Vagal, Cholinergic Regulation of

Pancreatic Polypeptide Secretion

T. W. SCHWARTZ,J. J. HOLST, J. FAHRENKRUG,S. LINDKLER JENSEN, 0. V.NIELSEN, J. F. REHFELD, 0. B. SCHAFFALITZKY DE MUCKADELL,andF. STADIL, Institute of Medical Biochemistry, University of Aarhus, Arhus;Department of Clinical Chemistry, Bispebjerg Hospital, Copenhagen;Department of Surgical Gastroenterology C, Rigshospitalet,Copenhagen, Denmark

A B S RA C T The effect of efferent, parasympatheticstimulation upon pancreatic polypeptide (PP) secre-tion was studied in three ways: (a) Plasma PP concen-trations increased in response to insulin-induced hypo-glycemia in both normal subjects, from 11 pM (9.5-12.5) to 136 pM (118-147), n = 8 (median and inter-quartile range) and in duodenal ulcer patients, from 33pM (21-52) to 213 pM (157-233), n = 7. The PP re-sponse to hypoglycemia was diminished by atropine innormal subjects (P < 0.005) and completely abolishedby vagotomy in the duodenal ulcer patients. (b) Electri-cal stimulation, 8 Hz, of the vagal nerves in anesthe-tized pigs induced an increase in portal PP concentra-tions within 30 s from 32 pM (28-39) to 285 pM (248-294), n = 12. Minimal stimulatory frequency was 0.5Hz and maximal stimulatory frequency 8-12 Hz. Atro-pine inhibited the PP response to electrical stimula-tion. Median inhibition with 0.5 mgof atropine/kg bodywt was 74%, range 31-90%, n = 6. The response waseliminated by hexamethonium. Adrenergic alpha andbeta blockade did not influence the release of PP inresponse to vagal stimulation. (c) Acetylcholine stimu-lated, in a dose-dependent manner, the secretion of PPfrom the isolated perfused porcine pancreas, half-maxi-mal effective dose being 0.19 ,uM; maximal PP outputin response to 5 min stimulation was 228 pmol, range140-342 pmol, n = 5. Atropine completely abolishedthis response.

The results of the present study together with thepreviously demonstrated poor PP response to food invagotomized patients, indicate that vagal, cholinergicstimulation is a major regulator of PP secretion.

This work was presented in part at the 11th Acta Endo-crinologica Congress, 20-23 June 1977, Lausanne, Switzer-land, Abstract 161. 1977. Acta Endocritnol. 212 (Suppl.): 106.(Abstr.)

Received for publication 18 Jtuly and in revised form 25October 1977.

INTRODUCTION

Humanpancreatic polypeptide was isolated by Chanceet al. as a side fraction during insulin purification (1).It is a 36 amino acid straight chain polypeptide with nosimilarity in sequence to any other known hormone.One or two aminoacids distinguish the known mam-malian pancreatic polypeptides (PP)l (1), whereas theyhave only 16 amino acids in common with the aviancounterpart (2).

PP has been localized to secretory granules in a spe-cific endocrine cell type in the pancreas distinct fromthe A, B, D, and D, cells (3, 4). PP cells are foundboth in the islets and in the acinar tissue and ductepithelium. In some species PP cells are numerousin the extra-insular tissue and concentrated towards theduodenal part of the pancreas, whereas e.g., in the isletsof the sheep the PP cells nearly outnumbers the insulincells (4).

The action of PP has not yet been established al-though PP has been shown to affect many gastrointesti-nal functions (5). At present it seems likely that PP actson the exocrine pancreatic tissue as a local regulator ofsecretion, in a way that may be conceived as an anti-cholecystokinin and secretin-enhancing action (6, 7).It has recently been suggested that PP may also affectsatiety (8).

During a meal PP concentrations in plasma increaserapidly (9-13). The fool-mediated PP response consistsof a rapid primary phase (5-30 min) which is elimi-nated by truncal vagotomy (10) and a prolonged sec-ondary phase (0.5-6 h) which in man is reduced butnot abolished by vagotomy (10). The present study wasundertaken to investigate the effect of efferent para-sympathetic stimulation on PP secretion.

' Abbreviationis used in this paper: DU, duodenal ulcer; PP,panereatic poly-peptides; VIP, vasoactive intestinal polypep-tide.

The Jouirnal of Clinical Investigation Volunme 61 March 1978 781-789 781

MIETHODS

Radioimnmu noassay for paticreatic polypeptideAnitiserum. Anti-PP serumll (146-5), at generouis gift from

Drs. R. E. Chance and N. E. Moon (Eli Lilly & Co., Indianap-olis, Ind.), was raised in a rabbit against subcutaneously in-jected highly purified bovine PP (BPP) coupled to albuminand mixed with complete Freund's adjuvant. The anltiserumwas used at a final dilution of 1/7.5 x 106.

