Increased calcium-channel currents of pancreatic β cells in neonatally streptozocin-induced diabetic rats

Increased

Calcium-Channel Currents of Pancreatic p Cells in Neonatally Streptozocin-Induced Diabetic Rats

Seika Kato, Hitoshi Ishida, Yoshiyuki Tsuura, Yoshimasa Okamoto, Kazuo Tsuji, Minoru Horie, Yasunobu Okada, and Yutaka Seino Using a whole-cell patch-clamp technique, voltage-dependent Ca *+-channel activities were found to be increased in cultured single p cells isolated from neonatally streptozocin-induced diabetic rats (NSZ rats). The current-voltage relationship and inactivation time course of Ba2+ currents via L-type Ca*+ channels were indistinguishable between NSZ and control rats. However, the current density observed in NSZ rats was significantly greater than that in control rats. Ba*+ currents via T-type Ca*+ channels were also found to be enhanced in NSZ p cells. The insulin-secretory capacity of cultured pancreatic islets in response to a depolarizing stimulus (20 mmol/L arginine or 30 mmol/L KCI) in the presence of 11.1 mmol/L glucose was augmented in NSZ rats, whereas that in response to 11.1 and 16.7 mmol/L glucose alone was significantly reduced. It is concluded that the impaired insulinotropic action of glucose in p cells in NSZ rats is not due to reduced activity of voltage-dependent Ca*+ channels. The fact that insulin secretion induced by a depolarizing stimulus was enhanced in NSZ rats may be related to the augmented activity of the voltage-dependent calcium current found in NSZ p cells. Copyright (c 1994 by W.B. Saunders Company

A

CHARACTERISTIC FEATURE of patients with non-insulin-dependent diabetes mellitus is a selective impairment of glucose-induced insulin secretion. However. administration of arginine is known to elicit insulin release from pancreatic p cells in hyperglycemic animal models to an extent similar to or at times larger than in normal animals.’ In a hyperglycemic animal model obtained by injection of streptozocin to neonatal rats (NSZ rats), insulin secretion has been similarly observed to be induced by arginine, but not by glucose.?-” It is generally accepted that an increase of the intracellular free calcium concentration ([Ca?+]J, triggered by activation of voltage-dependent CaL+ channels. plays a crucial role in the mechanism of insulin secretion. We have recently observed that the [Caz+], response to glucose is selectively impaired in pancreatic p cells of NSZ rats.5 This fact may not necessarily imply an inhibition of voltage-dependent Ca2+ channels, because the glucose sensitivity of adenosine triphosphate (ATP)sensitive potassium channels (KATP channels), whose inhibition by glucose brings about a depolarization resulting in activation of voltage-dependent Ca2+ channels, has recently been found to be impaired in p cells of NSZ rats.6 This decreased glucose sensitivity of KATP channels was attributed to the impairment of glucose metabolism.h In fact, it has been shown that the [Ca’+]i response to arginine was not impaired but instead enhanced in p cells of NSZ rats.5 Taken together, it is possible that the activity of voltagedependent Ca’+ channels is not impaired but rather enhanced in NSZ p cells. However, to date, no direct recording of Cal+-channel activity has been made in diabetic p cells. In the present study, we recorded voltagedependent Cal+-channel currents in p cells of NSZ rats using a whole-cell patch-clamp technique. Furthermore, the number of L-type Ca’+ channels was estimated from the total specific dihydropyridine-binding sites assessed by [3H]-nitrendipine-binding study. MATERIALS

