509 Glibenclamide but not other sulphonylureas stimulates release of neuropeptide Y from perifused rat islets and hamster insulinoma cells

R N Kulkarni, Z-L Wang, R-M Wang, D M Smith, M A Ghatei and S R Bloom Francis Fraser Labs, Department of Metabolic Medicine, Imperial College School of Medicine, Hammersmith Campus, Du Cane Road, London W12 0NN, UK (Requests for offprints should be addressed toSRBloom; Email: [email protected]) (R Kulkarni is now at Research Division, Joslin Diabetes Centre and Department of Medicine, Harvard Medical School, Boston, Massachusetts 02215, USA)

Abstract We have studied the effects of first and second generation n=6) enhanced glucose (8 mmol/l)-stimulated insulin sulphonylureas on the release of insulin and neuropeptide release. However, only glibenclamide stimulated the re- tyrosine (NPY) from hamster insulinoma tumour (HIT- lease of NPY from the islets (control 3·4
0·8 vs GLIB T15) cells and isolated rat islets. In the presence of 24·5
5 attomol/min per islet, P<0·01, n=6). Similar 5·5 mmol/l glucose all sulphonylureas stimulated insulin results were obtained in islets isolated from dexamethasone- release from the HIT cells (P<0·01 ANOVA, nd4) but treated rats. Glibenclamide treatment of HIT cells showed a only glibenclamide (GLIB, 10 µmol/l) stimulated the prompt insulin release (10 min) while NPY secretion was release of NPY (mean
... control 11·1
1·3 vs GLIB slower (60 min), suggesting that internalization of the 28·4
4·1 fmol/h per 106 cells, P<0001, n=16). In iso- sulphonylurea is required to stimulate NPY release. Gliben- lated perifused rat islets both glibenclamide (10 µmol/l) clamide, the most common oral therapeutic agent in type 2 (control 3·5
0·3 vs GLIB 6·3
0·2 fmol/min per islet, diabetes mellitus, is associated with release of the autocrine P<0·01, n=6) and tolbutamide (50 µmol/l) (control insulin secretion inhibitor, NPY. 4·7
0·1 vs TOLB 6·7
0·3 fmol/min per islet, P<0·01, Journal of Endocrinology (2000) 165, 509–518

Introduction islets has been demonstrated to occur secondarily to alterations in ionic fluxes in islets pre-exposed to gliben- Sulphonylureas enhance
-cell insulin release by blocking clamide but not tolbutamide (Rabuazzo et al. 1992). The the ATP-dependent K+ channel and are widely used in mechanism underlying the
-cell hyporesponsivity is not the treatment of type 2 diabetes mellitus. In chronic clear. However, among the sulphonylureas, glibenclamide therapy the mechanism of action of sulphonylureas is less is exceptional in accumulating progressively in
-cell rich clear. Studies have shown that the long-term use of these pancreatic islets in mice (Hellman et al. 1984) and auto- oral hypoglycaemic agents does not increase basal insulin radiography using [3H]glibenclamide shows silver grains to release or enhance insulin secretion in response to meta- be preferentially associated with
-cells compared with bolic stimuli in patients with type 2 diabetes (DeFronzo & non-
endocrine cells (Carpentier et al. 1986). Thus it is Simonson 1984, Karam et al. 1986). In experiments on possible that internalization of glibenclamide (Gorus et al. normoglycaemic rats, chronic treatment with sulphonyl- 1988), as well as causing a prolonged insulin secretory ureas results in lower insulin content and reduced insulin effect compared with tolbutamide, in some manner, con- responsiveness to metabolic stimuli (Duckwoth et al. 1972, tributes towards impaired insulin release following chronic Fineberg & Schneider 1980, Wajchenberg et al. 1980). exposure of islets to the drug. One interpretation of the lower insulin content in this Neuropeptide tyrosine (NPY), a 36-amino acid pep- context is that it is secondary to excessive stimulation of tide, is widely distributed in the central and peripheral
-cell function. A study reporting the effect of chronic nervous systems and in the pancreas (Adrian et al. 1983, exposure of sulphonylureas on insulin secretion shows a Gibson et al. 1984, Carlei et al. 1985). We and others have reversible impairment of insulin release in isolated rat islets shown that NPY mRNA is present in rat and human islets after a 24 h pre-exposure to glibenclamide (Gullo et al. and clonal
-cell lines (Jamal et al. 1991, Wang et al. 1994, 1991). Furthermore, desensitization of insulin release in rat Bennet et al. 1996, Myrsen-Axcrona et al. 1997). Both

Journal of Endocrinology (2000) 165, 509–518 Online version via http://www.endocrinology.org 0022–0795/00/0165–509  2000 Society for Endocrinology Printed in Great Britain

Downloaded from Bioscientifica.com at 09/24/2021 09:14:19AM via free access 510 R N KULKARNI and others · Effect of sulphonylureas on insulin and NPY release