Tracer. Highly Ipurified bovine PP was iodinlated by achloramine T method: 1 mCi 1251 (Behrinig-Werke AG, Mar-burg/LLahi, West Germclany), 0.24 numlol (1 ,ug) highly pturifiedBPP (a gift from Dr. Chance, Eli Lilly & Co.), and 18 nmol(5 ,g) chloramine T was mixed in a total volume of 0.023 ml/0.05 M sodium phosphate buffer, pH 7.4. The reactioni wasterminated after 20 s by the addition of 26 nmol (5 ,ug) sodiunmetabisulfite in 0.010 ml sodium phosphate buffer. Theni 1.2gnmol of potassium iodide in 0.1 nl wvas added followed by0.2 ml of equine serum. The tracer was purifiedl on a 10 x 1,00(mmSephadex G-50 superfine columni eluted with amlnmoniumbicarbonate,( 0.25 Xi, pH 8.2, at 4°C. The fractioins froml theascending part of the 1251_PP peak gave the lowest unspecificbinding without antibody (1-3%) and the highest binding inantibody surplus (90-95%). The specific radioactivity of thetracer was calculated to 200-250 /.Ci/4g (0.8-1.05 x 106 Ci/m10ol).

Standard. Highly purifiedl humclan PP (a gift from Dr.Chance, Eli Lilly & Co.) was usedl as standard.

Assaly condition.s. Incubation was carriedl oUt in 0.02 so-dium barbitonie buffer, pH 8.4, with 0.2% bovine serumii al-bumin (Ortho Pharmaceutical Corp., Raritani, N. J.) and 0.6mnMthiomersal. The incubationi volume was 1.5 ml. In assaysof plasma the saimple volume comprised 1/20 of the incubationvolume and in other assays at most 1/10 of the incubationvolume. To obtain comparable incubation coniditions "hor-mnone-free" plasma, twice treated with charcoal, was added tothe standard and the same volume of buffer was added to theplasmiia samples. Incubation for 3 days was used in all assaysexcept in assays on column eluates, where preincubation for2 days was followed by 1 day of incuibation with tracer.

Separation. Separation was performed with plasma-coated charcoal (Merck article 2186, Merck AG, Darmstadt,West Germany). To each tube 0.5 ml of a preincubated char-coal slurry, 0.25 ml charcoal 40 g/liter and 0.25 ml outdatedplasma, was added. Centrifugation was performed 20 mni afteraddition of charcoal. Control samples were included in eachcentrifugation.

Specificity. No displacenment of the tracer from antiserumcould be demonstrated with up to 10( nM of: monocomponenithuman insulin (Novo Research Institute, Bagsvaerd, Den-mark); monocomponent porcine pancreatic glucagon (NovoResearch Institute); synthetic ovine somatostatin (BeckmanInstruments, Geneva, Switzerland); synthetic human gastrin-17 (ICI Pharmaceutical Co., Alderly Park, Cheshire, England);highly purified porcine cholecystokinin (CCK 33); highlypurified porcinie secretin; and highly purified porcinevasoactive intestinal polypeptide (VIP) all generous giftsfrom Dr. V. Mutt, Karolinska Institute, Stockholm, Sweden,and highly purified porcine gastric inhibitory polypeptide, agenerous gift from Dr. J. Browni, University of BritishColumbia, Vancouver, Canada.

Sensitivity and precision. The experimental detectionlimit during equilibrium incubation was 2.8 pM and in pre-incubation assays 1.0 pM. The within-assay coefficient ofvariation was 6.8, 7.1, 8.0, and 6.8% in buffer standards of12, 25, 60, and 120 pM. The coefficient of variation was 9.7and 12.7% in plasma samples of 15.7 and 49.7 pM. The be-

tweeni-assay coefficient of variation tested in 10 consecutiveassays at concentrations of 30 and 75'pM was 12.8 and 12.1%.

Accuracy. Measurements of PP in plasma with added PP,and in dilutions of perfusate from the isolated panicreas (seebelow), porcine plasma, and human plasmla yielded resultswhich deviated <20% from the expected values. The follow-ing specimens were exposed to gel permeation chromiiatog-raphy: (a) perfusate from the isolated porcine pancreas duringstimulation with acetylcholinie; (b) porcine plasma obtainedfrom the portal vein during electrical stimulation of the vagusnerve; and (c) human plasma obtained during in.sulini hypo-glycenmia. The gel permeation chromatography was performedat 4°C on 10 x 1,000 mmSephadex G-5( superfine columnseluted with 1.0 Macetic acid with 2%equine serum and 0.2%bovine serumii albumin. The fractions were lvophilize(d andreconstituted in barbitonie buffer 0.02 M, pH 8.4. In everv caseonly one major immunoreactive peak with elution positioncorresponding to purified PP was found, Fig. 1. Wheni plasma

200-

150 -

x

E0C

w

Q-UJ

CL

-J0a.

0

wIr0z0L

100 -

50 -

0-

50 -

25 -

0-

75-

50-

25 -

0 -

A

B

6 25 50

ELUTION VOLUME

75 100

PERCENTAGE

FIGURE 1 Elution diagrams of PP immunoreactivity in: Aperfusate from the isolated perfused porcine pancreas, B hu-man plasma obtained during insulinhypoglycemia, C porcineplasma obtained during electrical stimulation of the vagus.The samples were applied to Sephadex G-50 superfine col-umns (10 x 1,000 mm) eluted at 4°C with 1.0 M CH3COOHcontaining 2%equine serum and 0.2% bovine serum albumin.The columns were all calibrated with 22NaCl and column Aalso with '25I-albumin. Abscissa: percent of volume betweenthe elution volume of 125I-albumin and 22NaCl. The arrow ontop indicate the elution position of highly purified human PP.

