Characterization of mixed syngeneic-allogeneic
and syngeneic-xenogeneic islet-graft rejections
in mice. Evidence of functional impairment of
the remaining syngeneic islets in xenograft
rejections.
O Korsgren, L Jansson
J Clin Invest.
1994;
93(3)
:1113-1119.
https://doi.org/10.1172/JCI117063
.
Allogeneic mouse islets or xenogeneic rat islets, or fetal porcine islets were implanted
under the renal capsule of C57BL/6 mice either alone or carefully mixed with syngeneic
islets. With this experimental model the syngeneic islets, although not rejected themselves,
are exposed to cytokines and inflammatory mediators released during either allograft or
xenograft rejection. No differences in insulin content could be observed between mixed islet
grafts and pure syngeneic islet grafts 6 wk after transplantation. Neither was there any
morphological evidence of a non-specific destruction of syngeneic islets. These findings
suggest that the mechanisms of both allograft and xenograft rejections are highly specific.
The hormone release from the mixed syngeneic-allogeneic grafts was similar to that from
pure syngeneic islet grafts. In contrast, a pronounced impairment of both the first and
second phases of insulin release was observed 2 wk after implantation in mixed
syngeneic-xenogeneic islet grafts. When perfusing the mixed islet graft after completed rejection of the
concordant xenogeneic rat islets (6 wk after implantation), the insulin release from the
remaining syngeneic mouse islets was identical to that of control grafts. However,
syngeneic mouse islets exposed to the rejection mechanism of the discordant xenogenic
pig islet-like cell clusters did not attain a complete functional recovery.
Research Article
Characterization of Mixed Syngeneic-Allogeneic
and
Syngeneic-Xenogeneic Islet-Graft
Rejections in Mice
Evidence of Functional Impairment ofthe Remaining Syngeneic Islets in Xenograft Rejections
OlleKorsgren and LeifJansson
DepartmentofMedical Cell Biology, Uppsala University, S-75123Uppsala, Sweden
Abstract
Allogeneic mouse islets or xenogeneic rat islets, or fetal porcine isletswere implanted under the renal capsule ofC57BL/6 mice either alone or carefully mixed with syngeneic islets. With this experimentalmodel the syngeneic islets, although not rejected themselves, are exposed to cytokines and inflammatory media-torsreleasedduring either allograft or xenograft rejection. No differencesininsulin content could be observed between mixed islet grafts and pure syngeneic islet grafts 6 wk after transplan-tation.Neitherwas there anymorphologicalevidence of a non-specificdestructionofsyngeneic islets. These findings suggest thatthemechanisms of both allograft and xenograft rejections arehighly specific.Thehormone release from the mixed synge-neic-allogeneic grafts was similar to that from pure syngeneic isletgrafts. Incontrast,a pronounced impairment of both the first and second phases of insulin release was observed 2 wk after implantation in mixed syngeneic-xenogeneic islet grafts. Whenperfusingthe mixed islet graft after completed rejection of the concordantxenogeneicrat islets (6wkafter implanta-tion), the insulin release from the remaining syngeneic mouse islets was identical to that of control grafts. However, synge-neic mouse islets exposed to the rejection mechanism of the discordantxenogenic pigislet-like cell clusters did not attain a completefunctionalrecovery.(J. Clin. Invest. 1994. 93:1113-1119.)Key words:transplantation* rejection-islets of
Langer-hans-cytokines-insulin release
Introduction
The immune response to allogeneically or
xenogeneically
transplanted tissues
comprises
bothspecific
andnonspecific
components. The latter
mainly
consists oftheinflammatory
reactioninflictedby thesurgical injury associated with
implan-tation. The
specific allograft reaction,
onthe otherhand,
de-pends on an activation of bothCD4'
andCD8+ Tlympho-cytes
(1).Bothindirect and directpathways
ofantigen
presen-tationcanstimulate activationandproliferation
of alloreactive Tlymphocytes
directed againstpancreatic
islets (2).During
these processes,alargenumberof
cytokines
arereleased, which notonly recruitimmunocompetent cells intothegraft,
butalso directly contributetocelldamage(3).Address correspondencetoDr.Olle Korsgren, Department of Medical CellBiology,BiomedicalCenter,POBox571,S-751 23Uppsala,
Swe-den.
Receivedfor publication26July1993andinrevisedform I
No-vember 1993.
The predominant cell mediating xenogeneic cellular im-mune responsesisprobably the CD4+ T lymphocyte. This con-cept isbasedon thefindingsthatmicetreatedwithananti-CD4 antibodydo not reject xenogeneic skin(4) orislet grafts(5), whereas CD8 + T cell depletion before transplantation was vir-tually withoutanybeneficial effectongraft survival. Immuno-suppressive drugsthatinhibit lymphocyte function or prolifera-tion, canreadilypreventallograft rejection, whereas theyare unable toaffect rejection ofacellularxenograft (6). This dis-crepancyimpliesthatdifferentcellularmechanismsare opera-tive inallograftorxenograft rejections, respectively (7, 8).