in vivo and in vitro studies have demonstrated that NPY pre-incubated with serum-free and glucose-free medium inhibits glucose-stimulated insulin secretion (Petterson for 2 h. Cells were then incubated in the absence or et al. 1987, Skoglund et al. 1991), probably acting via the presence of different concentrations of sulphonylureas in Y1 receptor (Morgan et al. 1998). The inhibitory effect of medium containing 5·5 mmol/l glucose. Culture medium NPY on insulin release in perifused islets isolated from was collected, centrifuged at 500 g for 5 min to remove both normal and dexamethasone-treated rats suggests it has cell debris and the supernatant stored at 20 Cfor an autocrine/paracrine role in the endocrine pancreas subsequent RIA for insulin and NPY. Cell viability and (Wang et al. 1994). In the same study, treatment of rat cell counts were assessed at the end of each experiment by islets with NPY antisera increased insulin release and it trypan blue exclusion using a haemocytometer. was suggested that NPY is an endogenous tonal inhibitor of insulin (Wang et al. 1994). Treatment of rats, islet isolation and perifusion In the present study, we have examined the effects of first and second generation sulphonylureas on their ability Male Wistar rats (250–300 g) were housed in the Biologi- to stimulate insulin and NPY secretion in two different cal Services Unit (ICSM, London, UK) and had free access cell models. We have used rat islets that retain the to water and food. All experiments carried out on the physiological micro-architecture between the different animals complied with the regulations of the British Home islet cell types. In addition, to study the direct effects Office. Two series of experiments were carried out. The of the sulphonylureas on the
-cells we have used first series involved the study of the effects of the sulpho- the predominantly insulin-secreting glucose-responsive nylureas on insulin and NPY secretion in islets isolated hamster insulinoma cell line HIT-T15. from normal rats. In a subsequent series, the effects of the sulphonylureas were tested in islets isolated from a group of rats which had been injected subcutaneously with dexa- Materials and Methods methasone (4 mg/kg per day) for 10 days and the observed effects were compared with islets isolated from control Chemicals animals injected with 0·9% saline. Rats were killed by CO2 asphyxiation and islets were isolated by means of Glibenclamide, tolbutamide, chlorpropamide, dimethyl collagenase digestion using a modified method of Lacy and sulphoxide (DMSO), dexamethasone and all other chemi- Kostianovsky as described earlier (Lacy & Kostanovsky cals and reagents, unless otherwise specified, were of 1967, Wang et al. 1994). Briefly, following isolation, Analar grade and obtained from Sigma Chemical Co. islets were perifused with a 50:50 vol/vol mixture of (Poole, Dorset, UK). Glipizide was a gift from Pfizer Krebs Ringer buffer (KRB) and CMRL-1066 culture (Sandwich, Kent, UK) and gliclazide was obtained from medium (GIBCO-BRL, Paisley, UK) supplemented Servier (Fulmer, Slough, UK). with BSA (4 g/l), bacitracin (40 mg/l), ascorbic acid (80 mg/l), aprotinin (50 000 Kallikrein international Cell culture units/l, Trasylol, Bayer AG, Leverkusen, Germany) and glucose (2·8 mmol/l). The buffer perifusing the islets was HIT-T15 (HIT) cells were obtained from the American gassed with O2:CO2, 95:5. The perifusion chambers were Type Culture Collection (ATCC, Rockville, MD, USA). immersed in a water bath at 37 C, and each chamber was Cells were routinely cultured in RPMI 1640 medium perifused with buffer at 0·4 ml/min and fractions were (GIBCO-BRL, Paisley, UK) with 10%, vol/vol, foetal calf collected for RIA. serum (FCS) supplemented with antibiotics as described earlier (Kulkarni et al. 1995a,b). All experiments were carried out between the passages 64 and 70. Plastic ware RIA was obtained from NUNC (Roskilde, Denmark). Culture media were assayed in duplicate for insulin by an established RIA using porcine insulin standard and Glu5 antibody at a final dilution of 1:150 000 (Albano et al. Experimental protocol 1972). The assay had a detection limit of 2 fmol/tube at Glibenclamide, glipizide and gliclazide were dissolved in 95% confidence limits with intra- and inter-assay coef- DMSO to a final concentration of 0·01% (vol/vol) and ficients of variation of 8·2 and 11·1% respectively. NPY in tolbutamide and chlorpropamide were dissolved in 50% the culture media was assayed using antibody YN10 at a ethanol (final concentration 1%, vol/vol) (Leclercq-Meyer final dilution of 1:30 000 (Allen et al. 1984). Aliquots of et al. 1991). Experiments involved plating cells, at a perifusion effluent were rotary vacuum-dried (Savant density of 3 to 4105 cells/well, in 24-well plates in Instruments, NY, USA) before analysis for NPY. The a total volume of 1 ml (Kulkarni et al. 1995a, Kulkarni assay had a detection limit of 0·2 fmol/tube at 95% et al. 1997). After reaching 70% confluence, cells were confidence limits and intra- and inter-assay coefficients washed with glucose-free RPMI 1640 medium, and of variation of 12·8 and 15·7% respectively. Potential