782 Schwartz et al.

wvas applicated without internal void voltume marker apparentPP immunoreactivitv was determined in anid closely after thevoid volume of the columiln (Fig. 1). A simiiilar apparent im-munoreactivitv coulcI l)e detected in the void volume when thehorm one-free charcoal-treated plasnma that usually was addedto the standards was applied. This apparent PP immuni oreac-tivitv could not be removed by specific PP immunoabsorption(antibody from the same rabbit was used for assay). WVethere-fore conelude that the present assay meassures only a peptidewhich elutes in the samiie position as the highly purified PP.Fasting levels of PP in plasma from y ouing people (20-40 yr)weere 13 pMI (range 2-30), it = 50, in agreenment with resultsobtained by Floyd et al. (13) with another PP antiserun, withiodinated highlx purified hulmian PP as tracer, and wvith anotherseparation techniique.

Insulin hiypoglycenmiaPatientts. Eight normiial sul)jects, two femlales an(d six

males, mediani age 22 yr, range 19-25, and seveni duodencalulcer (DU) patients, one female and six miales, mediani age 52vr, ranige 34-60( were studied after an overnight f:ast. Bloodwas taken from an aintecubital vein, cenitrifuged at 4°C acndserum was stored at -20°C before assayeed. Gastric juice wasaspirated continuously during the experimenit. Data concerni-ing this acid secretioni anid gastrini release have been published(14, 15).

Stimiulatiotn. Hypoglycemia was achieved by i.v. injectionof0.2 IU of instulin (Insulin LEO, Nordisk Insulinl LaboratoryGentofte, Denmark) per kilogramii body weight. Blood glucoseconicentration, measured by the glucose oxidase method de-creased in all subjects l)elow 1.8 mM.

In1hibitioni. The normual subjects were studied on separatedays with insulin alonie or after i.m. injection of 30 jig atropine/kg body wt given 25 min before the insulini. The DUpatientswere studied before and 13 wk (8-26) after their operation,trunical vagotomv, and pyloroplasty.

Electrical stimulation of the vagel nerves

Preparationi. 28 female pigs, DaIIish Landrace, weight: 25kg (20-32) were usedl. Food, but not water, was withdrawn 18h before experiments. Azaperone (4 mg/kg i.m.) was used forpremedicationi. Anesthesia was induced vith Halothane/N20)per 02 and maiintained with chloralose (.Merck AG, Darmstadt,West Germany) 70( mg/kg i.v. Polyethylene catheters wereinserted into the portal vein and left femoral arteryand vein. Intra-arteirial blood pressure and rectal temper-ature were continuously monitored (EMT 34 trans-ducer, mingograph, E. Sch6nander, Stockholm, Sweden,and temperature recorder, Z8, ellab, Copenhagen, Denmark).A laparotomy was performed, the pylorus was ligated, and gas-tric content was continuously drained by suction via an oro-gastric tube. The splanichnic nerves were cut bilaterally.Through a thoracotomy the vagal trunks were identified abovethe diaphragm anid cut below the heart. The peripheral endsof the nerves were passed through a bipolar platinum elec-trode. Blood was collected from the portal vein and the femoralartery, in ice-chilled tubes, containing 500 kallikrein inactiva-tor un1its aprotinin (Trasylol, The Bayer Co., New York) and50 IU heparin/ml blood. Samples were centrifuged at 40C andplasnma stored at -20°C until assayed.

Stimulation. 5 miD stimulations with constant-current (8mA) square-wave impulses (frequency 0.25-20 Hz, impulseduration 5 ins) were carried out by means of an electronicsquare-wave stimulator (Palmer, England) Applied voltage

andl eurrenit ais monitored on a dual-beam oscilloscope (type502A, tektron ie, Guernsey Ltd., Englandl).

Itnlhibitioni. Atropine, 0.00.5, 0.05, 0.5-mg/kg, wvas injectedlbefore repeated sull)maximal stimulationis in one pig, atropine0.5 mg/kg in six different pigs. Hexamiiethoniumt, 10( mg/kg wasusedl in four dlifferent pigs. In four other pigs submaximalstimulcationis without inhibition wvere repeated. In five differ-eut pigs combined alpha and beta blockade was performed:propraniolol (In{deral, ICI Pharmiiaceeuticatl Co.) was inljectedi.v. in a (lose of 1 mg/kg. Phenoxybenzarmine chloride,(Alf'redl Benizon, Copenhllagein, Denmllark) in a dose of 1 mg/kg(lillited in 100 ml isotoniic saline wvas inifused dutiring 1 h, afterwvhiiel the vagal stimuilation wvas repeatedl. In these experi-menits dextran, Macrodex 6%, Phlarnacia, Uppsala, Swveden,was infulsed to careftlly substitute blood anid fluiid losses.

Isolated1, perfused, porcinie pa ucreas;ace tylcholihie

Preparatiotn. The isolationi anid preparation procedure hasI)een dlescribed in detail elsew.,here (16). Five pigs, DanishlLandrace, weight: 26-3() kg, median 28 kg, were uised. Thepainereas was isolatedl together with their enitire vascular sup-ply; the dllodeilnlim was nlot incluided in the preparation butthe panereaticodniodealll vessels renmained intact. The pan-creases were perflisedl It 380C with Krebs-Ringer bicarbonatesolutioni enrieled with fumnarate, glutamiiate, and pyruvate all5 mM, Dextrani T 70 (Pharnmacia, Uppsala, Sweden) 4%, humanseruim calbumin (Trockeni Reinist, Behring Werke/Lhl}in, \VestGermany) 0.1%, aind 5 mMgluicose. The perfusate was oxy-geniated with 95% oxygen and 5% CO2 anld perfused throuighthe paciereats at a flo!v rate of 30( ml/min, thus the oxygencoinsuimiiptioni could be mlaintainiedl at 0.006-0.008 nml O2/I-inper g tissue. Flow rate, pressure, temiiperatuire, Pco2, Po2 and( p1of the perfusate were measuiredl at 20 mni intervals. The ef-fluent was collected everv m-inutte and stored at -20(C unltilassaved.

