Pituitary Adenylate Cyclase Activating Polypeptide Is an ...

THEJOURNAL OF BIOUXCALCHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 269, No. 2, Issue of January 14, pp. 1290-1293, 1994 Printed in U.S.A.

Pituitary Adenylate Cyclase Activating Polypeptide Is an Extraordinarily Potent Intra-pancreatic Regulator of Insulin Secretion from Islet P-Cells* (Received for publication, July 15, 1993, and in revised form, August 11, 1993)

Toshihiko YadaSS, Masaya Sakuradan, Kaori Ihidall, Masanori NakataS, Fusayoshi Muratall, Akira Arimura**, and Masatoshi Kikuchin From the Departments of $Physiology and Iknatomy, Kagoshima University School of Medicine, Kagoshima 890,Japan, the %Institutefor Adult Diseases, Asahi Life Foundation, lbkyo 160, Japan, and **United States-Japan Biomedical Research Laboratories, W a n e University Hebert Center, Belle Chasse, Louisiana 70037

Insulin secretion from pancreatic islets is controlled by peptides as well as by nutrients. We report here a novel, extraordinarily potent peptidergic regulation of insulin secretion. A 27-residue form of pituitary adenylate cyclase activating polypeptide (PACAP27)as low as to M stimulated insulin release from rat islets in a glucose-dependentmanner. PACAP27 also increased cytosolic free Ca2+concentration ([Ca2+Ii) in islet p-cells. Nitrendipine, a blocker of the L-type Ca2+ channel, abolished both [Ca2+Ijand insulin responses. Vasoactive intestinal peptide, a peptide exhibiting 68% amino acid homology with PACAP, also increased [Ca2+Ii in p-cells but only at concentrations in the nanomolar range, indicating that PACAP27 is 4 logs more potent.A 38-residue form of the peptide (PACAP38) stimulated insulin release and increased @-cell [Ca2+Ii in a manner similar to that of PACAP27. PACAP-likeimmunoreactivity was demonstrated in pancreatic nerve fibers,islets, and capillaries. The results indicate that PACAP is a physiologically occurring peptide in pancreas and that PACAP, in a glucose-dependent manner, activates p-cells presumably via a high affinity PACAP-selective receptor, raises [Ca2+liby increasing the activity of L-type Ca2+ channels,and consequently stimulates insulin release. PACAP appears to be byfar the most potent insulinotropic peptide known.

(PACAP27) or a 38-residue peptide (PACAP38) (6, 8) and is present in a variety of tissue, abundant in the central nervous system, pituitary, adrenal, and testis (9). PACAP-containing nerve fibers were demonstrated in brain and other tissue (7,

04 30

40

50

60

Time(min)

B

T#

600 Insulin secretion from pancreatic islets are under regulation by peptides,as well as by nutrientsand other substances (1,2). Some of the members of the secretinlglucagonMP1 family of peptides such as glucagon, which originates from islet a-cells, and truncated glucagon-like peptide-1 (tGLP-l), which originates from intestine, stimulate insulin release in a low concentrationrange of to M andarethoughtto be involved i n t h e physiological regulation of insulin release (3-5). Pituitary adenylate cyclase activating polypeptide (PACAP), originally isolated from ovine hypothalamus, is a new member of this family (6, 7).PACAP occurs as a 27-residuepeptide

* This work was supported in part by grants from the Ministry of Education, Science and Culture of Japan and from the Kodama Foundation for Research of Medical Science(to T. Y.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. Section 1734 solely toindicate this fact. Q To whom correspondence shouldbe addressed: Dept. of Physiology, Kagoshima University School of Medicine, 8-35-1 Sakuragaoka, Kagoshima 890, Japan. Tel.: 81-992-75-5225;Fax: 81-992-75-5231. The abbreviations used are: VIP, vasoactive intestinal peptide; [Ca2+Ii,cytosolic free Ca2+concentration; PACAP, pituitary adenylate cyclase activating polypeptide; tGLP-1, truncated glucagon-like peptide-1; KRB, Krebs-Ringer bicarbonate; BSA, bovine serum albumin.