Journal of Endocrinology (2000) 165, 509–518 www.endocrinology.org

Downloaded from Bioscientifica.com at 09/24/2021 09:14:19AM via free access Effect of sulphonylureas on insulin and NPY release · R N KULKARNI and others 511 cross-reactivity of glibenclamide in the NPY assay was effect at 1 µmol/l. Glipizide and gliclazide stimulated evaluated by performing the RIA in the presence and insulin release with a threshold effect at a dose of absence of the highest dose (10 µmol/l) of sulphonylurea 0·5 nmol/l. Cells treated with glipizide responded with a used in the experiments and also at ten times this maximal threefold rise in insulin release at 10 nmol/l that concentration. did not significantly change with further addition of glipizide. Northern blot analysis and peptide content in cell extracts NPY release Among the sulphonylureas tested, only HIT cells grown in 60 mm culture dishes were treated glibenclamide stimulated the release of NPY compared with the sulphonylureas for 12 h and harvested with with untreated controls (Fig. 2B). There was a 1·4-fold 2·5 ml of guanidinium isothiocyanate solution. Samples   increase in NPY at the dose of 0·5 µmol/l (control were stored at 70 C for subsequent extraction of RNA 11·1
1·3 vs glibenclamide (GLIB) 16·1
1·1 fmol/h per using the acid guanidinium isothiocyanate–phenol– 106 cells; P<001; n=16) and a 2·5-fold increase at chloroform procedure (Chomczysnki & Sacchi 1987) or 1 µmol/l (GLIB 27·3
3·2 fmol/h per 106 cells; P<0·001; treated with acid ethanol overnight at 4 C for estimation + n=16). There was no significant further increase in NPY of peptide content (Kulkarni et al. 1995a). Poly(A) RNA at 10 µmol/l compared with 1 µmol/l (1 µmol/l NPY was prepared from total RNA by oligo(dT12–18) cellulose 27·3
3·2 vs 10 µmol/l NPY 28·4
4·1 fmol/h per 106 chromatography (Sambrook et al. 1989). Filters were cells; P=NS; n=16). The EC50 for NPY release was initially probed for NPY mRNA using a specific rat NPY 181
9 nmol/l. There was no significant increase in NPY cRNA probe, then stripped and re-probed for hamster release in response to any of the other sulphonylureas at insulin mRNA using a cDNA probe (Bretherton-Watt the maximal tested concentrations (for concentrations et al. 1989, Kulkarni et al. 1995a). Total RNA from rat see Fig. 1A-D) (control 11·1
1·3; gliclazide 9·1
1·2; hypothalamus was also probed for NPY and served as a glipizide 10·8
2·1; tolbutamide 12
1·2; chlorpropa- reference for NPY mRNA (Jamal et al. 1991). Normaliz- mide 10·3
0·8 fmol/h per 106 cells; P=NS, control vs ation of stripped filters was carried out by rehybridization gliclazide or glipizide or tolbutamide or chlorpropamide, with radiolabelled oligo(dT12–18) and the bands were n=4–6). quantified by an Autograph direct capture surface counter using Saxon Micro software as described earlier (Austin Time-course kinetics of insulin and NPY release in et al. 1994, Kulkarni et al. 1995a). Beta actin was used as response to glibenclamide To evaluate whether the control for loading (Austin et al. 1994). Results are insulin and NPY release in response to glibenclamide expressed in arbitrary relative units of insulin or NPY stimulation showed different kinetics, HIT cells were mRNA. incubated with 10 µmol/l of the sulphonylurea for 10, 30, 60, 120 and 240 min. Figure 3A shows a significant Statistical methods increase in insulin release beginning at 10 min compared
Cell and islet experiments were carried out on 6–16 with control untreated cells (control 0·44 0·05 vs GLIB
6 < different occasions in triplicate. Means of triplicate values 0·67 0·05 pmol/10 cells, P 0·02, n=4–6). However, were used to obtain an n=1. Results are expressed as NPY release was comparable to untreated cells up to means
... Experiments with gliclazide were carried 30 min and was significantly increased only at 60 min

6 out on four different occasions in triplicate due to a limited (control 20·1 2·6 vs GLIB 51·8 5·2 fmol/10 cells, < supply of the compound. Data was analysed by one-way P 0·01, n=4–6, Fig. 3B). These data indicate that intern- ANOVA with subsequent post hoc Tukey tests. Values are alization of the sulphonylurea is likely to be necessary to considered significantly different if P<0·05. promote NPY secretion. There was no cross-reactivity of the sulphonylurea in the NPY RIA at the concentrations used (or at 100 µmol/l) which excludes the possibility that Results the increase in NPY release in response to glibenclamide is secondary to an interference in the assay. Clonal cells Insulin release All sulphonylureas stimulated insulin Effects of glibenclamide on insulin and NPY mRNA secretion in a dose-dependent manner (Fig. 1A-D, Fig. HIT cells incubated with glibenclamide (1 µmol/l) in the 2A). The second generation sulphonylureas were more presence of glucose for 12 h did not show a significant potent than tolbutamide and chlorpropamide (CHLOR) increase in either insulin or NPY mRNA compared with in stimulating insulin release. A maximal sevenfold in- untreated cells. The size of the NPY transcript in the HIT crease was observed with 1 mmol/l of chlorpropamide cells was confirmed by the presence of a band of similar (control 0·4
0·1 vs CHLOR 2·9
0·8 pmol/h per 106 size in the lane loaded with total RNA prepared from rat < cells; P 0·01; n=6). The calculated EC50 for glibencla- hypothalamus (Jamal et al. 1991) (Fig. 4, upper panel). The mide was 76·9
4·6 nmol/l with the observed maximal positive control (11 mmol/l glucose) showed a 1·5-fold www.endocrinology.org Journal of Endocrinology (2000) 165, 509–518

Downloaded from Bioscientifica.com at 09/24/2021 09:14:19AM via free access 512 R N KULKARNI and others · Effect of sulphonylureas on insulin and NPY release

Figure 1 Dose-dependent increase in insulin secretion in response to treatment with different sulphonylureas. HIT cells were treated with increasing concentrations of second generation (A) gliclazide (n=4), (B) glipizide (n=6), or first generation (C) tolbutamide (n=6) and (D) chlorpropamide (n=6) sulphonylureas for 120 min. *P<0·02, **P<0·01 vs control. Results are presented as means
S.E.M.