AND METHODS

Animals NSZ rats were prepared by subcutaneous injection of 50 mgikg (Sigma, St Louis. MO) to male Wistar neonatal rats

streptozocin

Metabolism,

Vol 43, No 11 (November), 1994: pp 1395-1400

(aged 1.5 days)’ and used in experiments between 9 and 12 weeks of age. Age-matched untreated male Wistar rats were used as the control. The body weight of NSZ rats (263 ? I1 g, n = 8) was not statistically different from that of control rats (259 2 10 g, n = 8). Blood samples were collected via the left renal vein immediately before isolation of the pancreatic islets under pentobarbital anesthesia (40 mg/kg). The plasma glucose concentration in the fed state as measured by the glucose oxidase method was 9.4 ? 0.2 mmol/L (n = 8) in control rats and 18.1 2 1.1 mmol/L (n = 8. P < ,005) in NSZ rats. Assessment of the Insulin-Secretory CapaciQ of Cultured Islets Islets of Langerhans were isolated by collagenase digestion from animals of both groups as previously described,’ and were cultured in RPM1 1640 medium containing 11.1 mmol/L glucose for 24 hours. Insulin secretion induced by glucose alone (5.5. 11.1, and 16.7 mmol/L) or by a depolarizing stimulus (20 mmol/L arginine or 30 mmol/L KCI) in the presence of 11.1 mmol/L glucose was examined in cultured islets by the batch incubation method. as previously reported.’

Measurements of Ba2+ Currents in Cultured p Cells The culture of dispersed islet cells was performed on small cover glasses according to a previously described method.” Whole-cell current recording@ were made from p cells at 25” to 27°C with a patch-clamp amplifier (EPC-7, List Electronics. Darmstadt. Ger-

From the Deparrment ofMetabolismand Clintcal Nutritionand the Third Department of Internal Medicine, Kvoto University School of Medicine. Kvoto; and the Department of Cellular and Molecular Physiology. National Institute for Physiological Sciences. Okazaki. Japan. Submitted September 20. 1993; accepted Janus? 19, 1994. Supported by Grants-in-Aid for Scientific Research from the Ministn, of Education. Science. and Culture and the Research Committee of Experimental Mod& for intractuble Diseases of the Minisfn, of Health and Welfare of Japan, and also by grants ,from the Japan Diabetes Foundation and Otsrtka Pharmaceutical. Tokyo, Japan. Address reprint requests to Seika Kate. MD, Department of Metabolism and Clinical Nutrition, Qoto Universiy School of Medicine. 54 Shogoin Kawahara-cho, Sakyo-ku, Kyoto 606. Japan. Copyright 0 I994 by W.B. Saunders Compuny 0026-0495l94/3311-0012$03.0010 1395

KATO ET AL

many). Patch pipettes (resistance, from

borosilicate

Corning,

Midland,

2 to 4 megaohms)

glass capillaries, Ml).

coated

fire-polished.

solution containing (in mmol/L)

with

were pulled

Sylgard,

Capacity of Cultured Islets From NSZ and

Control Rats (n = 5)

and filled with a pipette

Insulin-Secretory

130 CsCI, 10 tetraethylammonium

4 ATP-Na2, 0.4 GTP-Na?, IO EGTA, and 10 HEPES (pH 7.2 with CsOH). Ba?+ currents were recorded in a bath solution composed of (in mmol/L) 30 BaCIL. 116 CsCl, 10 tetraethylammonium chloride. 1.2 MgCll, and 10 HEPES (pH 7.4 with CsOH). Because insulin-producing cells have been characterized by their larger size, y.ln the experiments were performed using individual cells greater than 10 km in diameter. The total exposure time of these cells to the glucose-free bathing solution did not exceed 10 minutes. Ba?’ currents via L-type Ca?+ channels were measured by applying depolarizing step pulses (from -40 mV to - +50 mV, in IO-mV steps, l-second duration at 0.1 Hz) from a holding potential of -45 mV. After recording L-type Ca’+-channel currents. Ba?+ currents via T-type Ca2+ channels were assessed by measuring nitrendipine-resistant Ba?+ currents upon depolarizations to -20 mV (OS-seconds duration) from a holding potential of -90 mV in the presence of 10 bmol/L nitrendipine. since deep negative holding potential can release the T-type Ca’+ channels from inactivation.‘” Data currents were stored on video tape via a pulse code modulator (PCM 501. Sony, Tokyo, Japan) and later analyzed by personal computer using software routines provided by Dr S. Morishima (Kyoto Universiiy). Leak currents were estimated by applying -65mV pulses before each experiment, and only cells that showed minimal leak currents (< 10 PA) were used for data acquisition. Inward currents were converted to current densities after dividing by the cell membrane capacitance. which was evaluated by the C-slow value of the EPC-7. Under these conditions, Ba” currents showed little rundown up to 3 minutes. chloride,