Stimutlationi. After a resting period of 30) miii 5mIistimulil-tion period(s with acetvlcholine 0.05, 0.1, 0.5, 2.5, and 100/.M (final concenitrationi in perfusiate) were carried out sepa-rated by 15 min restinig periods.

Inhibitioni. Atropine, 0.5 uiM (final concenitrationi in per-fusate), vas infuised 5 min before and( durinig sttimulition withacetyleholine, 0.5 jNIM.

StatisticsThe results are, if not otherwise iildicated, preseinted as

medlianii coneentrationi followed bw the iinter(luartile rainge inbrackets. The \Vilcoxon imiatched-pairs signed-ranks test orwheni appropriate the Mann-Whitney test or the Friedmauniitw%o-wNay analysis of variance were used for statistical evaluia-tion (17). Differencees resultinig in P values <0.05 were con-sidered signiificant.

RESULTS

Inisulin hypoglycemliaDurinig insulini hypoglycemia the plasmla concenitra-

tions of PP in the eight normal subjects increased froml11 pM (9.5-12.5) to a maximum of 136 pM (118-147)at 45 min after the injection of insulin; PP levels werestill elevated 180 mmi after the iiujection, 46 pM (29-55), Fig. 2. Atropine did not significanitly alter basal

Vagal, Cholinergic Regulation of Pancreatic Polypeptide Secretiotn 783

x 100

'aECL

0L

700

- 600

- 500

-toL00 )

x

300 CL0-

200 -

150 -

200 *_

100 0 100 -0-

-60 -30 0 30 60 90 120 150 180

TIME (Minutes)

FIGURE 2 Plasma PP concentrations, median concentrationand interquartile range (hatched area) in 8 normal subjectsduring insulin hypoglycemia (0.2 U x kg-' body weight) with-out (0) or with (0) atropine (0.03 mg/kg, i.m.) injected 25min before insulin.

50 -

0

PP concentrations, 10.1 pM (9.3-11.5) 5 min before and9.5 pM (9.0-14.0) 25 min after atropine. However, theresponse to insulin hypoglycemia, from 9.5 pM (9.0-14.0) to a maximum of 31.5 pM (28.5-31.5) at 45 minafter insulin, was diminished by atropine (P < 0.005)at all points from 30 to 120 min after insulin injection.

Basal PP concentrations in the DU patients beforethe operation were 33 pM (21-52), significantly higherthan those of the younger normal group. The PP re-sponse pattern to insulin hypoglycemia before the op-eration was similar to that of the normal group, Fig. 3.However, the maximal PP concentrations were higherin the DU patients, 213 pM (157-233). After truncalvagotomy basal PP concentrations, 17.5 pMI (9.5-27.0),were lower than preoperative levels (P < 0.05), and notdifferent from the basal PP levels in the normal group.No increase in PP concentrations was observed dur-ing insulin hypoglycemia in the DUpatients after trun-cal vagotomy, Fig. 3 (P > 0.10, Friedmann two-wayanalysis).

Electrical stimulationFrequency-response studies were carried out in

three pigs, Figs. 4 and 5. Half-maximal stimulatory fre-quency was 2.5, 3.3, and 5.6 Hz, and maximal stimula-tion was achieved with frequencies of 8, 12, and 12 Hzrespectively. Increase in portal plasma PP concentra-tions could be detected during stimtilatioin with one

- 60 -30 0 30 60

TIME (Minutes)90 120

- 1000

- 900

800

- 700

- 600

-500 X

0.(L

-L00

- 300

- 200

- 100

0

FIGURE 3 Plasma PP concentrations, median concentrationand interquartile range (hatched area) in seven duodenal ulcerpatients during insulin hypoglycemia (0.2 U/kg body wt) be-fore (0) and after (0) truncal vagotomy.

7x

E0.

0-0L

2.5

2.0

1.5

0.5

0 60 120

MINUTES

FIGURE 4 Portal plasm11a PP coieenitratiolis dutrinig gradedelectrical stimulation with fre(uenecies from )0.25 to 20 Hz, in-dicated by bars, of' the vagal nerves in a pig.

784 Schwartz et al.

100

z0I-

-JD

x(I)-J

Ix

I

LiL0

0

z

0:wLa-

75

50

25

.

0

A

00

300 -

7

x

O 200 -E0.

a-0L

100-

A

0

0.3 1 3 10

STIMULATIONS PER SECOND

FIGURE 5 Integriate(I PP responise in three indclividual pigsdutiring gradedl stimlltiation of the vagal nerves. The experimenltC1emoInstrated in Fig. 4 is inidicatedl hy (0). The responisesshowvn have been callcilated b) initegrationi of PP conieentra-tioin cturves dturing stimulclationi periods minutsi prestimulatorylevels, if basal PP levels was niot reached before repeatedstimulationi a caleuilated lbasal level was suhtracted.

impulse every other seconid, l)ut not with stimuitilationisevery 4th s.