increase in insulin mRNA (control, 5·5 mmol/l glucose was observed when the glucose concentration was raised 100
6% vs 11 mmol/l glucose 167
10%, P<0·05, from 2·8 (time 0 to 40 min) to 8 mmol/l (time 40 to n=5) (Fig. 4, lower panel). In parallel experiments, 80 min). Assessment of viability of islets at the end of each measurement of peptide content in the glibenclamide- experiment showed a significant decrease in insulin release treated cells showed a significant reduction in insulin (time 161 to 200 min) when glucose was lowered to (control 11·21
1·0 vs GLIB 9·2
0·9 pmol/106 cells, 2·8 mmol/l and a further significant elevation in insulin P<0·05, n=3). By contrast, although a reduction was also release (time 201 to 220 min) when glucose concentration evident in the NPY content this did not reach statistical was raised to 16 mmol/l. significance (control 94·4
8·8 vs GLIB 85·1
7·2 fmol/ During the treatment period (time 80 to 120 min), 106 cells, P=ns, n=3). addition of glibenclamide at 10 µmol/l showed a further 80% increase in insulin secretion from treated islets com- pared with islets from untreated chambers (control Isolated islets from normal rats 3·5
0·3 vs GLIB 6·3
0·2 fmol/min per islet; P<0·01, Following the observations in the HIT cells, two sulphonyl- n=6) (Fig. 5A). Treatment with tolbutamide (50 µmol/l) ureas – one first generation compound (tolbutamide) and showed a 63% increase in insulin release compared one second generation compound (glibenclamide) – were with islets from untreated chambers (control 4·1
0·1vs examined for their effects on insulin and NPY release in tolbutamide (TOLB) 6·7
0·3 fmol/min per islet; perifused islets isolated from normal rats. P<0·01, n=6) (Fig. 5B).

Effects on insulin release In all the islet experiments Effects on NPY release The secretion pattern of NPY a significant increase in insulin secretion (68%, P<0·001) from perifused rat islets is shown in Fig. 6. The pattern of

Journal of Endocrinology (2000) 165, 509–518 www.endocrinology.org

Downloaded from Bioscientifica.com at 09/24/2021 09:14:19AM via free access Effect of sulphonylureas on insulin and NPY release · R N KULKARNI and others 513

NPY release in islets from normal rats, we therefore carried out the same experiment in islets from rats treated with dexamethasone. Pilot studies using tolbutamide (50 µmol/l) showed a significant increase in insulin release during the treatment period (control 6·4
1·0vsTOLB 9·8
0·8 fmol/min per islet; P<0·01, n=6), while no differences were observed in NPY release (control 10·8
2·0 vs TOLB 11·2
1·1 attomol/min per islet; P=NS, n=6).

Effects on insulin release Insulin secretion increased when glibenclamide (10 µmol/l) was added to the islets (time 81 to 120 min) (control 7·5
0·1 vs GLIB 10·3
0·3 fmol/min per islet, P<0·01, n=6) (Fig. 7A). During the 40 min immediately after treatment with glibenclamide (time 121 to 160 min), insulin secretion was significantly lower in the treated islets (control 7·3
0·1vs GLIB 6·3
0·1 fmol/min per islet, P<0·01, n=6). No differences were observed during the subsequent periods of perifusion.

Effects on NPY release During glibenclamide treat- ment (time 81 to 120 min) an increase in NPY release was observed (control 12·1
0·3 vs GLIB 22·3
0·7 attomol/ min per islet, P<0·01, n=6) (Fig 7B), and continued to be elevated after glibenclamide was withdrawn, during the subsequent 40 min (time 121 to 160 min) (control 11·2
0·5 vs GLIB 17·9
0·5 attomol/min per islet, P<0·01, n=6). No significant differences were observed between treated and untreated chambers during the other periods.

Figure 2 Glibenclamide stimulates the release of (A) insulin and (B) NPY in HIT cells. HIT cells were treated with increasing Discussion concentrations of glibenclamide and incubated for 120 min as described in Materials and Methods. *P<0·02, **P<0·01 vs control. Results are presented as means
S.E.M., n=16. The present study shows a dose-dependent stimulatory effect of the second generation sulphonylurea, glibencla- mide, on insulin and NPY release from clonal HIT cells. NPY release was similar to that reported earlier by us Glibenclamide also increased NPY secretion in islets (Wang et al. 1994) with an inverse correlation to the isolated from normal and dexamethasone-treated rats. glucose concentration in the buffer. NPY release from Treatment with glibenclamide did not significantly glibenclamide treated islets showed an eightfold increase increase insulin or NPY mRNA in HIT cells. The other compared with islets in untreated chambers (control sulphonylureas tested – tolbutamide, chlorpropamide, 3·4
0·8 vs GLIB 24·5
5·9 attomol/min per islet; gliclazide and glipizide – stimulated insulin release but not P<0·01, n=6) (Fig. 6A). No significant differences were NPY release. observed in NPY release from islets treated with tolbuta- First generation sulphonylureas like tolbutamide have mide (TOLB, 50 µmol/l) compared with controls (control been largely replaced by second generation drugs like 2·7
0·3 vs TOLB 3·1
0·4 attomol/min per islet; glibenclamide that are short-acting and more potent P=NS, n=6) (Fig. 6B). (Lebovitz & Feinglos 1983, Paice et al. 1985). Although the sulphonylureas continue to be used in the treatment of type 2 diabetes mellitus, long-term clinical studies have Isolated islets from dexamethasone-treated rats reported a secondary failure in increasing insulin secretion We and others have shown that the islet secretion of NPY (Groop 1992). The cause of this ‘secondary-failure’ has increases when rats are treated with dexamethasone been the subject of several investigations and have been (Wang et al. 1994, Myrsen et al. 1996). Following the ascribed to ‘
-cell exhaustion’ (Schauder et al. 1977, observations in this study that glibenclamide increased Gullo et al. 1991, Rabuazzo et al. 1992) and/or ‘
-cell www.endocrinology.org Journal of Endocrinology (2000) 165, 509–518

Downloaded from Bioscientifica.com at 09/24/2021 09:14:19AM via free access 514 R N KULKARNI and others · Effect of sulphonylureas on insulin and NPY release

