Insulin secretion stimulated by allogeneic lymphocytes in an inbred strain of mice

Insulin secretion stimulated by allogeneic

lymphocytes in an inbred strain of mice.

J B García, … , O H Pivetta, J C Basabe

J Clin Invest.

1986;

77(5)

:1453-1459.

https://doi.org/10.1172/JCI112457

.

Effects of intraperitoneal injection of allogeneic lymphocytes on insulin secretion were

studied in incubated pancreas slices from BALB/c mice. Injection of allogeneic lymphocytes

from C57BL/6J (H2b) mice increased insulin secretion, both in basal and 11-mM

glucose-stimulated conditions. This effect was only present when at least 5 X 10(6) or 1 X 10(6) cells

were injected (in basal and stimulated conditions, respectively). Glucose-induced insulin

secretion (3.3-27.5 mM) was significantly increased in pancreata from mice injected with

allogeneic lymphocytes. No effect was observed when glucose was not included in the

incubation medium. Intraperitoneal injection of Dextran 70 produced no change in

glucose-elicited insulin secretion. There were no differences in glucagon and somatostatin (SRIF)

secretion obtained from pancreas of mice injected with allogeneic or syngeneic

lymphocytes. Injection of allogeneic cells increases insulin secretion (basal and both

phases of 11 mM glucose-stimulated secretion). Puromycin significantly inhibited the

second phase of insulin secretion. These results suggest that: Injection of allogeneic

lymphocytes raises both basal and glucose-stimulated insulin secretion. This effect seems

to be connected with the major histocompatibility complex, and to be related to the number

of allogeneic cells injected. Injection of allogeneic lymphocytes seems to sensitize the beta

cell response to glucose stimulus. Neither glucagon nor SRIF secretion are altered by

alloantigen injection. The stimulatory effect of allogeneic lymphocytes is related, at least in

[…]

Research Article

Insulin Secretion Stimulated

by

Allogeneic Lymphocytes

in

an

Inbred Strain

of Mice

J.B. Garcia, M.C. Venturino, E. Alvarez, L.FabianodeBruno, M. Braun,0. H.Pivetta,and J.C.Basabe

Fundacion Laboratorios deInvestigacionesPedidtricas, Hospital General deNifios"Dr.Pedro de Elizalde,"Centro de Estudios Farmacol6gicosydePrincipiosMaturales, andInstitutoNacional de Genetica Medica,Buenos Aires, Argentina

Abstract

Effects of intraperitoneal injection of allogeneiclymphocyteson insulin secretion were studied in incubated pancreas slices from BALB/c mice. Injection of allogeneic lymphocytes from C57BL/ 6J (H2b) mice increased insulin secretion, both in basal and 11-mMglucose-stimulated conditions. This effect was only present when atleast 5 X 106 orI X 10' cells were injected (in basal

andstimulated conditions, respectively). Glucose-induced insulin secretion(33-27.5 mM) was significantly increased in pancreata from mice injected with allogeneic lymphocytes. No effectwas observed when glucose was not included in the incubation me-dium. Intraperitoneal injection of Dextran 70 produced no change in glucose-elicited insulin secretion. There were no differences in glucagon and somatostatin (SRIF) secretion obtained from pancreas of mice injected with allogeneic or syngeneic lympho-cytes. Injection of allogeneic cells increases insulin secretion (basal and both phases of 11 mM glucose-stimulated secretion). Puromycin significantly inhibited the second phase of insulin

secretion.Theseresults suggest that: (a) Injection of allogeneic

lymphocytes raises both basal and glucose-stimulated insulin secretion.(b) This effect seems to be connected with the major histocompatibility complex, and to be related to the number of allogeneic cells injected. (c) Injection of allogeneic lymphocytes

seemstosensitize the beta cell response to glucose stimulus. (d) Neither glucagon nor SRIF secretion are altered by alloantigen injection. (e) The stimulatory effect of allogeneic lymphocytes is related, at least in part, to insulin synthesis.

Introduction

Interaction between the endocrine and immune system is com-plex and the study of potentially critical factors needs multidis-ciplinary analysis.

Numerousfindings support the existence of regulatory cir-cuits between theimmuneand endocrine systems. Antigen

in-jectionintomiceinducesarapid increase ingonadotropinsblood levels, the pharmacological suppression ofwhich results in a blockadeof antibodysynthesisandspecifictolerance(1).

RestingTlymphocytes have noinsulinreceptors, but they

developthemdenovoafter nonspecificmitogen stimulation by

Concanavalin A (2, 3). Insulin provided to receptor-bearing lymphocytesincreases intermediary metabolism above

mitogen-Addressreprintrequests to Dr. Basabe,FundacionFLIP, Montes deOca 40 (1270),BuenosAires, Argentina.