During submaximal stimuitilation, 8 Hz, PP conicenitra-tions in portal plasmia increased within 30 s from 32 pMi(28-39) to 285 pNM (248-294), ni = 12, Fig. 6. PP con-cenitrations in portal plasma, anid in arterial plasmiia (Fig.8), remainied elevated throughout the 5 min stinilulationperiod. The PP response could be reproducedl in thesaime pig duiring repeated stimultlationi 3 h afterthe primary stimutlation. The integrated PP re-sponise durinig restimulationi averaged 91% of the pri-m-ar responise (rainge: 84-129%) ni = 5, Fig. 7.

Atropinie iii cl dose of 0.05 mi,g/kg was stufficienit toinihihit the PP release in response to submaximal elec-trical stimilulation of the vagal nierves, Fig. 8. However,even with 0.5 mg/kg of atropine the degree of inhibitionvaried; miiediani inihibition: 74% (ranige: 31-90%) n= 6, Fig. 7. Inijectioni of' hexamiiethoniumlil 10 miig/kg le-fore vagus stinudliationi elimliinatedl the PP responise,Figs. 9 anid 7. Comiibined alpha (phenoxybenzaminiie)alnd beta (propraniolol) adreniergic blockade dicl niot sig-nificaintly inifluienice the PP responise to vagal stimutla-tioll, n = 5, Fig. 7.

Cessationi of the electrical stimiutilation resuilted in a

-15 -10 -5 0 5 10

TIME (MINUTES)

1500

- 1000

Ex

cmC)

0L0-

- 500

15 20

FIGURE 6 Portal plasim1a PP concenItration1s, m11ediaii conIeen1-trationi anl interdlliartile rainge (hatchedI area) in responise to8 Hz electrical stimtilation, inidicated 1v bar, of the vagalnerves in pigs, n = 12.

0

00

100 -

-: Z 75 -:y O

0- F-

LLO D 50-

Z F-

) 25-wLLJ0 O

0 -

/0

E

0

/-6

RES ATR HEX ADR

FIGURE 7 The effect of (lifferent blocking agenit uipoin 8 Hzelectricial, vagal stimlllliationi of PP secretioni. Ordiniate indi-cates initegratted PP response cluirinig restimulaltioni as percentof initegratedl PP responise dutiring the primary stimutilationl.RES.-restimulation without anyx blocking agenits. ATR. Atro-pinie 0..5 miig/kg, HEX.-hexamethoniiumil, 10 Imlg/kg, ADR-com-bined atlpha, phenoxybenzamine 1 miig/kg, an(d beta, proprain-olol 1 miig/kg, adreniergic blockade.

Vagal, Clholitnergic Reguilationi of Patncreatic Poltlpeptide Secretion 8

* 6

785

300

200

7x

E

0-

100 -

x

500 0-

60

MINUTES

FIGURE 8 The effect of increasing doses of atropine on theplasma PP concentrations in the portal vein (*) and in thefemoral artery (0) during repeated 8 Hz stimulations, indi-cated by bars, of the vagal nerves in a pig.

rapid decline in PP concentrations, Fig. 6 and 8. Aftera more rapid decrement during the first five min, thesemilogarithmic disappearance curve for PP was linearfrom 5 to 15 min after cessation of the stimulation,suggesting a half time of disappearance for endoge-nously released PP of 5.7 (4.5-6.9) minutes; ii = 12.

The isolated, perfused, porcine pancreasBasal output of PP from the isolated pancreas was

2.6 pmol/min (range: 1.4-3.6 pmol/min). Acetylcholine

13 -

12 -

11-

I

U -

- 8z

0. 7

63-zW 5-zo 4

2 -

1-11- 15 -10

increased the secretion of PP from the isolated pancreasin a dose-dependent manner, Fig. 10 and 11. Thelowest dose tested, 0.05 uM, stimulated the PP secre-tion in all experiments. Half-maximal effective dose ofacetylcholine was 0.19 ,uM (range: 0.07-0.80 ,uM).Maximal output of PP in response to 5 min stimulationswas 228 pmol (range: 140-342 pmol). In all experi-ments the stimulation of PP secretion was abolished byatropine (Fig. 10).

DISCUSSION

The present study shows that efferent, vagal stimula-tion releases PP, and that this is probably mainly medi-ated by acetylcholine. Together with the previouslydemonstrated poor PP response to food in vagotomizedpatients (10) these data indicate that vagal, cholinergicstimulation is of major importance for the secretionof PP.

Insulin hypoglycemia. The increase in concentra-tion of PP in plasma during insulin hypoglycemia issimilar to the increase observed during a meal (11,13, 18). This response is closely correlated to thehypoglycemia (11, 13). The fact that atropine stronglyinhibits the PP response (Fig. 2) indicates that vagalactivity and not hypoglycemia per se stimulates the PPcell. Accordingly low glucose concentrations did notenhance the release of PP from the isolated pancreas(unpublished observations). On the other hand thestudy with atropine dose not prove that the PP responseto hypoglycemia is mediated by a peripheral, choliner-gic transmission, because atropine might also act on thecentral nervous system (19).