Table 1. Insulin-Secretory

(Dow

Capmty

NSZ

Control

5.5

1.76 + 0.37*

4.56 + 0.52

11.1

4.69 + 0.60*

6.46 2 0.50

16.7

8.37 + 0.91 t

20.02 + 2.64

52.28 r 8.90

40.71 k 4.27

5 MgCl:.

Nitrendipine Binding in Pancreatic Met Cells The number of specific nitrendipine-binding sites in pancreatic was assessed according to a method previously described.” The dispersed islet cells were incubated for 2 hours at 23°C in a buffer containing 25 mmol/L HEPES and 1 mmol/L CaCIz, pH 7.4. with 0.5 nmol/L [3H]-nitrendipine (73.0 Ciimmol, NET 741, New England Nuclear. Boston. MA) in the absence or presence of 1 PmoliL nonradioactive nitrendipine. The reaction was terminated by adding 1 mL of a terminating buffer containing 25 mmol/L Tris hydrochloride and 1 mmol/L CaClz, pH 7.4. The mixture was rapidly filtered through Whatman GFiF filters, followed by washing three times with 5 mL of the terminating buffer. Radioactivity of the filters was counted in 10 mL of an aqueous scintillation cocktail (Aquasol-2, New England Nuclear). The specific binding of [3H]-nitrendipine was calculated by subtracting nonspecific binding ([3H]-nitrendipine bound in the presence of I kmol/L nonradioactive nitrendipine) from total [3H]nitrendipine binding. Protein content was measured by the method of Lowry et aI.” Results were expressed as the mean i SE. Statistical significance was evaluated by an unpaired Student’s I test or the chi-square test. islet cells

RESULTS

Insulin-Secretory Capacity of Cultured

Islets of NSZ Rats

As shown in Table 1, insulin secretion induced by 5.5, 11.1, and 16.7 mmol/L glucose was significantly lower in islets of NSZ rats than in controls (P < .05). In the presence of 11.1 mmol/L glucose, the insulin-secretory capacity in response to 20 mmol/L arginine and 30 mmol/L

Glucose, mmol/L

Arginine, 20 mmol/LS KCI, 30 mmol/LZ

(1,115 + 190)X

(630 k 80)

27.67 2 4.27

22.23 Y! 3.04

(590 + 91)”

(344 t 47)

NOTE. Values in parentheses are the percentage expression of the insulin secretion induced by 11.1 mmol/L glucose alone. *Statistically different from the control at P < .05. tstatistically different from the control at P < ,001, Sin the presence of 11 .l mmol/L glucose.