Receivedfor publicationIISeptember 1984 and in revisedform 31 October 1985.

stimulated levels (4). Furthermore, insulinaswell asother

li-gands, which increase intracytoplasmatic cyclic guanosine monophosphate, augment lymphocyte-mediated cytotoxicity (5). Thesefactorsstronglysuggestaphysiologicalrole for insulin in

immunoregulation.

In a previoussetofexperiments, we found thatsheep red

blood cell injectioninto inbred mice caused an increase in insulin

secretion fromincubatedpancreasslices(authors' unpublished data).

Thefollowing studywasdesignedtoinvestigatetheresponse of mice endocrine pancreastoalloantigen injection.

Methods

Animal

model

We used6-9-wk-old male and female, fed inbred BALB/c,C57BL/KsJ,

andC57BL/6J miceprovided by the Instituto Nacional deGenetica M&dica(BuenosAires).

Preparation

ofcell

suspensions

Cellsuspensionswereprepared bytwomethods:(a) The tissue(spleen)

wassubmittedto asoft mechanical disruption in cold saline solution for homogenization; itwaswashed threetimes,andconcentrationwas ad-justedto50X 106cellspermilliliter, (b)Inotherexperiments spleen lymphocyteswerepurified by centrifugation inaFicoll-Hypaquegradient.

No differences were observed between the results obtained by both methods. (Insulin secretion,ng/100mgwettissueper20minof incu-bation: Basal, syngeneiclymphocytes, method(a) 4.1±0.8vs.method

(b)4.0±0.7; allogeneiclymphocytes,(a) 13.2±0.6vs.(b) 12.5±0.5; stim-ulated,syngeneic lymphocytes,(a) 62±3.5 vs.(b) 61.3±5.2; allogeneic lymphocytes,(a) 133.2±10.4vs.(b) 114.4±4.6).Therefore,weused the first method described in all furtherexperiments.

Experimental model

Antigen injection. 10' spleencellsfrom BALB/c(H2d), C57BL/KsJor C57BL/6J (H2b)miceresuspendedinavolume of 0.2 mlwereinjected intraperitoneally into BALB/c mice. The animalswerekilled 30, 60,

120, and 300minafterinjection and each mouse'spancreaswasremoved in orderto beincubatedasspecifiedin"Incubationtechnique,"below.

In a separate set ofexperiments, Dextran T70 (Pharmacia Fine

Chemicals,Piscataway, NJ),a conventional solubleantigen,at a

con-centration of 50or500,gdiluted in 0.2mlsaline,wasinjected

intra-peritoneally into BALB/c mice. The animalswerekilled 90 minafter injection and each animal'spancreaswasremovedtobeincubatedas

discussed in"Dynamicsof insulinsecretion,"below.

Bidirectionallymphocytesculture(BALB/cvs. C3H).Purified lym-phocytessuspensionswereprepared from spleen cells by Ficoll-Hypaque gradient.Afterwashing, theywereresuspendedat aconcentration of I0' cells/ml in RPMI 1640 with5% fetal calfserumandantibiotics, and

culturedinplasticmignontubesat37'C for different timeperiods (Table I). Replicatecultureswere pooled, centrifugatedat 1,500 rpm for 10 min, and the supernatantswereusedwithout furthermanipulations.

Dynamicsofinsulinsecretion. 107spleen cells fromBALB/corfrom C57BL/6J micewereinjected intraperitoneallyintoBALB/cmice. The micewerekilled 90minafter injection andpancreatawereremoved for

perifusion.

J.Clin.Invest.

© The AmericanSocietyforClinicalInvestigation,Inc.

0021-9738/86/05/1453/07 $1.00

Table I. Insulin Secretion Stimulated by Supernatant of Bidirectional Lymphocytes Culture

Basal glucose, 3.3 mM Stimulatedglucose, 11 mM Culture

Experimental Control time Experimental Control

h

5.5±0.5 5.6±0.8 0.5 48.3±5.7 39.1±1.8

5.4±0.5 5.1±0.3 1 57.6±5.0 54.0±2.2

7.0±0.4 5.5±0.4 3 96.2±2.8* 54.3±4.5

5.7±0.6 4.6±0.4 6 88.7±5.8* 53.6±1.8

4.5±0.3 4.5±0.3 12 54.4±7.2 45.9±3.6

Insulinsecretion(ng/I00mg w.t./20min, X±SEM)produced by pan-creasfromBALB/cmice incubated inpresenceofsupernatantof bi-directional lymphocyteculture(BALB/c vs.C3H) diluted 1/10 (exper-imental). Control pancreata were incubated in the presence of super-natantofBALB/c orC3Hlymphocytesculture.