Figure 3 Glibenclamide stimulation of HIT cells shows different time-course kinetics of (A) insulin and (B) NPY release. HIT cells were treated with 10 mol/l of glibenclamide and incubated for 10, 30, 60, 120 and 240 min as described in Materials and Methods. *P<0·02, **P<0·01 vs control. Results are presented as means
S.E.M., n=4–6.

desensitization’ to glucose (Hoenig et al. 1986, Bolaffi et al. 1988, Purrello et al. 1989) or other secretagogues (Grodsky 1989). An in vitro experiment (Gullo et al. 1991) has shown that 24 h exposure of islets to 100 nmol/l gliben- clamide did not deplete the insulin content, suggesting that mechanisms other than insulin depletion are operative which contribute to the reduced insulin. HIT cells express and secrete NPY-like immunoreac- tivity (Jamal et al. 1991). Further, NPY has been shown to be present in rat islets and to act as a paracrine factor inhibiting insulin secretion (Wang et al. 1994). Using immunocytochemistry and in situ hybridization, NPY has been localized to insulin-containing cells in rat islets (Waeber et al. 1993, Myrsen et al. 1995, Myrsen-Axcrona et al. 1997) and in addition has been shown to be localized to non-
-cells including islet glucagon and islet somatostatin cells (Teitelman et al. 1993, Myrsen et al. 1995). We observed an increase in NPY release associated with the insulin stimulatory effect, in HIT cells, only in response to glibenclamide but not to other sulphonylureas. This observation was confirmed using isolated perifused rat islets. Both tolbutamide and glibenclamide, however, increased insulin release in rat islets to levels comparable with those reported in other islet studies (Gullo et al. 1991, Rabuazzo et al. 1992). Our observation that among the sulphonylureas only glibenclamide is able to stimulate NPY release may be related to several possibilities. In HIT cells we observed a significant increase in insulin release at a threshold of 0·1 nmol/l of glibenclamide while a 100- fold higher concentration of the sulphonylurea was required to detect a significant increase in NPY release Figure 4 Glibenclamide does not alter insulin or NPY expression compared with controls. Earlier studies have reported that in HIT cells. Representative Northern blot (upper panel) showing glibenclamide is exceptional among the sulphonylurea the effect of glibenclamide (12 h treatment) on NPY and insulin class of drugs to accumulate progressively in insulin- mRNA using poly(A)+ RNA prepared from HIT cells. A significant containing cells in rodent islets (Hellman et al. 1984). To increase in insulin mRNA was observed only in the lane with the positive control (Glu, 11 mmol/l glucose). The lower panel shows test the possibility that accumulation of glibenclamide is a quantitation of insulin mRNA expression in arbitrary units a prerequisite to stimulate NPY release, we performed a (mean
S.E.M., n=5). *P<0·05, glucose vs control.

Journal of Endocrinology (2000) 165, 509–518 www.endocrinology.org

Downloaded from Bioscientifica.com at 09/24/2021 09:14:19AM via free access Effect of sulphonylureas on insulin and NPY release · R N KULKARNI and others 515

Figure 5 Glibenclamide and tolbutamide stimulate insulin Figure 6 secretion from isolated rat islets in a perifusion system. Glucose Glibenclamide but not tolbutamide stimulates NPY concentration was 2·8 mmol/l from 0 to 40 min and 161 to release from isolated rat islets in a perifusion system. Glucose 200 min, 8 mmol/l from 40 to 160 min, and 16 mmol/l from 161 concentration and other details are as described in Fig. 5. Filled to 200 min. The total treatment period was 40 min from 81 to bars represent control islets and hatched bars represent islets 120 min. Filled bars represent control islets and hatched bars treated with (A) glibenclamide (10 mol/l) or (B) tolbutamide represent islets treated with (A) glibenclamide (10 mol/l) or (50 mol/l) during 81–120 min. Glucose concentrations in the (B) tolbutamide (50 mol/l) during 81–120 min. Glucose concen- perifusion buffer are indicated on the right Y-axis (dashed line). *P<0·01, treated vs control. Results are presented as trations in the perifusion buffer are indicated on the right Y-axis
(dashed line). *P<0·01, treated vs control. Results are presented as means S.E.M., n=6. means
S.E.M., n=6. this period (12 h). It is possible that longer periods of treatment and higher concentrations are required to alter time-course experiment using HIT cells. While a prompt the transcription of the insulin and NPY genes. secretion of insulin was observed as early as 10 min of Secondly, in a recent report, Myrsen-Axcrona et al. incubating the HIT cells with glibenclamide, a significant (1997) proposed different pathways for NPY and insulin release of NPY was evident only at 60 min. The delayed secretion in the clonal rat cell line RINm5F. In the release of NPY and the lack of cross-reactivity of the context of treatment of
-cells with sulphonylureas, the sulphonylurea in the NPY RIA supports the concept presence of separate pathways for the two
-cell peptides that this action of the sulphonylurea is likely linked to may be related in part to the reported molecular hetero- internalization. Further, glibenclamide has been shown to geneity and intracellular localization of the sulphonyl- enhance insulin secretion largely by mobilizing stored urea receptor (Ozanne et al. 1995). Glibenclamide, in granules, with newly synthesized hormone representing particular, has been shown to bind to a 140 kDa protein – only a small proportion of this output (Schatz et al. 1975). different from the binding site occupied by other This observation is in agreement with our finding of the sulphonylureas such as glimepiride or other analogues lack of an effect of glibenclamide on either insulin or NPY containing a sulphonyl moiety (Kramer et al. 1994). In mRNA expression, indicating that the drug is probably addition, sulphonylureas have been shown to stimulate effective only in the release of the stored peptides during exocytosis by directly activating components of the www.endocrinology.org Journal of Endocrinology (2000) 165, 509–518