(-)

-15

-14

-13

-12

Log PACAP27 (M) FIG.1. Stimulation of insulinreleasefromratislets by PACAP27. A, insulin release from perifused islets in KRB containing The thick line signifies the period of 8.3 (0)or 2.8 m~ glucose (0). exposure to 10-13 M PACAP27.B , concentration response of PACAP27 to stimulate insulin release from islets in static incubation for 60 min in KRB containing 8.3 (filled column) or 2.8 mM glucose (open column). Eachpoint represents the mean * S.E. for six experiments. *, p c 0.05; #, p c 0.01 versus base-line level preceding PACAP stimulation (A) or control without PACAP ( B ) .

1290

PACAP: Potent Intra-pancreatic Regulator of Insulin Release

1291

[Caz+li Measurements-[Ca2+li in single islet cells was measured as described previously(20).In brief, isolated rat islets were dispersed into single cells under Ca-free conditions. The single cells were plated on coverslips and cultured for 1 4 days in Eagle's minimal essential medium with 5.6 m~ glucose and 10%fetal bovine serum. The cells were incubated with 11.1~fura-Yacetoxymethylester(21)for 30min at 37 "C. The cells were superfused with KFtB at 1d m n i a t 37 "C, and fura-2 fluorescence at 510 nm following excitation a t 340 and 380 nm every 2.5 s was detected by intensified charge-coupled device (ICCD) camera, and the ratio image was produced using an Argus-50 system (Hamamatsu Photonics) (13).Alternately, fura-2 fluorescence was detected every '1s by photomultiplier using a P1 system (Nikon). Ratio was converted to [Ca2+li according to calibration curves. We took the datafrom the cells that responded to glucose and tolbutamide (20). Identification of Islet &Cells"ARer the [Ca2+limeasurement, the cells on coverslips were fixed in 10% formalin. Procedures for immunostaining were carried out a t room temperature. Endogenous peroxidase activity was inhibited by 0.3% H202.ARer pretreatment with 1% BSA, cells were incubated with guinea pig anti-insulin antibody (Dako Corp., Carpinteria, CA) for 3 h. Biotinylated goat anti-guinea pig IgG (1:lOO) (Vector Laboratories, Burlingame, CA) was applied for 60 min, EXPERIMENTAL PROCEDURES followed by incubation with horseradish peroxidase-labeled streptaviSolutions and Chemicals-Insulin release experiments were carried din (1:lOO) for 60min. Location of the antigen was visualized by diamiout in KRB composed of (m~):121.7 NaCl, 4.4 KCl, 1.2 KH2P04, 2.0 nobenzidhe reaction (22). Immunohistochemistry with PACAP27 Antibody-Pancreases of CaCl,, 1.2 MgSO,, 23.0 NaHC03 at pH 7.4 supplemented with 0.1% bovine serum albumin (BSA)and equilibrated against a mixture of 95% Wistar rats were removed, fixed in 10% formalin, and embedded in 02,5%C02 at 37 "C. KRB used for single cell experiments contained 10 par&. Following peroxidase inhibition and pretreatment with BSA, m~ Hepes and a reducedHCO; (5.0 m ~ )PACAP27 . and PACAP38 were sections of 4 p were incubated in a rabbit antisera against synthetic from American Peptide Co. (Sunnyvale, CA) and Peptide Institute PACAP27 (88121-6) (10) a t 1:2000 dilution for 72 h at 4°C. Subse(Osaka, Japan). VIP was from Peptide Institute. All the peptides were quently, goat anti-rabbit IgG was applied for 60 min followed by incudissolved in distilled water with 0.1% BSAin a concentration of lo4 M bation with protein A-CG conjugate (23) for 60 min. Final reaction products were intensified by photochemical silver reaction (24) using and stored a t -20 "C until used. Insulin Release"Is1ets of Langerhans were isolated from 8-12-week- silver acetate as an ion donor, followed by fixation with Ilford photoold Wktar rats by collagenase digestion.Groups of 100 islets were first graphic fixer Word Proprietary Ltd., Mt.Waverly, Australia). Sections incubated for 1h in KRB with 2.8 m~ glucose. For perifusion studies, were counterstained with nuclear fast red. Control experiments were islets were then placed in a chamber and perifused a t 1d m i n . m e30r carried out by omitting the primary antisera or using nonimmune rabmin,the perifusate was collected everyminute and insulin determined bit antiserafollowed bya staining of the section by solid-phase absorpby radioimmunoassay. For static incubation studies, islets were incu- tion method (25). bated for 60 minand aliquots assayed for insulin. The calculated values were expressed as mean 2 S.E. for the number of observations (n). RESULTS AND DISCUSSION Student's t test was used for statistical analysis of the results. StimuIn the presence of 8.3 IILM glucose, PACAP27 at to lated insulin release a t each time point was comparedwith the mean of M stimulated insulin release in a biphasic manner: an initial the base-line level preceding stimulation. For static incubation data, the values of insulin release for 60min with and without PACAP were sharp rise followed by a moderate elevation (Fig. lA).Insulin release from islets under static incubation was also increased compared.