*P <0.001;notedifferencesbetweenexperimentaland controlin stimulated Il-mMglucose.

Incubation technique. The incubation technique described by Martin

and Bambers (6) was employed.After killingtheanimals bydecapitation,

weexcisedthe splenicportion of each mouse's pancreas, and removed

the adiposetissue, lymphatic vessels, and nodules. Slices of pancreas weretransferred into flasks containing 2 ml of incubation buffer and werepreincubated for 15minat4VCin an ice-cold water bath under constantgassing withamixture of95%02and 5%CO2.The slices were

transferred intonewflasks withfresh bufferandincubated for 15min

inaDubnoff incubator (recuperation period).Then the slices were

trans-ferred into flasks with fresh buffer and incubated for 20min at370C (basalperiod).Thepancreatic tissuewasfinally transferred intoafourth

groupof flasksandincubated for20 min at370Cin the presenceof 11 mMglucose asstimulus for insulinsecretion(stimulatedperiod).

Theincubation mediumwasKrebs-Ringer bicarbonate(pH, 7.38-7.42) supplemented with 60mg%glucose, Ig%bovineserumalbumin,

and1,000 KIU/mlproteaseinhibitor Trasylol.

Perifusion technique. The technique described by Burr et al. (7) was

used. Theperifusion mediumand theprocedure toobtaintheslices of

pancreas were similar to those described above. Glucose, 11mM, used asstimulus for insulin secretion,wasadded to the perifusion medium at3min. Theproteolytic effecton the hormonalsecretionwas avoided

byusing Trasylol inthebufferand bycollecting the samples on 0.25 M EDTAintubes at4°C,which wereimmediately frozen at -20°C.

Hormonal

dosage

Insulin.Insulinwasdeterminedbythe method of Herbertetal.(8) mod-ified by Crestoetal.(9).Theporkmonoiodine insulin of

high

specific

activitywaslabeled with

1251,

separated bymeansofastarch

preparative

gel chromatography,andpurified bypassagethrougha_SephadexG-100

column. The ratstandard insulin wasobtained from Novo Research Laboratories(Denmark). The anti-insulinantibodyprovedtobe suffi-ciently"nonspecific"todisplacethe labeledinsulin b both therat stan-dard andourexperimentalinsulin. This methodalloweddetermination within0.4-32ng/mlrange.

Glucagon.GlucagonwasdeterminedbytheSerono BiodataGlucagon

polyethylene glycol (PEG). Kit, Hypolab S.A.(Switzerland),batches 10727

and 10810. With minor modifications in the antiserum dilution,this method alloweddetermination withina 15-1,000ng/mlrange.

Somatostatin

(SRIF).'

Somatostatinwasassayedwith thetechnique

1.Abbreviations used in thispaper:SRIF,somatostatin;Tyr,

tyrosinyl.

of Arimuraetal.(10).Antisera1650against SRIF,SRIFstandard,and

tyrosinyl

(Tyr')-SRIF

were obtained from UCB Bioproducts,

Peptide

Department (Bruxelles, Belgium).Thesensitivityoftheassay,calculated astheleastamount ofSRIFcausing significant displacementsof '25I-Tyr'-SRIFfrom antibody (P<0.005),was 1.5-2.0pg/tube;serial di-lutionswereparalleltothe standardcurve. Neitherinsulin, glucagon, pancreatic polypeptide,norgastrinused in the standardcurve

_displaced

'25I-Tyr'-SRIF

from the antiserum underradioimmunoassayconditions. Statistics

One-wayanalysisof varianceorStudent'sttestwereused

(1 1).

Results

Intraperitoneal

injection

of

C57BL/6J

(H2b)

lymphocytes

into

BALB/c

(H2d)

mice

_{significantly}

increased insulin

secretion,

both in basal

(3.3

mM

_glucose)

and stimulated

(11

mM

glucose)

conditions.Thisrisewasobservedwhenthemiceweresacrificed

30, 60,

and 120 minafter

injection,

but not when

_they

were

sacrificed 300 min after

injection

(Fig. 1).

Thesameeffectwas obtained when

allogeneic lymphocytes

were

injected

intrave-nously

(insulin

as

ng/

100mgwettissueper20 min

incubation;

(a)

basalconditions:

allogeneic

injected

mice,

8.2±0.5vs.

syn-geneic injected mice, 4.0±0.4,

P<

0.001;

(b)

stimulated

con-ditions:

allogeneic

injected

mice,

106±8.7vs.

