972) FEBS Lett. 26,237-240

268

6 Laube, H., Schatz, H., Nierle, C., Fussg/inger, R. and Pfeiffer, E.F. (1976) Diabetologia 12,441-446

7 Welch, B.L. (1947) Biometrika 34, 28-35 8 Curray, D.L., Bennett, L.L. and Grodsky, G.M. (1968)

Endocrinology 83, 572-584 9 Morris, G.E. and Korner, A. (1970) Biochim. Biophys.

Acta 208,404-413 10 Dunbar, J.C. and Fo~, P.P. (1974) Diabetologia 10, 27-

35 11 Levy, J. and Malaisse, W.J. (1975) Biochem. Pharmacol.

24,235-239 12 Tanese, T., Lazarus, N.R., Devrim, S. and Recant, L.

(1970) J. Clin. Invest. 49, 1 394-1 404

13 WoUheim, C.B., Kikuchi, M., Renold, A.E. and Sharp, G.W.G. (1978) J. Clin. Invest. 62,451-458

14 Kikuchi, M., Wollheim, C.B., Siegel, E.G., Renold, A.E. and Sharp, G.W.G. (1979) Endocrinology 105, 1 013- 1019

15 Henquin, J.-C. (1980) Diabetologia 18, 151-160 16 Yajima, M., Hosoda, K., Kanbayashi, Y., Nakamura, T.,

Nogimori, K., Nakase, Y. and Ui, M. (1978) J. Biochem. 83,295-303

17 Katada, T. and Ui, M. (1980) J. Biol. Chem. 255, 9580- 9588

18 Katada, T. and Ui, M. (1979) J. Biol. Chem. 254,469- 479