10). ! h o binding sites, type I receptors specific forPACAP and type I1 receptors shared with VIP, were demonstrated (11). PACAP exhibits a variety of biological activities, which include activation of the adenylate cyclase in pituitary and other cells (6, 7), release of pituitary hormones from perifused pituitary cells (7, 121, [Ca2+Iiincrease in hippocampal neurons (13) and pituitary cells(14, 151, catecholamine release from adrenal chromaflin cells (16),amylase release from exocrine pancreas (17, 18), and vasodilating effect (7, 19). However, the role of PACAP in theendocrine pancreas is poorly understood. The present study was aimed at exploring a possible role of PACAP in the regulation of insulin secretion. Here we report that PACAP as low as to M activates rat islet p-cells via a specific PACAP receptor, increases [Ca2+Ii,and consequently stimulates insulin release in a glucose-dependentmanner. "he PACAP-like immunoreactivity is demonstrated in the nerve fibers and islets of rat pancreas.

FIG. 2. [Ca2+Iiresponse to PACAP27 in insulin-ptmitive islet &cells. A and B , pseudo-color image of [Ca2+liin islet cells before (A) and during exposure to 10-13 M PACAP27 ( B ) in the presence of 8.3 m~ glucose. C, subsequent immunwytochemical staining of the same cells with antibody against insulin. The majority of the positively stained cells, including those numbered 1 , 2 , and 3 (cell 1, cell 2 , and cell 3). had responded to PACAP27 with [Ca2+Iiincreases. Insulin-negative cells with relatively smaller diameter were also seen. Bur indicates 50 p.D, [Ca2+liresponses to 8.3 m~ glucose, 10-13 M PACAP27, and lo4 m tolbutamide in insulin-positive cell 1, cell 2, and cell 3. Glucose concentration is indicated in the upper column. Thick lines s p e w the period of exposure to agents, andarrows indicate the beginning of PACAP administration. Dots with TA and TB signify the time when the images in A and B were taken, respectively.

.

.1

* 1

.2

PACAP: Potent Intra-pancreatic Regulator of Insulin Release

1292 A

Glucose 8.3 rnM

G2.81 PAC1613

I

B

A

G2.81 Glucose 8.3 rnM

G2.d

Glucose 8.3 mM

TB34

6001

PACAP27 1(r'3

6001

a

3

-;s

0

0

O J d -;5

-i4

-;3

-;2 -;1 Log PACAP27 or VIP (M)

I

I

-io

-b

-b

FIG.3. Characteristics of [Caa+li responses to PACAP27 and VIP. A, glucosedependence of p-cell [Ca2+liresponse to 10-13 M PACAP27. Panel is representative of four experiments ( n = 4). B , p-cell [Ca2+],response to lo-* M VIP ( n = 6). Slanting lines indicate an omission of recording for 7 min. C , concentration response relationships for the frequency (%) of p-cells with [Ca2+l,response to PACAP27 in the presence of 8.3 (W) or 2.8 m~ glucose (0) and to VIP in 8.3 m~ glucose (A). Each point represents 5-20 cells.