_{syngeneic injected}

mice,

49.8±8.8,

P<

_0.01;

n = 5 in all

groups).

Furthermore,a

significant

increasein insulin secretioninduced

_by

intraperito-neal

injection

of

_{allogeneic lymphocytes,}

was

_only

observed when atleast

5)X

106 cells

_{(basal conditions)}

or 1 X 106 cells

(glucose-stimulated

_conditions)

were

_injected

(Fig. 2).

Itcanalsobeseenin

Fig.

_1,

that when

BALB/c

(H2d)

mice

were

injected

with

lymphocytes

from

C57BL/KsJ

(H2d) mice,

therewas no

change

in theamountof insulinsecreted. TableIshowsthat thesupernatant of bidirectional

lympho-cytes

cultures,

obtainedat3 and 6

h,

enhances

_{glucose-stimulated}

insulinsecretion from

BALB/c

incubated micepancreas. Nodifferencesin

glycemic

levels between animals

intraper-itoneally injected

with

_allogeneic

or

syngeneic lymphocytes

were

observed

(glycemic

levels

expressed

as

mg/dl,

90 min after

in-jection:

(a)

allogeneic

lymphocyte-injected

mice: 113±8vs.

(b)

syngeneic

lymphocyte-injected

mice,

137±1

1,

P

_NS).

In vivo

peripheral

insulin levels of

_{allogeneic lymphocyte}

intraperitoneally injected

mice

(3.9±0.2 ng/ml)

showedno

dif-ferences when

compared

with

syngeneic

lymphocyte-injected

animals

(3.6±0.2 ng/ml).

Table IIshows that Dextran

70,

injected

intraperitoneally,

50or_{500 ,g in} 0.2ml

_saline,

did not

_modify

eitherthe basal orthe

glucose-stimulated

insulin secretion.

Theamountof insulinsecretion

produced

by

different

glu-coseconcentrations obtained from the pancreas of animals

in-jected

with

syngeneic

or

allogeneic

lymphocytes,

canbeseenin Table III. Inboth groups insulin levelswere

significantly higher

than control values

(3.3

mM

glucose),

and the levels attained were

higher

as

glucose

concentration increased

(but

only

until 16.6 mM

glucose,

since therewasa

plateau

in the responseto

16.6 and 27.5mM

glucose

in both

groups).

Ifthe results obtained in TableIIIare

expressed

intermsofadose responserelation

between

glucose

concentration in theincubationmedium and insulin response

levels,

the

injection

of

allogeneic lymphocytes

Xp 0.01

xx pxO.001

ii

TIME: 30'

BASAL PE

xx

xx

7-I

.,

100.

lF~t-1

Figure 1.Insulinsecretion (X±SE)

L4C ANS. inbasalconditions, 3.3mMglucose

(A)andstimulated conditions(B)

so from incubatedslicesbelongingto

BALB/c mice, injectedwithC57BL/

Nss s s s KsJ, BALB/clymphocytes, or with

salinesolution; 30, 60, 120,and 300

666! 5fiL6ffi1|1 | 19 minbefore theexperiment(n = 6in

60' 120' 300' TIME: 30, 60' 120' 300' eachgroup). Injectedwith: X,saline;

X, BALB/c (H2d); ,C57BL/KsJ

ERIOD (Glucose 3.3mM) STIMULATED PERIOD (Glucose 11 mM) (H2d);O,C57BL/6J (H2b).

Table III also shows that the stimulatory effect ofalloantigen injection on insulin secretion wasnot observedin a medium

containing0 mMglucose.

In our experimental conditions allogeneic lymphocyte

in-jection didnotmodify glucagon secretion, eitherinbasalorin

both phases of5.5 mM glucose/20 mM arginine-stimulated conditions,as canbe observedinTable IV (thepointatwhich

the maximal valuewasattained,wasconsidered representative

ofthe first phase of glucagon release, generallyat5 min of peri-fusion;and the valueoccurringat 40minas representative of

thesecondphase).

Somatostatin secretionpattern, both inbasal and

glucose-stimulated conditions, wasnotaltered by the injection of

allo-geneic cells (Fig. 3).

Fig. 4showsinsulin secretion patternsstimulated with 11 mMglucose from perifused slices ofpancreasofBALB/cmice

injected withallogeneicorsyngeneiclymphocytes. Injectionof allogeneic lymphocytes inducedapatternofbasal and stimulated hormoneresponse,significantly higher than control.(Areaunder

thecurveinnanogramsof secreted insulinper 100mgof

pan-creatictissue,X+SEM;pancreasof mice injected with allogeneic

lymphocytesvs.pancreasof mice injected with syngeneic

lym-phocytes. Basal: minutes 1-3, 9.8±0.5 vs. 5.2±0.5,P<0.001.