Downloaded from Bioscientifica.com at 09/24/2021 09:14:19AM via free access 516 R N KULKARNI and others · Effect of sulphonylureas on insulin and NPY release

induced a similar increase in NPY to that observed in islets from saline-treated rats. Interestingly, however, after with- drawal of glibenclamide, the NPY release continued to be high in the islets from the dexamethasone-treated group during which insulin release was observed to be signifi- cantly lower. Therefore, similar to the study in the INS-1 cells (Waeber et al. 1993), it is possible that when insulin secretion is increased after treatment with dexamethasone, any additional stimulation such as glibenclamide treatment may lead to an autocrine/paracrine inhibitory mechanism resulting in the secretion of NPY which would restrain over-secretion of insulin. These findings along with the observation that NPY expression is induced by dexa- methasone with a concurrent alteration in insulin release in rat islets and RINm5F cells (Wang et al. 1994, Myrsen-Axcrona et al. 1997) collectively suggest that insulin and NPY secretion from the islets/
-cells is complex and may be regulated differently. Type 2 diabetes patients unable to control hyperglycae- mia on diet therapy alone are usually treated with oral hypoglycaemic agents, such as glibenclamide at a dose of 2·5 to 5 mg per day orally. These patients show serum levels ranging from 400 to 450 nmol/l four hours after ingestion of the drug (Grill et al. 1986, Ikegami et al. 1986). Assuming the concentration of the drug in the blood vessels supplying the pancreas and the islets is similar, this concentration would be in a similar molar range as that reported in this study required to stimulate insulin and NPY release. Taken together the data in the present study shows an Figure 7 Glibenclamide stimulates insulin and NPY release from increase in insulin and NPY release following treatment perifused islets isolated from dexamethasone-treated rats. Glucose with glibenclamide in HIT cells and islets isolated from concentration and other details are as described in Fig. 5. Filled bars represent control islets and hatched bars represent islets normal and dexamethasone-treated rats. It is possible that treated with glibenclamide (10 mol/l) during 81 to 120 min. the released NPY has an autocrine function to prevent (A) Insulin release, (B) NPY release. Glucose concentrations in the excess secretion of insulin similar to the purported function perifusion buffer are indicated on the right Y-axis (dashed line). < of NPY secreted by GLP-1 stimulation in INS-1 cells *P 0·01, treated vs control. Results are presented as (Waeber et al. 1993). It is interesting that the other means
S.E.M., n=6. sulphonylureas did not stimulate NPY release and this may be related to the reported internalization of glibenclamide secretory machinery independent of the sulphonylurea particularly to islet
-cells and/or to the heterogeneity of receptor (SUR):KATP channels (Eliasson et al. 1996). The sulphonylurea receptors. Whether the action of glibencla- presence, localization and characterization of these differ- mide in releasing NPY – an autocrine insulin inhibitory ent mechanisms of action in the HIT cells may be useful to factor – contributes in any way towards the secondary understand the secretory effects of glibenclamide affecting failure associated with the use of the oral hypoglycaemic different
-cell peptides. agent in type 2 diabetes requires further investigation. Finally, an earlier study has shown that GLP-1 (glucagon-like peptide-1) (7–36) amide stimulates insulin and NPY release from INS-1 insulinoma cells (Waeber Acknowledgements et al. 1993). The authors suggested that the incretin hormone may serve to autoregulate insulin release during This study was supported by a grant from the Medical periods of excessive secretion by stimulating a local inhibi- Research Council, UK. The authors would like to thank tor of insulin release such as NPY. Since dexamethasone Pfizer (UK) and Servier (UK) for their donation of the treatment of rats is known to result in excess insulin pure compounds for the studies. RNK was the recipient of secretion (Ogawa et al. 1992), we carried out experiments an Overseas Research Students Award and the Weston to assess the effects of glibenclamide stimulation on islets Scholarship and wishes to thank the Special Trustees of isolated from dexamethasone-treated rats. Glibenclamide Hammersmith Hospital for support.

Journal of Endocrinology (2000) 165, 509–518 www.endocrinology.org

Downloaded from Bioscientifica.com at 09/24/2021 09:14:19AM via free access Effect of sulphonylureas on insulin and NPY release · R N KULKARNI and others 517