KC1 in NSZ islets was greater than in control islets. When these values were expressed as the percentage of the insulin secretion induced by 11.1 mmol/Lglucosc alone. NSZ islets exhibited significantly augmented responses to these stimuli (P < .M; Table 1). Whole-Cell Ba2+ Currents in p Cells qf NSZ Rats Sustained inward currents were recorded upon application of l-second depolarizing pulses to more than -30 mV from a holding potential of -45 mV in single p cells not only of control but also of NSZ rats. as shown in Fig 1. These currents were maximal at approximately 0 mV. and were decreased with a shifting of the command pulse level to more positive potentials. Although the membrane capacitance of NSZ p cells (7.2 2 0.3 pF, n = 32) was not significantly different from that of control cells (6.5 t 0.4 pF. n = 25), the voltage-dependent inward currents recorded from NSZ 0 cells (Fig 1B) were markedly greater than those from control cells (Fig 1A). When the current densities as calculated by dividing the current by the membrane capacitance were plotted against voltage, essentially similar bell-shaped current-voltage curves (I-V curves) were obtained in both control and NSZ cells (Fig 2). Current densities in NSZ rats were significantly larger than in control rats at -20 to +30 mV (P < .05). Slow time-dependent inactivation was observed after voltage steps positive to -10 mV. The inactivation time course was well fitted by a single exponential. and the time constant (T) was not significantly different between the two groups (432 2 9 milliseconds in 32 NSZ cells and 447 t 12 milliseconds in 25 control cells at 0 mV). The currents in cells from both control and NSZ rats were almost completely inhibited by 10 pmol/L nitrendipine (Fig 3A and B). The activation of Ba’+ currents upon depolarizations positive to -30 mV from a relatively shallow holding potential (-45 mV), their slow timedependent inactivation during depolarizations, and the inhibition by nitrendipine are characteristic of currents through L-type Ca:+ channels.l”.lJ Nitrendipine-resistant Ba7+ currents could be transiently elicited by depolarizations to -20 mV from a holding

Ca2+ CHANNELS

IN HYPERGLYCEMIC

(3 CELLS

1397

A

control

B

NSZ

-20 mV

Fig 1. Representative BaZ+ currents activated by depolarizing pulses from a holding potential of -45 mV in p cells in the (A) control rat and (B) NSZ rat. Whole-cell current recording was performed using a Cs+-rich pipette solution and a bath solution containing 30 mmol/ L BaZ+. Cell membrane capacitance was 5.6 pF for the control cell and 5.4 pF for the diabetic cell.

+20

mv

Cl0 -10

mV mV 200

0 mV

potential of -90 mV in the presence of 10 pmol/L nitrendipinc (Fig 4, upper traces). The rapidly inactivating Ba2+ currents were found to be significantly greater in NSZ p cells than in control /3 cells (10.3 ? 0.9 pA/pF in 32 NSZ I3cells and 4.32 ? 0.6 pA/pF in 17 control p cells, P < ,001). The inactivation time constant (T) of 18.7 + 1.4 milliseconds in NSZ cells (n = 10) was not statistically different from the value of 18.4 t 2.0 milliseconds in control cells (n = 10). When these cells were kept at -45 mV. the transient Ba’+ current was never induced by a depolarization step in the presence of 10 pmol/L nitrendipine (Fig 4, lower traces).

Nitrendipine-Binding

Sites in NSZ Islets

The number of specific nitrendipine-binding sites was significantly lower in NSZ islets (4.26 * 0.82 fmolimg protein. n = 5. P < .05) than in control islets (7.54 + 0.82, n = 5). DISCUSSION

In the present study, we have directly demonstrated that the L-type voltage-dependent Ca’+ (Ba’+) current was not impaired, but rather it was significantly augmented in I3 cells of NSZ rats. Therefore it appears that an impairment of glucose-induced insulin secretion in NSZ rats is not associated with reduced activity of the Ca?+ channel, but rather with insufficient depolarizations in response to