The total elimination of the PP response to hypo-

2 -

0-0-

- 5 0 5 10 15 20 25 30TIME (MINUTES)

FIGURE 9 Portal plasma PP concentrations, median andtotal ranige (hatched area), during electrical stimulation, in-dicated by bar, of the vagal nerves without (-) or with(0) pretreatment with hexamethonium, 10 mg/kg body wt,n = 4. The ordinate indicates fractional PP concentration ob-tained by division of actual concentration vith concentrationat time zero.

0.05 01 0.5 2.5 100 05

0 20 00 60 80 100 120 100

TIME (minutes)

10

-5

-0

cE

160

FIGURE 10 The effect of increasing concentrations of acetyl-choline (0.05-100 ,uM), indicated by closed bars, upon PPconcentrations in the perfusate from an isolated porcine pan-creas. Addition of atropine, 0.5 ,tM, is indicated by open bar.

786 Schwartz et al.

3 -

100 -

0-4

I-

4

2so

z

LLJ

cKLl 11

c

, o01

ACETYLCHOLINE (MOL xIl-I x 10-6)

FIGURE 11 Initegrated PPoutptutfrom0 the isopainereas in response to graded stimulationi within five individual pancreases. The ouitput wasintegrationi of PP concentration curves from tiacetvlcholine started to PP levels retturnied toplevels, minus this concentrationi, mutltiplied x

iate.

gly cemia by truncal vagotomy (Fig. 3)indicordance with findings by Adrian et al.(1Ethe increment in PP concentration duringi]glycemia is caused by vagal activity. Otheractivated during hypoglycemia, e.g. the athe sympathetic nervous system, which miaffect the secretion of PP, are entirely depeintact vagal innervation.

Electrical stimulation of the vagal 10large increase in PP concentrations in p(during electric stimulation of the vagal nerthat efferent vagal stimulation is a potent reanism for PP. The increase in plasma conciPP during vagal stimulation does reflect iisecretion, since portal bloodflow was unchcreased up to 40%during these stimulationover, a similar increase could be detectectemic circulation, Fig. 8, which proves acreased splanchnic output.

That the threshold frequency for PP releone stimulation per second and that maxin-is achieved with 8-12 Hz stimulation is iIwith the supposed physiological discharglperipheral autonomic nervous system. Bfrom single or a few nerve fibres of bothand parasympathetic nerves a continuousimpulses with frequencies at -1 Hz is fcbasal conditions, and during intense reflexthe discharge rate is in the order of 10 Hthermore, similar frequency-response curvtained in studies on electrical, vagal stimulsecretion (22).

The elimination by hexamethonium of

o sponse duiring vagal stimuitilation (Fig. 9) indicates that* a niicotinic receptor isintercalated in the peripheral

vagal pathway distal to the diaphragm. This is in accord. with the coincept of a classic parasympathetic pathway

with a peripheral, postganglionary neuirone activatedbynicotiniic receptors.

The inhibitorx effect of atropine (Fig. 8)indicatesthat the trainsillissio from the postganglionary neuronieto the PP cell is, at least in part, mediated by acetyl-choliine acting on amuscarine receptor. In agreenmentwith the effect of atropine upon postganglionary, vagalstimulatioin of other gastrointestinal functions (19),large amiiounts of atropine are necessary toinhibit elec-

10 loo trical, vagal PP release. However, even with a largedose of atropiine, 0.5mg/kg, only 75% of the response

latedc poreinie couldl)e abolished (Fig. 7); thus other transmitter sub-acetxlcholine stances apart from acetylcholine could be involved. Itcalecultedcl) isnOt likely that the catecholamine cointaiining vagalresfinfuionofv fibres (23) play anymajor role during electrical stimula-restimullastorv

with the flO tion because combined alpha and beta adrenergicblockade does not influence the PP response, Fig. 7.Furthermore electrical stimiiulation of the splanchniicnerves does not enhance PP secretion (unpublished

icates, in ac- observations). Peptidergic transmission might be in-S), that all of volved in the neural stimulation of PPl)ecause e.g.nstslini hypo- VIP (24) aind gastrin (25) have been found in the vagus.

mlechanllismslS Furthermore VIP containiing cell bodies are in soIm1etdrenals; anld species localized in pancreatic ganglia,2 aind VIP is re-ight possil)ly leased by an atropine-resistant mechanism during elec-Zndent o01 all trical, vagal stimulatioin (20). Finally, VIP andmembers

of the gastrin family of gastrointestinal horniones cainervets The release PP (18,26). However, the possible involvementortal plasmiia of such substances in the electrical, vagal stimulationves conifirmiis of PP secretion is presently difficult to delineate.lease mech- The isolated porcine pancreas. Low doses of acetyl-entrations; of choline stimulated the secretion of PP from the isolatedocreased PP pancreas. The half-maximal effective dose of acetylcho-anged or in- line (0.2 ,uNl) for release of PP from the isolated porcines (20). MIore- pancreas is similar to the D50 of acetylcholine for stimu-I in the sys- latioin of the motor activity in the gastrointestinal mus-

In acttial in- culature (27) and for stimulation of insulin release fromthe isolated perfused rat pancreas (28). It is reasonable

ase is b)elow to assuime that acetylcholine acting on muscarinic re-llal seretionl ceptors also is involved in the neural stimulation of PPn agreemilent secretioni, since atropine inhibited vagal stimulation,

rate of' the Figs. 7 and 8, and because the effect of acetylcholiney/ recording oIn the isolated pancreas was completely abolished bysympathetic an e(quimolar dose of atropine, Fig. 10. The amount ofdischarge of PP released during maximal stimulation with acetyl-)und during choline from the isolated porcine pancreas could very

stiimiulation well account for the increase in plasma PP concentra-[z (21). Fur-zes were ob-ation of acid 2 L-I. Larsson, J. Fahrenkrug, J. J. Holst, and 0. B.