160

80

0 -20

Z

z

_,___S

LYMPHOCYTES

C4

C4

First phase:minutes 4-8,42±0.5vs.25±0.6,P<0.001. Second

phase: minutes 10-40, 447±13vs.325±16, P<0.001).

Insome groupsPuromycin, 80

,g/ml,

wasaddedtothe

pen-fusion medium. In allcasesthis drugwasperifused during30

min immediately before sample collection, andwasthen

peri-fusedfor 40 minineachexperiment.Attheconcentration used, Puromycin inhibits 95-99% of ['4Cjvaline incorporation into pancreatic proteins.

The second phase of insulin secretion induced by 11 mM

glucose inpancreasfrom mice injected with allogeneic cells, is

clearly inhibited byPuromycin (Fig. 5). (Area underthecurve

innanogramsof secretedinsulinper100mgofpancreatic tissue,

X±SEM; withoutPuromycin, 448±12; with Puromycin, 307±3,

P<0.001). In thiscasetheamountandpattern ofinsulin

se-cretionweresimilartothoseobtainedwhenpancreatafrom mice

injected with syngeneic lymphocytes were stimulated with 11 mMglucose inabsenceof Puromycin. (Area under thecurve,

325±13ng/100mgpancreatic tissue).

At5mMconcentration,glucosecausedonlyonephase in-sulinsecretion. Injection ofalloantigeniccellsincreasedthis first

phaseof insulinsecretionandclearlyinducedabiphasic secretion patternwitharemarkablepotentiationofthe secondphase (Fig.

6;P<0.001 from 10to40min.Areaunder thecurvein

nano-11mM.glu.

2BmM.glu.

Figure 2. Insulin secretioninducedby 11 mM(top)or

5.5mMglucose (bottom)frompancreaticslices be-longingtoBALB/cmiceinjected90minbefore the experimentinpresence of different concentrations of allogeneic (o)orsyngeneic (-) lymphocytes (X±SEM,

n =6ineachpoint;P<0.001).

10-S.

-S

0

E

0 0

'I,

o

z

N.S.

(P)

xx

I -1 xx

F-I

Table II. Dextran 70 Injection and Insulin Secretion

Insulin secretion

Injection Basalperiod Stimulatedperiod

ng/J00mgw.t./20 min ng/100mgw.t./20 min

Dextran 50

_Ag

5.2±0.5 51.1±4.8

Dextran500,gg 4.5±0.4 59.3±6.9

Saline 5.3±0.2 58.6±6.0

Effect of intraperitoneal Dextran 70 injection on the basal and 1-mM glucose-induced insulin secretion from incubated pancreas of inbred BALB/c mice (n = 5 in each group). w.t., wet tissue.

grams of secreted insulin per 100 mg of pancreatic tissue,

X+SEM;

pancreasof miceinjected with syngeneic lymphocytes, 59±3; injected withallogeneic lymphocytes, 177±4). Puromycin significantly inhibits the second phase of insulin secretion (P

<0.001) from 20 to 40 min perifusion. However, the "residual" secondphase pattern wasstillsignificantly higherthanthe one obtained from pancreata of mice injected withsyngeneic cells

(areas under the curve, 147+7.0 and 59±2.7 ng/100 mg pan-creatictissue,respectively).

Discussion

Ourfindings show that endocrine pancreas can respond to

an-tigenic stimulation with an increase in insulin secretion. This rise has already been observed after 30 min of antigenic

intra-peritonealinjection. It remains high even after2h, butdisappears after5h.Thishormonal response toantigenicstimulation does

not seemtobeexclusive of pancreas,since Pierpaoli and Maes-troni (1, 12, 13)foundanincrease infollicle-stimulatinghormone and luteinizing hormone plasmatic levels, 2 hafter allogeneic injection. This shows a response of the hypothalamo-hypophy-sealaxisto animmune aggression.

Our results related to hormonal response to antigen stimu-lation were characterized by the following findings: (a) When

0

25C C

_E E L L:

I

0 20

TIME (min.)

Table III. Insulin Secretion Induced byDifferentGlucose Concentration

Insulinsecretion

Allogeneiclymphocyte Syngeneic lymphocyte

Glucose injection injection

mM ng/100 mg w.1./20min ng/100 mg w.t./20min

0 8.32± 1.08 6.76±0.56

3.3 14.04±0.64 6.36±0.66

5.5 82.64±4.12 41.84±3.80

11 105.84±1.72 43.68±8.68

14.4 122.36±2.13 50.04±3.28

16.6 138.92±4.44 99.60±6.44

27.5 141.96±3.04 101.56±4.16

Insulin secretion induced by different glucose concentrations from in-cubated pancreas slices of BALB/c mice sacrificed 90 min after synge-neic orallogeneic lymphocyte injection (mean±SEM; n = 6 in each group). w.t., wet tissue.