I

pA

_

0.1 s

glucose due to decreased glucose sensitivity of KATP channels secondary to impaired glucose metabolism.h The present finding of increased activity of L-type Ca?+ channels in NSZ I3 cells may account for our recent observation that the [Ca’+], response to arginine in I3 cells of NSZ rats is greater than in control rats.5 The increased [Ca?+], response to the depolarizing agent may be causatively related to the insulin secretion of supranormal capacity in response to arginine and high K+ in islets of NSZ rats (Table 1). L-type Ca’+-channel currents exhibited identical inactivation kinetics and nitrendipine sensitivity in both control and NSZ 13 cells. The voltage dependency of the L-type Ca?+ channel was also indistinguishable between both groups, as judged by the I-V curves (Fig 2). Therefore, it is likely that the channel properties per se are not grossly altered in NSZ I3 cells. On the other hand, the number of specific binding sites for nitrendipine, which is considered to be related to the number of L-type Ca?+ channels, was decreased in NSZ islets. Therefore, the increased L-type Cal+ currents seem not to be due to the increase in the number of channels. but rather increases in the single-channel conductance or open probability may be responsible for the increased current density under hyperglycemic conditions. It is possible that an altered level of phosphorylation or dephosphorylation of the channel protein is associated with increased singlechannel conductance or open probability, since a number of

1398

KATO ET AL

PA/M 20

40

putative phosphorylation sites have been shown to exist by the molecular cloning of the p-cell L-type Ca7+ channel.” For the precise mechanism to be determined. singlechannel recordings of L-type Ca’+ channels arc necessary. In rat pancreatic p cells, T-type and L-type Ca’+ channels have been reported to be present.“’ T-type Ca’+channel currents are known to show more rapid timedependent inactivation, to become apparent when the channel is activated by a depolarization (-40 mV to + -20 mV) from a deep negative holding potential (-100 mV - -80 mV). and to be insensitive to dihydropyridine.“’ Therefore, nitrendipine-resistant, rapidly inactivated Ba2+ currents observedin the present study arc considered T-type Ca’+-channel currents. These currents were also found to be enhanced in p cells of NSZ rats. However, an exact role of T-type Ca?+ channels in the excitationsecretion coupling of 0 cells is still unknown. In normal p cells, the L-type Ca’+-channel activity recorded in the perforated patch or cell-attached patch configuration is known to be enhanced by a high conccntration of glucose in the bathing solution.lh In the present study, the enhanced activity of L-type Ca’+ channels in p cells of NSZ rats was observed in the absence of ambient glucose under the conventional whole-cell configuration. There is a possibility that the increased Ca?+-channel activity had been induced by some metabolic alterations in diabetic p cells exposed to prolonged hyperglycemia in vivo before culturing. Antibodies in the sera from insulin-dependent diabetic patients have been found to be capable of activating L-type Cal+ channels of p cells, thereby triggering the pro-

mV

)I

\

-60

t

Fig 2. Current-voltage relationship of the voltage-dependent Ba*+ currents recorded from control (0) and NSZ (W) p cells kept at -45 mV. In both curves, no significant difference was observed in the distribution of voltages at which maximal currents were obtained (2 test, P = .21). Each symbol represents the mean of 25 to 32 experiments. lP < .05. tP c ,001.

A

control

j-

__z;;;_

10 JIM Nitrendipine ,,_icni&“.&,-“-.++-J-“~--’ m“ 0 UM

B

NSZ

) 200

PA

1 0.1

s

Fig 3. Nitrendipine sensitivity of Ba2+ currents elicited by a voltage step to 0 mV from the holding potential of -45 mV. Representative inward currents recorded from a B cell (5.0 pF) of the (A) control rat and that (5.0 pF) of the (B) NSZ rat are shown. Inward currents recorded at other voltage steps were also inhibited by nitrendipine in both groups in an irreversible manner.

Ca2+ CHANNELS IN HYPERGLYCEMIC

A

p CELLS

1399

control

B

US2

ti.p.=

H.P.=

L

20 PA

-45

mV

-90

H.P.’

mv

-45

mv

L7.0 PA

I

50 ma

50 ma

Fig 4. Representative Ba2+ currents in response to depolarizing pulses to -20 mVfrom two different holding potentials ([HP] = -90 mVforthe upper traces and -45 mV for the lower traces) in (A) a control p cell and (B) an NSZ )3 cell of the same size (5.0 pF) in the presence of 10 pmol/L nitrendipine.