Schaffalitzkv de Muckadell. Innervation of the pancreas byvasoactive intestinal polypeptide inimunoreactive nerves. In

F the PP re- preparation.

Vagal, Cholinergic Regulation of Pancreatic Polypeptide Secretiotn 787

tions during maximal, vagal stimulation in the samespecies.

From these three studies taken together it can beargued: that efferent, vagal stimuilation is a potentrelease mechanism for PP; that the vagal pathwayincludes a peripheral neurone activated by nicotinic re-ceptors below the diaphragm; and that the final media-tor probably mainly is acetylcholine acting on mus-carinic receptors.

The importance of efferent, vagal stimulation of thePP secretion in human physiology is illustrated by stud-ies on vagotomized patients. In a prospective studywith each patient being his own control we found thattruncal vagotomy eliminated the first 15-25 min of theotherwise rapid increase in plasma PP concentrationsduring a meal, and also significantly reduced the sec-ondary prolonged response.,10). These findings havepartly been supported by the results of Adrian et al.(18) and have been confirmed by Taylor et al. (29).It has been suggested that the elimination byvagotomy of the primary PP response might bedue to impaired gastric emptying after the operation(30). However, the afferent branch of the vago-vagal reflex for PP release is activated by (pre-gastric) classic, cephalic stimulation (31), and by gastricfood- ar d distention-receptors (26). We therefore con-clude that vagal stimulation of the PP secretion isresponsible for the rapid, primary increase in plasmaPP concentrations during a meal and that it also con-tributes to the prolonged secondary PP response.

Vagal activity does not seem to influence the basal PPsecretion in young normal subjects, because atropinedoes not suppress their basal PPconcentrations (Fig. 2).However, the elevated PP levels found in some DUpa-tients were normalized by vagotomy (10) (Fig. 3). Thedifference between the DU patients and the normalgroup may yet be due to difference in age becausethe patients studied were relatively old, and PP levelstend to increase with age (13). However, fluctuationsin PP concentrations and in the basic acid output inDUpatients are correlated, and moreover atropine nor-malizes the elevated PP (Schwartz, Olbe, and Sten-quist, unpublished observations). Thus, possibly someDU patients have elevated vagal activity, reflected inincreased basal PP concentrations.

ACKNOWLEDGMENT

Jetta Bach Haulrik, Evy D0rge, and Ulla S0gaard are gratefullyacknowledged for their skillful technical assistance. Wearealso very grateful to Dr. R. E. Chance and Dr. Nancy E. Moon,Lilly Research Laboratories, Eli Lilly & Co., Indianapolis,Ind. U. S. A. for donating highly purified pancreatic polypep-tides and antisera.

This study was supported by grants from the Danish Medi-cal Research Council (J.nr. 512-7190) and from Landsforenin-gen for Sukkersyge.

REFERENCES

1. Chance, R. E., F. M. Lin, M. G. Johnson, N. E. Moon,D. E. Evans, W. E. Jones, and J. E. Koffenberger. 1975.Studies on a newly recognized pancreatic hormone withgastrointestinal activities. 57th Annual Meetinig of the Eni-docrine Society. 1: 265/183A. (Abstr.)

2. Kimmel, J. R., L. J. Hayden, and H. G. Pollock. 1975.Isolation and characterization of a new pancreatic pol%-peptide hormone. J. Biol. Chemn. 250: 9369-9376.

3. Larssoin, L.-I., F. Sundler, and R. Hakansoni. 1975. Im-muniohistochemical localizationi of humani pancreaticpolxpeptide (HPP) to a population of islet cells. Cell Tis-sue Res. 156: 167-171.

4. Larsson, L.-I., F. Sundler, and R. Hakanson. 1976. PaIn-creatic polypeptide-a postulated new hormone: identifi-cation of its cellular storage site by light and electronmicroscopic immnunocytochemiistry. Diabetologia. 12:211-226.

5. Lin, T. M., and R. E. Chance. 1974. Gastrointestinalactions of a new bovine pancreatic peptide (BPP). In En-docrinology of the gut. W. Y. Chey and F. P. Brooks, Edi-tor. Charles B. Slack, Inc., Thorofare, New Jersey. 143-145.

6. Lin, T. M., D. C. Evans, and R. E. Chance. 1974. Actionof a bovine pancreatic polypeptide (BPP) on pancreaticsecretioni in dogs. Gastroeniterology. 66: 198-852 A(Abstr.)

7. Lin, T. M., D. C. Evans, R. E. Chance, and G. F. Spray.1977. Bovine panicreatic peptide: Action on gastric andpancreatic secretion in dogs. An. J. Physiol. 232: E311-E315.

8. Malaisse-Lagae, F., J-L. Carpentier, Y. C. Patel, W. J.Malaisse, and L. Orci. 1977. Pancreatic polypeptide: Apossible role in the regulation of food intake in the mouse.Hypothesis. Experientia (Basal). In press.