L

30

40

50

60

80

70

90

100

Time (min)

byPACAP27 at to 10-l' M (Fig. lB). Ina lowerconcentration of glucose (2.8 mM),PACAP27 failed to alter insulin release (Fig. 1,A and B 1. In previous reports, PACAP27 at to M stimulated insulin release from perfused rat pancreas (26). In our experiments, PACAP27 at lo-' M also increased M. insulin release but to a lesser extent than at Since cytosolicCa2+in islet p-cells plays a crucial role in the regulation of insulin secretion (1,2,27), we examined whether PACAP27 couldact directly on islet p-cells and increase [Ca2+Ii. PACAP27 at M,administered under superfusion condition, increased [Ca2+liin a large population of islet cells, as demonstrated by a change in thepseudo-color image (Fig. 2, B versus A). Following [Ca2+Iimeasurement, cells were immediately fwed and stained using an antibody to insulin. The majority of the islet cells that exhibited [Ca2+li response to PACAP27 were stained positively with anti-insulin antibody (Fig. 2, A X ) , indicating that the p-cell is the direct target of PACAP27. In this experiment, 12 of 15 insulin-positive cells (80%) responded to M PACAP27. PACAP27 increased [Ca2+Iiin a rather transient and monophasic pattern (Fig. 2 0 ) . The data also confirm the previous reports that glucose produces a bimodal change in [Ca2+]i:an initial decrease and subsequent increase (28, 29), now in immunocytochemically identified p-cells (Fig. 2 0 ) . Inthe presence of 2.8 m~ glucose, PACAP27 at M was without effect on [Ca2+li(Fig. 3 A ) . PACAP27 increased [Ca2+li in a concentration-dependent manner, and a maximal effect was obtained at M and a t to lo-@M,where about 80% of p-cells responded(Fig. 3C). In theCa2+-freeKRB made with 0.1 mM EGTA and no added Ca2+, neither the increase in [Ca2+li M PACAP27. nor the release of insulin was induced by Nitrendipine (1p ~ ) a, blocker of the L-type Ca2+channel, inM PACAP27, hibited both[Ca2+Ii and insulin responses to and after washing out the drug both responses were restored (Fig. 4, A and B ) . The data suggest that PACAP27 stimulates Ca2+influx by enhancing the activity of the dihydropyridinesensitive L-type Ca2+channel, which is present in p-cells and activated by glucose (30, 31). The [Ca2+Iiincrease and insulin release in response to PACAP27 shared common properties: a similar concentration-

FIG.4. Complete and reversible inhibitionof [Ca2+lrand insulin responses by nitrendipine (NTD). A, &cell [Ca2+liresponse to 10-13 M PACAP27, whichwas muchlarger inthe amplitude than the 8.3 m~ glucose-induced oscillation, was inhibited by 1 p~ NTD.Slanting lines indicate an omission of recording for 11min ( n = 6 ) .B , stimulated release of insulin from perifused islets by 10-13 M PACAP27 was inhibited by 1 p~ nitrendipine. Panel shows the mean * S.E. for fiveexperiments. *, p < 0.05;#, p < 0.01 versus base-line level preceding PACAP stimulation.

B

G2.8IGlucose 8.3rnM 400-1

PACAW:

1013

:@

Q- 300

+=200

-e

b- 100 30

40

50 60 time (min)

70

FIG.5.Insulin and [Caz+li responses to PACAP38.A, PACAP38 at 10-13 M stimulated insulin release from perifused rat islets in amanner . similar t o PACAP27 at 10-13. Glucoseconcentration was 8.3 m ~Panel gives the mean * S.E. for six experiments. * , p < 0.05;# , p < 0.01 versus base-line levelpreceding PACAP stimulation. B , PACAP38 a t M increased [Ca2+liin p-cells (n = 6).