BALB/c mice were injected withallogeneic lymphocytes from C57BL/KsJ mice (which have thesameH2 complexasthe re-ceptorbut inadifferent genetic background),ourexperimental

model showed no increase in insulin secretion. Thissuggests that the increasecouldbe related to themajorhistocompatibility complex. (b) The stimulatory effect of alloantigen on insulin

secretion was related to the number oflymphocytes injected. (c) Injection of a conventional soluble antigen, such as Dextran T70, didnotproduce any changein the insulin secretion stim-ulated by glucose. (d) Injectionofallogeneiccells did notinduce

any increasein theinsulinrelease in absenceofglucose inthe incubation medium. (e)Theresponsethatdepicts the

interre-lation betweendifferentglucoseconcentrations and insulin se-cretion was shifted to the left due to theinjectionofallogeneic lymphocytes. Thiscould indicate thatalloantigensensitizedthe

beta cell toglucose stimulation.

The fact that allogeneic lymphocyte injectedmice did not

produce any effect on peripheral insulin levels is surprising. However,consider that in determinedconditions,insulin

stim-1

Figure3.Somatostatin secretion pattern induced by 11 mMglucosefrom perifused pancreas slices, belong-ing to BALB/c mice injected 90 min before with allo-40 geneic (--e--)orsyngeneic(-*-) lymphocytes

TableIV.EffectofStimulated AllogeneicandSyngeneicCells

InjectiononGlucagonSecretion

First phase, Secondphase, Basal,minuteI minute5 minute 40 Allog. Sing. Allog. Sing. Allog. Sing.

Mean 163 160 1,355 1,322 1,267 1,137

SEM 7 14 69 82 39 57

2

10 20

TIME (minutes)

Figure 4.Insulinsecretion pattern, stimulated by 11mM glucose,

from perifusedpancreasslices belonging to BALB/c mice injected 90 minbefore withallogeneic(C57BL/6J)orsyngeneic (BALB/c)

lym-phocytes(X±SEM,n=5).

ulus canincrement portal hormonal levels with little or no change

inperipheralinsulin concentration (14, 15). In our experimental arrangement, peripheral insulin levels were slightly altered but notenough (P NS) toinduce any appreciable change on glycemic concentration.

Reports on the immunohistochemistry and electron mi-croscopyof the endocrine pancreas disclosed connections among a,

fi,

andi cells(16). On the basis of this and other reports, the

existence of interregulationamonginsulin,glucagon, andSRIF has been postulated (16-18). This part of the study was devised

keeping inmindthefundamental role of the insulin-glucagon

Glucagon,pg/min per100mgwettissue.

Glucagon secretioninducedby 20/5.5mMarginine/glucosefrom

perifused pancreasslices ofBALB/c mice injected 90minbeforewith

allogeneic (Allog.)orsyngeneic(Sing.) cells (mean±SEM; n =5).

Areaunder thecurve:(minutes 3to8, pg/100mgw.t.)allogeneic lym-phocytes 4,719±272 vs. syngeneic lymlym-phocytes, 5,304±353;P,NS.

balance in metabolic control and the fact that SRIF regulates

the secretion ofboth hormones. Inourexperimentalset,injection

ofallogeneic lymphocytes did not modify either glucagon or SRIF secretion. Thissuggeststhatthe stimulatory effect of al-loantigen injectionactsonlyonpancreatic data cell secretion.

Constantpancreas stimulus with 11 mM glucose produces abiphasic insulin secretionpattern. Thefirst phase represents

the exocytosis of granules formed and aligned in thesystem of microtubules and microfilaments of the beta cell. The second phase, which is slower andmoreprolonged, is the result of the release of both accumulated granules and granules with insulin synthetized denovo(19). Stimulation withallogeneic

lympho-cytesproducedanincrease in both insulin secretion phases. This seems to indicate that stimulusacts on boththe fast and the slowinsulin releasing pools. The second phase ofglucose-induced

insulin secretioncanbepartially inhibited (20-30%) by protein synthesisinhibitors. For this reason it has beenpostulated that this provisionary phase is relatedtoprecursorsynthesis ( 19-21). Since injection of allogeneic lymphocytes sensitized the

sec-ond phase ofglucose-induced insulin secretion, the question that arises is whether this effect might be relatedtoinsulin synthesis

ornot.Ascanbeseenin Figs. 5 and 6, Puromycin addedtothe perifusion medium significantly inhibited the second phaseof theglucose-stimulated insulin secretion. Although indirectly, this

Figure 5. EffectofPuromycin(opencircles)on in-sulin secretioninduced by 11 mMglucosefrom perifusedpancreasofmiceinjected 90min before

40 with allogeneic (dotted lines) or with syngeneic

(fulllines) lymphocytes(n=5in each group).