9. Floyd, J. C., R. E. Chance, M. Hayasha, N. Moon, and S. S.Fajans. 1975. Concentrations of a newly recognized pan-creatic islet polypeptide in plasma of healthy subjectsand in plasma and tumors of patients with insulin-se-creting islet cell tumors. Clin. Res. 23: 535A. (Abstr.)

10. Schwartz, T. W., J. F. Rehfeld, F. Stadil, L-I. Larsson,R. E. Chance, and N. Moon. 1976. Pancreatic-polypeptideresponse to food in duodenal-ulcer patients before andafter vagotomy. Lantcet. 1: 1102-1105.

11. Floyd, J. C., S. S. Fajans, and S. Pek. 1976. Regulationin healthy subjects of the secretion of human pancreaticpolypeptide, a newly recognized pancreatic islet polypep-tide. Trans. Am. Physicians. 89: 146-158.

12. Adrian, T. E., S. R. Bloom, M. G. Bryant, J. M. Polak, P. H.Heitz, and A. J. Barnes. 1976. Distribution and releaseof human pancreatic polypeptide. Gut. 17: 940-944.

13. Floyd, J. C., S. S. Fajans, S. Pek, and R. E. Chance. 1977.A newly recognized pancreatic polypeptide; plasma lev-els in health and disease. Recent. Prog. Horm. Res. 33:519-570.

14. Stadil, F., J. Malmstrom, J. F. Rehfeld, and M. Miyata.1974. Effect of atropine on hypoglycemic release of gas-trin in man. Acta Physiol. Scand. 92: 391-398.

15. Stadil, F., and J. F. Rehfeld. 1974. Gastrin response toinsulin after selective, highly selective, and truncal vagot-omy. Gastroenterology. 66: 7-15.

16. Lindkeer Jensen, S., C. Kuhl, 0. Vagn Nielsen, and J. J.Holst. 1976. Isolation and perfusion of the porcine pan-creas. Scand. J. Gastroent. 37 (Suppl.): 57-61.

17. Siegel, S. 1956. Nonparametric statistics for the behavioralsciences. McGraw-Hill Book Co., New York. 1.

788 Schwartz et al.

18. Adrian, T. E., S. R. Bloom, H. S. Besterman, A. J. Barnes,T. J. C. Cooke, R. C. G. Russell, and R. G. Faber. 1977.Mechanism of pancreatic polypeptide release in man.Lancet. 1: 161-163.

19. Innes, I. R., and M. Nickerson. 1975. Atropine, scopol-amine, and related antimuscarinic drugs. In The pharma-cological basis of therapeutics. L. S. Goodman and A.Gilman, Editors. Macmillan, Inc., New York. 516-522.

20. Schaffalitzky de Muckadell, 0. B., J. Fahrenkrug and J. J.Holst. 1977. Release of vasoactive intestinal polypeptide(VIP) by electrical stimulation of the vagal nerves. Gas-troenterology. 72: 373-375.

21. Hillarp, N. A. 1960. Peripheral autonomic mechanisms.In Handbook of Physiology. Am. Physiol. Soc. Washing-ton. Sec. 1. Vol. 2: 989-990.

22. Sjodin, L. 1975. Gastric acid responses to graded vagalstimulation in the anaesthetized cat. Digestion. 12: 17-24.

23. Liedberg, G., K. C. Nielsen, C. H. Owman, and N.-O.Sjoberg. 1973. Adrenergic contribution to the abdominalvagus nerves in the cat. Scand. J. Gastroent. 8: 177-180.

24. Said, S. I., and R. N. Rosenberg, 1976. Vasoactive Intes-tinal Polypeptide: Abundant immunoreactivity in neural

cell lines and normal nervous tissue. Scienice (WXVashi.D. C.). 192: 907-908.

25. Uvrn.s-WVallensten, K., J. F. Rehfeld, L.-I. Larsson, acndcB. Uvnas. 1977. Heptadecapeptide gastrin in the vagalnerve. Proc. Natl. Acad. Sci. In press.

26. Schwartz, T. W., and J. F. Rehfeld. 1977. NIechanism ofpancreatic-polypeptide release. La icet I: 697-698.

27. Clark, A. J. 1927. The reaction between acetylcholine andmuscle cells. Part II.J. Physiol. 64: 123-143.

28. Loubatieres-Mariani, NI. MI., J. Chapel, R. Alric, and A.Loubatieres. 1973. Studies of the cholinergic receptors in-volved in the secretion of insulin using isolated perfusedrat pancreas. Diabetologia. 9: 439-446.

29. Taylor, I. L., M. Impicciatore, and J. H. W7alsh. 1977.Effect of atropine and vagotomy on the pancreatic poly-peptide response to a meal. Gastoenterology. 72: A-116-1139.

30. Bloom, S. R., and T. E. Adrian. 1977. Mechanism of pan-creatic polypeptide release. Lancet. I: 698.

31. Schwartz, T. W. B. Steenquist, and L. Olbe. 1977. Vagalregulation of PP secretion. In Proceeding from the gut hor-mone symposium, Lausanne, 18-19th June 1977. S. R.Bloom Editor. Churchill Livingston, Edinburgh. In press.

Vagal, Cholinergic Regulationi of Patncreatic Polypeptide Secretion 7789)