response relationship, glucose dependence(Figs. 1and 3), and nitrendipine inhibition (Fig. 4). The results indicate that PACAP-induced insulin release is mediated by increased [Ca2+Ii.This is in accord with the notion that an increase in [Ca2+Iiis thetriggering event for exocytosis in p-cells (32-34). A transient [Ca2+Iiincrease associated with biphasic insulin release, observed in this study, is in agreement with the idea that a transient [Ca2+liincrease could initiate a sustained insulin release (33, 35). PACAP exhibits 68% amino acid homology with VIP (61, and type I1 receptors for PACAP are shared by both peptides (11). VIP also increased [Ca2+Iiin p-cells in the presence of 8.3 m~ glucose in aconcentration-dependent manner over the range of to M (Fig. 3, B and C ) .VIP at M or lower failed

PACAP: Potent

a & h t PACAP!27. A,nerve fibers (n), islets (i), andblood capillaries (c) were stained, while acinar cells remained unstained. B, high magnification of islets. Intensely stained cells were localized at the central portionof islets. Bars indicate 50 pm.

Intra-pancreatic Regulator of Insulin Release

1293

,

I

to evoke the [Ca2+li response. Hence, PACAP27 appeared 4 logs more potentthan VIP (Fig. 3C). Therefore, it seems likelythat PACAP27 at low concentrations reacts with a specific PACAP receptor in p-cells which may have an extraordinarily high

aEnity.

M stimuA38-residue formof the peptide (PACAP38)at lated insulin release from perifusedislets and increased [Ca2+Ii in p-cells, in an almost identical manner to PACAP27 at M (Fig. 5). The effectsof PACAP38were also glucose-dependent and inhibited by nitrendipine.2 PACAP38 at and M also increased insulin release, whereas it was previously reof little effect ported that PACAP38 in a nanomolar range was on insulin release in vivo in mouse (36). The discrepancy between the two studies may bedue to the different experimental conditions employed. We studied PACAP-like immunoreactivity in rat pancreas using a rabbit antiserum against PACAp27, which recognizes both PACAP27 and PACAP38, but not VIP and other related peptides (10). PACAP-like immunoreactivity was observed in the nerve fibers, blood capillaries, and islets (Fig. 6A). The central part of islets was intenselystained while the periphery was vague (Fig. 6B). The staining disappeared when the antisenun waspreabsorbedwith PACAP27, but notwith VIP. Therefore, the staining is considered specific for PACAP. The PACAP-like immunoreactivityin the nerve fibershas been also previously reportedin pancreas of rat and mouse (36).Whether the PACAP-like immunoreactivity observed in islets reflects the peptide releasedfrom the nerves or that produced by islet cells is yet to be determined. The present study demonstrated that PACAP in the subpicomolar range activates islet p-cells and stimulates insulin release in a glucose-dependent manner.Thus, PACAP appears to be at least 3 logs more potent than VIP, tGLP-1 (4, 201, glucagon (31, and other insulinotropic peptides. PACAP also occurs in pancreatic nerve fibers and islets. The h d i n g suggests that PACAP could serve as a neuronal and/or local hormonal regulator of the glucose-induced insulin secretion.

Acknowledgments-We thank Kayoko Itoh and K a o n Yamamoto for technical assistance andDr. Geoffrey W. G. Sharp for carefully reading the manuscript.