References impairs insulin secretion in isolated rat pancreatic islets. Journal of Endocrinological Investigation 14 287–291. Adrian TE, Allen JM, Bloom SR, Ghatei MA, Rossor MN, Roberts Hellman B, SehlinJ&Taljedal IB 1984 Glibenclamide is exceptional GW, Crow TJ, TatemotoK&PolakJM1983Neuropeptide Y among hypoglycaemic sulphonylureas in accumulating progressively distribution in human brain. Nature 306 584–586. in
-cell-rich pancreatic islets. Acta Endocrinologica 105 385–390. Albano DJM, Ekins RP, Maritz G & Turner RC 1972 A sensitive Hoenig LC, MacGregor LC & Matschinsky PM 1986 in vitro precise radioimmunoassay of serum insulin relying on charcoal exhaustion of pancreatic
-cells. American Journal of Physiology 250 separation of bound and free hormone moieties. Acta Endocrinologica E502–E511. 70 487–509. Ikegami H, Shima S, Tanaka A, Tahara Y, Hirota M & Kumahara Y Allen JM, Yeats JC, Adrian TE & Bloom SR 1984 Radioimmunoassay 1986 Interindividual variation in the absorption of glibenclamide in of neuropeptide Y. Regulatory Peptides 8 61–70. man. Acta Endocrinologica 111 528–532. Austin C, Nandha KA, Legon S, Khandan-Nia N, LoG&BloomSR Jamal H, Jones PM, Byrne J, Suda K, Ghatei MA, Kanse SM & 1994 Distribution and developmental pattern of neuromedin U Bloom SR 1991 Peptide contents of neuropeptide Y, vasoactive expression in the rat gastrointestinal tract. Journal of Molecular intestinal polypeptide Y, and beta-calcitonin gene-related peptide Endocrinology 12 257–263. and their messenger ribonucleic acids after dexamethasone Bennet WM, Wang ZL, Jones PM, Wang RM, James RF, London treatment in isolated rat islets of Langerhans. Endocrinology 129 NJ, Ghatei MA & Bloom SR 1996 Presence of neuropeptide Y 3372–3380. and its messenger ribonucleic acid in human islets: evidence for a Karam JH, Sanz N, SalamonE&NolteMS1986Selective possible paracrine role. Journal of Clinical Endocrinology and unresponsiveness of pancreatic beta-cells to acute sulfonylurea Metabolism 81 2117–2120. stimulation during sulfonylurea therapy in NIDDM. Diabetes 35 ffi Bola JL,BrunoL,HeldtA&Grodsky GM 1988 Characteristics of 1314–1320. desensitisation of insulin secretion in fully in vitro systems. Kramer W, Muller G, Girbig F, Gutjahr U, Kowalewski S, Hartz D Endocrinology 122 1801–1809. &SummHD1994Differential interaction of glimepiride and Bretherton-Watt D, Gilbey SG, Ghatei MA, BeachamJ&BloomSR glibenclamide with the beta-cell sulphonylurea receptor. II 1989 Failure to establish islet amyloid polypeptide (amylin) as a Photoaffinity labeling of a 65 kDa protein by [3H]glimepiride. circulating beta cell inhibiting hormone in man. Diabetologia 33 Biochimica et Biophysica Acta 1191 278–290. 115–117. Carlei F, Allen JM, Bishop AE, Bloom SR & Polak JM 1985 Kulkarni RN, Smith DM, Ghatei MA & Bloom SR 1995a A67kDa Occurrence, distribution and nature of neuropeptide Y in the rat protein mediates the pituitary adenylate cyclase-activating pancreas. Experientia 41 1554–1557. polypeptide and vasoactive intestinal peptide-stimulated insulin Carpentier JL, Sawano F, RavazzolaM&Malaisse WJ 1986 secretion in a hamster clonal
-cell line. Journal of Endocrinology 147 Internalisation of 3H-glibenclamide in pancreatic islet cells. 121–130. Diabetologia 29 259–261. Kulkarni RN, Wang Z, Akinsanya KO, Bennet WM, Wang RM, Chomczysnki P & Sacchi N 1987 Single-step method of RNA Smith DM, Ghatei MA, Byfield PGH & Bloom SR 1995b isolation by acid guanidinium thiocyanate–phenol–chloroform Pyroglutamyl-phenylalanyl-proline amide attenuates thyrotropin- extraction. Analytical Biochemistry 162 156–159. releasing hormone stimulated insulin secretion in perifused rat islets DeFronzo RA & Simonson DC 1984 Oral sulfonylurea agents and insulin-secreting clonal beta cell lines. Endocrinology 136 suppress hepatic glucose production in non-insulin dependent 5155–5164. diabetic individuals. Diabetes Care 7 (Suppl 1) 72–80. Kulkarni RN, Wang ZL, Wang RM, Hurley JD, Smith DM, Ghatei Duckworth WC, Solomon SS & Kitabchi AE 1972 Effect of chronic MA, Withers DJ, Gardiner JV, Bailey CJ & Bloom SR 1997 Leptin sulfonylurea therapy on plasma insulin and proinsulin levels. Journal rapidly suppresses insulin release from insulinoma cells, rat and of Clinical Endocrinology and Metabolism 35 585–591. human islets and, in vivo, in mice. Journal of Clinical Investigation 100 Eliasson L, Renstrom E, Ammala C, Berggren P-O, Bertorello AM, 2729–2736. Bokvist K, Chibalin A, Deeney JT, Flatt PR, Gabel J, Gromada J, Lacy PE & Kostianovsky M 1967 Method for the isolation of intact Larsson O, Lindstrom P, Rhodes CJ & Rorsman P 1996 islets of Langerhans from the rat pancreas. Diabetes 16 35–39. PKC-dependent stimulation of exocytosis by sulfonylureas in Leclercq-Meyer V, Akkan AG, MarchandJ&Malaisse WJ 1991 pancreatic
cells. Science 271 813–815. Effects of glimepiride and glibenclamide on insulin and glucagon Fineberg SE & Schneider SH 1980 Glipizide versus tolbutamide, an secretion by the perfused rat pancreas. Biochemical Pharmacology 42 ff open trial. E ects on insulin secretory patterns and glucose 1634–1637. 18 concentrations. Diabetologia 49–54. Lebovitz HE & Feinglos MN 1983 The oral hypoglycaemic agents. In Gibson SJ, Polak JM, Allen JM, Adrian TE, Kelly JS & Bloom SR Diabetes Mellitus, Theory and Practice, edn 3, pp 591–610. Eds 1984 The distribution and origin of a novel brain peptide, M Ellenberg&HRifkin. New York: Medical Examination neuropeptide Y, in the spinal cord of several mammals. Journal of Publishing. Comparative Neurology 227 78–91. Gorus FK, Schuit FC, In’t Veld PA, Gepts W & Pipeleers DG 1988 Morgan DGA, Kulkarni RN, Hurley JD, Wang ZL, Wang RM, Interaction of sulphonylureas with pancreatic
-cells. A study with Karlsen AE, Bloom SR & Smith DM 1998 Inhibition of glucose glibenclamide. Diabetes 37 1090–1095. stimulated insulin secretion by neuropeptide Y is mediated via the Grill V, EfendicS&Regitz G 1986 Potentiation by previous nutrients Y1 receptor and inhibition of adenylyl cyclase in RIN 5AH rat of glibenclamide-induced insulin release in man. An effect which insulinoma cells. Diabetologia 41 1482–1491. is counteracted by meal-induced retardation of drug absorption. Myrsen U, AhrenB&Sundler F 1995 Neuropeptide Y is expressed Scandinavian Journal of Clinical and Laboratory Investigation 46 in subpopulations of insulin- and non-insulin-producing islet cells 527–532. in the rat after dexamethasone treatment: a combined immuno- Grodsky GM 1989 A new phase of insulin secretion: how will it cytochemical and in situ hybridisation study. Regulatory Peptides 60 contribute to our understanding of
-cell function? Diabetes 38 19–31. 673–678. Myrsen U, AhrenB&Sundler F 1996 Dexamethasone-induced Groop LC 1992 Sulfonylureas in NIDDM. Diabetes Care 15 755–772. neuropeptide Y expression in rat islet endocrine cells. Rapid Gullo D, Rabuazzo AM, Vetri M, Gatta C, Vinci C, Buscema M, reversibility and partial prevention by insulin. Diabetes 45 Vigneri R & Purrello F 1991 Chronic exposure to glibenclamide 1306–1316. www.endocrinology.org Journal of Endocrinology (2000) 165, 509–518