2 E

0-0

I

20-

15-oi

i

E a o 10

z

z

5.

o

.I-

1-I

*-eC57BL/6J

o- -o BALB/C

30

I_E

Alo

i

t 0

21

a,1

a)

Z

l)

0 U

0

E

F

20

Time

0

oi

E

I1

Figure 6.Effect of Puromycin(solidcircles)onthe insulinsecretion induced by 5.5 mM glucose, from perifusedpancreasof miceinjected 90minbefore with allogeneic (dotted lines) or syngeneic(full lines)lymphocytes (n=5 in eachgroup).

showed that the increased sensitivity ofthe beta celltoglucose

induced byallogeneiclymphocyteswasrelated,atleastinpart,

toinsulinsynthesis.

Glucose ata5.5-mMconcentration causedonlyonephase

insulin secretion. Alloantigen injection, in turn,modified not onlytheamountbut also thepatternof 5.5mMglucose-induced hormone secretion, substitutinga monophasic by a biphasic

profile. Incidentally,the change intheinsulinsecretionpattern caused by the injection of allogeneic lymphocytes, where 5.5 mM glucose is usedasstimulus,wassimilartothatelicitedby

glibenclamide, sodium salicylate,ortheophylline,all substances

with previously reported stimulatory effectsoncyclicAMP

pan-creaticlevels (22-24). Althoughanycomment ortentative

ex-planation would be, at'present, only speculative, these results

couldsuggest thatthe increase in beta cell sensitivitytoglucose

causedby alloantigen injection mightbemediated, atleastin

part,byaneffectoncyclicAMPlevels.

Insulin secretion by thepancreaticbetacell is regulated in

bothanimmediate(nutrients, neurotransmitters, hormones) and

achronic(ontogenic, endocrine,andnutritional factors) fashion;

that isto say, there isa multifactorial regulation for asingle

processofsecretion (25, 26).Our resultssuggestthat theimmune

system should be consideredamongtheinsulinsecretion

regu-latory factors.Uptonow wedonotknow thecharacteristics of

thishormonalresponsetoantigenic stimulation.Itmaybe that

theimmuneinteractionproduces the releaseofsomelymphokine

or interleukin capable ofstimulating the beta cell. Although

thedata obtained with thesupernatants of bidirectional

lym-phocytes culturesseemto supportadirect effectonthepancreas,

theydonotruleoutthatallogeneiclymphocytes might also in-ducethesynthesisorrelease of nonspecific factor(s) from

non-lymphoidtissues'that increase insulin secretion.

Regardless ofwhether these speculationsprovecorrect,our

study supplies newdata thatsupporttheexistenceofan

inter-relationbetweeninsulinand the immunesystem.

Acknowledgments

We wishtothank Miss P. I.Castellaniand Mrs. M.M. de Cometfor excellenttechnical assistance.WeareindebtedtoDr.R.Gutman from

theEndocrinology Unit ofItalianHospital for theantiinsulinserum,

andtoBayerArgentinafor thesupplyofTrasylol.

This workwassupportedin partbygrants8458/80fromtheConsejo

Nacionel deInvestigacionesCientificasyT6cnicasand 10235/82from

SubsecretariadeCienciayTecnica. Thispaperwasalsosupportedin partbythe Latin American Insulin Grant(Nordisk, Denmark)andby FundacionAlbertoJ. Roemmers.

References

1. Pierpaoli, W., andG. J. M. Maestroni. 1977. Pharmacological controlof the immune responseby blockade of theearly hormonal changes following antigen injection.CellImmunol. 31:355-363.

2.Helderman, J. H.,and T. B. Strom. 1977.Emergenceof insulin

receptorsuponalloimmune T cellsinthe rat. J. Clin.Invest. 59:338-344.

3. Gozes, Y., J.Caruso, and T. B. Strom. 1981. The absence of

crypticinsulin receptorsonrestingTlymphocytes.Diabetes.

30:314-316.

4. Helderman, J. H., andL.Edwards. 1981. Role of insulin in the

intermediary metabolismofactivatedthymic-derivedlymphocyte.J. Clin.

Invest. 67:1636-1642.

5. Strom, T. B.,R. A. Bear, and C. B. Carpenter. 1975. Insulin-induced augmentation oflymphocyte mediated citotoxicity. Science (Wash. DC). 187:1206-1208.