T. Yada and M. Sakurada, unpublished data.

REFERENCES 1. Wollheim, C. B., and Sharp, G. W. G. (1981) Physiol. Reu. 61,914-973 2. F'rentki, M., and Matschinsky, F.M. (1987) Physwl. Rev. 67,1185-1248 3. Pipeleers, D., Veld, P. I., Maes, E., and Winkel, M. V. D. (1982)P m . Natl. Acad. Sci. U.S.A. 79,7322-7325 4. Mojsov, S.,Weir, G. C., and Habener, J. F. (1987) J. Clin. Invest. 79,616-619 5. Orskov, C.(1992) Diabefologia 95,701-711 6. Miyata, A,. Arimura, A,, Dahl, R. R., Minamino, N., Uehara, A, Jiang, L., Culler, M. D., and Coy, D. H. (1989) Eiochem. Eiophys. Res. Commun. 164. 567474 7. Arimura, A. (1992) Regul. Pepf. 37,287303 8. Miyata, A, Jiang, L., Dahl, R. D., Kitada, C.,Kubo, K., wino, M., Minamino, N., and Arimura,A. (1990) Eiochem. Eiophys. Res. Commun. 170,643-648 9. Arimura, A, Somogyvari-Vlgh, A,, Miyata, A., Mizuno, K, Coy, D. H., and Kitada. C. (1991) Endocrinology 129,2787-2789 10. Koves, K., Arimura, A,, Somogyvari-Vigh, A, Vigh, S., and Miller, J. (1990) Endocrinology 127,264-271 11. Gottsehall, P. E., Tatsuno, I., Miyata, A,, and Arimura, A (1990)Endocrinology 127,272-277 12. Goth,M. I., Lyons, C.E., Canny, B.J., andThorner,M. 0.(1992)Endocrinology 130,939-944 13. Tatsuno, I., Yada, T., Vigh. S.. Hidaka, H., and Arimura, A. (1992)Endocrinology 131,7341 14. Canny, B. J., Rawlings, S. R.. and h n g , D.A. (1992) Endocrinology 130, 211-215 15. Yada, T., Vigh, S., and Arimura, A. (1993) Peptides 14,235-239 16. Isobe., K, Nakai, T., and Takuwa, Y. (1993) Endocrinology 132,1757-1765 17. Mungan, Z.,Ertan, A, Hammer, R. A,, and Arimura, A (1991) Peptides 12, 559-562 18. Raufman, J.-P., Malhotra, R., and Singh, L. (1991) Regul. Pept. 36,121-129 19. Warren, J. B., Donnelly, L. E., Cullen, S., Robertson, B. E., Ghatei, M. A, Bloom, S. R., and MacDermot, J. (1991) Eur J. Pha-01. 1S7,131-134 20. Yada, T, Itoh, K, and Nakata, M. (1993) Endocrinology, in press 21. Grynkiewicz, G., Poenie, M., and Tsien, R. Y. (1985)J. B i d . Chem. 260,34403450 22. Graham, R. C.,and Karnovsky, M. J. (1966) J. Histochum. Cytochem. 14, 291302 23. Roth, J. (1982) J. Hisfochem. Cytochem. 30,691-696 24. Hacker, G. W., Grimelius, L., Danscher, G.,Bernatzky, G., Muas, W., Adam, H., and Thurner, J. (1988) J. Histofechnol. 11.213-221 25. Koves, K , and Arimura, A (1989)J. Hisfochem. Cytochem. 37,903-908 26. Kawai, K., Ohse, C., Watanabe, Y., Suzuki, S., Yamashita, K, and Ohashi, S. (1992) Life Sci. S O , 257-261 27. Sakurada, M., Kanatsuka,A., Saitoh, T., Makino, H.,Yamamura, K., Miyazaki, J., Kikuchi, M., and Yoahida. S.(1993) Endocrinology 132,2659-2665 28. Grapengiesser, E., Gylfe, E., and Hellman, B. (1988) Eiochem. Eiophys. Res. Commun. 160,419-425 29. Yada, T., Kakei, M., and Tanaka, H. (1992) Cell Calcium 13,69-76 30. Ashcroft, F. M.,and Rorsman, P. (1989) Pmg. Eiophys. Mol. Eiol. S4,87-143 31. Smith, P. A,, Rorsman, P..and Ashcroft, F.M. (1989) Nature 342.550453 32. Penner, R.,and Neher, E. (1988) J. Exp. Eiol. 139, 329-345 33. Takasawa, S., Nata, K, Yonekura, H., and Okamoto, H. (1993) Science 259, 370373 34. Ammala, C., Ashcroft F. M., and Rorsman, P. (1993) Nature SSS.356358 35. Yada, T.,Russo, L. L., and Sharp, G. W.G. (1989) J. BWZ. Chem. 264.24552462 36. Fridolf, T., Sundler, F., and Ahren, B. (1992) Cell Illssue Res. 269,275-279