grammed death of the p cells. I7 Therefore, it is also possible that chronic Ca2+ overload due to increased Ca?+-channel activit:y is involved in the gradual loss of intact p cells in non-insulin-dependent diabetes mellitus. In conclusion, L-type: voltage-dependent calcium currents are found to be increased in p cells of NSZ rats. These facts may well explain the fact that insulin release evoked by nonglucose

depolarizing stimuli non-insulin-dependent

is preserved or often diabetes mellitus.’

augmented

in

ACKNOWLEDGMENT The authors thank Dr A.F. James for comments on the manuscript, T. Tanimura for his technical assistance, and H. Tachikawa and K. Iwasa for their secretarial work.

REFERENCES 1. Leahy JL. Bonner-Weir S, Weir GC: p-cell dysfunction induced by chronic hyperglycemia. Diabetes Care 15:442-455, 1992

streptozotocin in the rat during 23:889-895. 1974

2. Weir GC. Clore ET, Zmachinski CJ. et al: Islet secretion in a new experimental model for non-insulin-dependent diabetes. Diabetes 31):590-595, 1981

8. Hamill OP. Marty A, Neher E. et al: Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers Arch 391:X5-100, 1981

3. T!suji K. Taminato T, Usami M, et al: Characteristic insulin secretion in the streptozotocin-induced NIDDM Metabolism 37:1040-1044. 1988

feature of rat model.

9. Pipeleers DG, In’t Veld PA. Van de Winkel, et al: A new in vitro model for the study of pancreatic A and B cells. Endocrinology 117:806-816. 1985

4. Giroix MH. Portha B, Kergoat M, et al: Glucose insensitivity and amino-acid hypersensitivity of insulin release in rats with non-insulin-dependent diabetes. Diabetes 32:445-45 1. 1983

10. Hiraiart M, Matteson DR: Na’ channels and two types of CaZ+ channels in rat pancreatic B cells identified with the reverse hemolytic plaque assay. J Gen Physiol91:617-639, 1988

5. Tsuji K. Taminato T, Ishida H, et al: Selective impairment of the cytoplasmic Ca?+ response to glucose in rat pancreatic B cells of streptozocin-induced non-insulin-dependent diabetes. Metabolism 42: 1424-1428. 1993

11. Hoenig H. Russo LL. Ferguson DC: Characterization of calcium channels of a glucose-responsive rat insulinoma. Am J Physiol256:E488-E493.1989

6. Tsuura Y, Ishida H, Okamoto Y. et al: Impaired glucose sensitivity of ATP-sensitive K+ channels in pancreatic p-cells in streptozotocin induced NIDDM rats. Diabetes 41:861-865, 1992 7. Portha

B. Levacher

C. Picon L, et al: Diabetogenic

effect of

the perinatal

period.

Diabetes

12. Lowry OH. Rosebrough NJ, Farr AL, et al: Protein measurement with the Folin phenol reagent. J Biol Chem 193:265-275. 1951 13. Hopkins WF, Satin LS, Cook L: Inactivation kinetics and pharmacology distinguish two calcium currents in mouse pancreatic B-cells. J Membr Biol 119:229-239. 1991

1400

14. T&n RW. Ellinor PT, Home WA: Molecular diversity of voltage-dependent Ca?+ channels. Trends Pharmacol Sci 12:349354,199l 15. Seino S, Chen L. Seino M. et al: Cloning of the cul subunit of a voltage-dependent calcium channel expressed in pancreatic B cells. Proc Natl Acad Sci USA 89:584-588. 1992

KATO ET AL

16. Smith PA, Rorsman P. Ashcroft FM: Modulation of dihyby glucose metabolism in dropyridine-sensitive Ca?’ channels mouse pancreatic p-cells. Nature 342:5X)-553. 1989 17. Berggren LJ. Larsson 0. Rorsman P. et al: Increased activity of L-type Ca’- channels exposed to serum from patients with type I diabetes. Science 261:86-90, 1993