Downloaded from Bioscientifica.com at 09/24/2021 09:14:19AM via free access 518 R N KULKARNI and others · Effect of sulphonylureas on insulin and NPY release

Myrsen-Axcrona U, Karlsson S, SundlerF&Ahren B 1997 pancreatic islets following long-term administration of high or low Dexamethasone induces neuropeptide Y (NPY) expression and tolbutamide doses. Metabolism 26 9–15. impairs insulin release in the insulin-producing cell line RINm5F. Schatz H, NielreC&Pfeiffer EF 1975 (Pro-)insulin biosynthesis and Journal of Biological Chemistry 272 10790–10796. release of newly synthesized (pro-)insulin from isolated islets of rat Ogawa A, Johnsson JH, Ohneda M, McAllister CT, Inman L, Alam T pancreas in the presence of amino acids and sulphonylureas. & Unger RH 1992 Roles of insulin resistance and beta-cell European Journal of Clinical Investigation 5 477–485. dysfunction in dexamethasone-induced diabetes. Journal of Clinical Skoglund G, Gross RA, Bertrand GR, AhrenB&Loubatieres- Investigation 90 497–504. Mariani M-M 1991 Comparison of effects of neuropeptide Y and Ozanne SE, Guest PC, Hutton JC & Hales CN 1995 Intracellular norepinephrine on insulin secretion and vascular resistance in localization and molecular heterogeneity of the sulphonylurea perfused rat pancreas. Diabetes 40 660–665. receptor in insulin-secreting cells. Diabetologia 38 277–282. Teitelman G, Alpert S, Polak JM, MartinezA&HanahanD1993 Paice BJ, Patterson KR & Lawson DH 1985 Undesired effect of Precursor cells of mouse endocrine pancreas coexpress insulin, sulfonylurea drugs. Adverse Drug React Acute Poisoning Rev 4 glucagon and the neuronal proteins tyrosine hydroxylase and 23–36. neuropeptide Y, but not pancreatic polypeptide. Development 118 1031–1039. Petterson M, Ahren B, Lundquist I, BottcherG&Sundler F 1987 Waeber G, Thompson N, Waeber B, Brunner HR, Nicod P & Neuropeptide Y: intrapancreatic neuronal localization and effects on Grouzmann E 1993 Neuropeptide Y expression and regulation in a insulin secretion in the mouse. Cell Tissue Research 248 43–48. differentiated rat insulin-secreting cell line. Endocrinology 133 ff Purrello F, Veltri M, Gatta C, GulloD&VigneriR1989 E ects 1061–1067. of high glucose on insulin secretion by isolated rat islets and Wajchenberg BL, Nery M, Leme CE, Silveira AA, Fioratti P & purified
-cells and possible role of glycosylation. Diabetes 38 Germek OA 1980 Effect of prolonged gliclazide treatment on blood 1417–1422. glucose and plasma insulin responses in obese patients with maturity- Rabuazzo AM, Buscema M, Vinci C, Caltabiano V, Vetri M, Forte onset diabetes. Clinical Pharmacology and Therapeutics 27 375–378. F, Vigneri R & Purrello F 1992 Glibenclamide and tolbutamide Wang ZL, Bennet WM, Wang RM, Ghatei MA & Bloom SR induce desensitization of insulin release in rat pancreatic islets by 1994 Evidence of a paracrine role of neuropeptide-Y in the different mechanisms. Endocrinology 131 1815–1820. regulation of insulin release from pancreatic islets of normal and Sambrook J, Fritsch EF & Maniatis T 1989 Molecular Cloning: A dexamethasone-treated rats. Endocrinology 135 200–206. Laboratory Manual, edn 2. Plainview, New York: Cold Spring Harbor Laboratory Press. SchauderP,ArendsJ&Frerichs H 1977 Onset and reversibility of Received 25 August 1999 changes in secretory function and composition of isolated rat Accepted 13 January 2000

Journal of Endocrinology (2000) 165, 509–518 www.endocrinology.org

Downloaded from Bioscientifica.com at 09/24/2021 09:14:19AM via free access