6. Martin,J. M.,and G. Bambers. 1965. Insulin secretion in glu-cosamineinducedhyperglycemiainrats. Am. J.Physiol.209:797-803.

7. Burr,I.M.' L.Balant,W.Stauffacher,andA. E. Renold. 1971.

Adrenergicmodifications ofglucoseinducedbiphasicinsulin release from perifusedrat pancreas.Eur. J. Clin.Invest. 1:216-224.

8. Herbert, V., K.Lau, C. W.Gottlieb, and S. J. Bleicher. 1965.

Coated charcoalimmunoassayofinsulin.J. Clin. Endocrinol. Metab.

25:1375-41384.

9.Cresto,J.C.,I. Dujovne,P. J. Castellani,E. A.Mitta, S. F. de

Majo,and V. G.Foglia. 1972. Insulinradioimmunoassay bythe charcoal dextrantechnique. Diabetologia.8:292-295.

10.Arimura, A., H.Sato,D. Coy,and A.Schally. 1975.

Radioim-munoassayofGH-releaseinhibitinghormone. Proc. Soc.Exp.Biol.Med. 148:784-789.

11.Rattcliffe,J. T.1969.Elements of mathematical statistics.Oxford University Press,London. 120-132.

12.Pierpaoli, W., and G. J. Maestroni.1978.Pharmacologicalcontrol of thehormonallymodulated immuneresponse.II.Blockadeofantibody

production byacombinationofdrugs actingonneuroendocrinefunction.

10

c

i

Es 0

0

I01

Z

.a

44 n

0 20

Time

Itspreventionbygonadotrophins andcorticotrophin.ImmunoL 34:419-430.

13.Pierpaoli,W., and G. J. Maestroni. 1978. Pharmacological control

of the hormonally modulated immuneresponse. III. Prolongation of

allogeneic skin graftrejection andprevention ofrunt disease by a

com-binationofdrugsactingonneuroendocrine functions.J.Immunol.120:

1600-1603.

14. Porta, D., Jr., L. Girardier, J. Seydoux, Y. Kanasawa, and J. Pasternak. 1973. Neural regulation of insulin secretion in the dog. J.

Clin.Invest. 52:210-214.

15.Chieri,R., J. M. S.Farina,J.Halperin, and J. C. Basabe. 1975.

Effectof cephalic glucoseinfusiononinsulin secretion. Diabetologia.

11:1-6.

16.Orci, L., and R. H. Unger. 1975. Functional subdivision of islets

ofLangerhans andpossible role ofD-cells. Lancetii:1243-1244. 17,Itoh,H., L.Mandarino,and L.Gerich. 1980.Antisomatostatin

gammaglobulin augmentssecretion ofbothinsulin andglucagon in

vitro. Diabetes. 29:693-696.

18.Basabe, J. C.,0.H.Pivetta,L.Fabiano de Bruno, J. C. Cresto, andN. Aparicio. 1983. A temporal study of somatostatin secretion and

its inhibitory effectingeneticallydiabetic mice(C57BL/KsJ-db/db).

En-docrinology 113:1927-1934.

19.Lacy, P. E. 1970. Beta cell secretion.Fromthestandpoint ofa patholobiologist. The Banting Memorial Lecture 1970. Diabetes. 19: 895-905.

20. Curry,D.L.,L.Bennett,andG. M.Grodsky. 1968. Dynamic of insulin secretion bytheperifused ratpancreas. Endocrinology. 83:

572-584.

21.Basabe,J.C.,N.L6pez,J.Viktora, andF.Wolff. 1970. Studies

oninsulin secretion in the perifusedratpancreas:effect of diazoxide and A025. Diabetes. 19:271-281.

22. Kuo, W. H., D.Hodgkins, andF. Kuo.1973.Adenylatecyclase inislets of Langerhans. J. Biol. Chem. 248:2705-2711.

23.

Burrj

I.,L. Balant, W. Stauffacher, andA. E. Renold. 1970. Perifusion ofratpancreatic tissue in vitro: substrate modification of

the-ophylline-inducedbiphasic insulin release. J. Clin.Invest. 11:2097-2105. 24.Arata,M., L.Karabatas,M. E.Fernandez,J.C.Cresto,E.Astolfi,

andJ.C. Basabe. 1983. Effect ofsalicylatepoisoningoninsulinsecretion.

Studiesonitsmechanism of action.ActaDiab. Latinoamer.

20:241-249.

25.Malaisse,W.J.1973.Insulin secretion: multifactorialregulation

forasingleprocessof release.Diabetologia.9:167-173.