Cytokines, especiallyIL-l andTNF,areknown toimpair islet,B cellfunctioninvitro,and toultimatelycause celldeath (9). Based on these observations, it has been proposedthat cytokinesreleasedfrom activated macrophages withinthe pan-creatic isletsmay cause aprogressivep celldestruction, finally leadingtoinsulin-dependentdiabetes mellitus
(IDDM)'
(10). However,this notionhas beendifficulttoverifyinvivo,since direct effectsexertedbythecytokinesareimpossibleto sepa-ratefrom effects inducedby theimmunocompetentcells. How-ever,onepossiblemodel has beenrecently described by Sutton and co-workers, who used mixed syngeneic-allogeneic islet grafts (11). Theyconcluded that there was an absolute specific-ityof the allograft rejection within the mixed islet grafts, with-outanyassociated destruction ofthesyngeneic islets. However, nofunctionalevaluation oftheremaining syngeneic isletswas performed.Theaim ofthe present
investigation
was tostudythespeci-ficity ofboth
allograft
andxenograft
rejections.
Theexperimen-tal model with mixed
islet-grafts
may also show ifcytokines
released invivoin the very closevicinity ofthe remaining synge-neic isletscaninduceafunctional
impairment
similartothat observedinearly phasesof IDDM. For this purpose,syngeneic
mouseisletswereeither transplanted alone,or
carefully
mixed with either allogeneic mouseislets orxenogeneic
rat or fetal porcine islets.Thefunctionalcapacity
ofthemixedisletgrafts
was
characterized
byperfusing
theisletgraft-bearing
kidney,
eitherduringorafter
rejection.
Insulin extractionsand immu-nohistology were used for quantitativeanalysis
ofsurviving
islet/3 cells.
The specificity ofboth concordant and discordant
xeno-graft rejectionswasfoundtoequalthatseenin
allografts.
How-ever, the mechanisms mediating thexenograft
rejection
ex-ertedan
inhibitory
effectontheglucose-stimulated
insulinre-leasefromthe
remaining
syngeneic
mouseislets,
whichwas notseen
during
allograft rejection.
Thiseffect ismostlikely
me-diated by productsreleasedfrom activated
macrophages,
andit resembles theimpairment
of / cell function seen inearly
phasesofIDDM.Furthermore,the
finding
thatsyngeneic
islets1.Abbreviations usedinthis paper:BKs,B6 andC57/Ks
(mice);
ICC,islet-like cellclusters;IDDM,
insulin-dependent
diabetesmellitus.J.Clin.Invest.
©)TheAmericanSocietyfor ClinicalInvestigation,Inc. 0021-9738/94/03/1113/07 $2.00
mixed with discordant xenogeneic porcine islets showed
perma-nentfunctional impairment, whereas the syngeneic islets could recover completely when mixed with concordant xenogeneic rat islets, may indicate that xenogeneic rejections depend also on thephylogenetic disparity between donor and recipient spe-cies.
Methods
Animals. Male inbred C57BL/6 (B6) mice, 3-5 moold, servedas
recipients in allexperimentswith theexceptionofgroup E(seebelow),
where athymicB6 nu/nu mice(Bomholdtgaard, Ry,Denmark)were
used. Male and female inbred B6 andC57BL/Ks (BKs)mice bredat
theBiomedical Center(Uppsala, Sweden),originallyobtained from theJacksonLaboratory (BarHarbor, ME),3-5moold,wereusedas
syngeneicandallogeneicislet donors. MaleSprague-Dawleyratsfrom alocalbreedingcolonyatthe BiomedicalCenter,weighing - 300 g,
were used toobtain rat islets for concordantxenogeneic
transplanta-tions. Fetalpigsservedasdonorsforthediscordantxenogeneic
implan-tations. The pregnantsows werebredbyalocal farmer andwerekilled by means ofaslaughteringmaskonday70±5 of pregnancy(termis
115d). The fetuses were collected from the uterus,placedonice,and transportedtothelaboratory.Porcine islet-like cell clusterswerethen prepared as described below. All rodents had freeaccess totapwater
and pelleted food (type R34; AB AnalyCen, Lidkoping, Sweden)
throughout theexperimentalperiod.
Isletisolation andtransplantation. Pancreatic islets from rodents were preparedby a collagenase(Boehringer Mannheim, Boehringer, Germany)digestionmethod (12). Groupsof - 75 mouseor ratislets
were culturedfree-floating for periods of either 1-2 d (mouse) or 4-5 d (rat) in medium RPMI 1640(FlowLaboratories, Irvine,United
King-dom)supplementedwith 10%(vol/vol)calf serum (Statens Bakterio-logiskaLaboratorium, Stockholm, Sweden). The culture dishes were kept at370Cinagasphaseconsistingof 5%CO2inhumidified air, and theculture mediumwaschangedevery secondday.
Preparation and culture of fetal porcine pancreas has been de-scribed in detail ( 13 ).Briefly,thepancreatic glandswereminced into
fragmentswithasize of 1-2 mm3.After 10mg/ml collagenase
-di-gestion (BoehringerMannheim)duringsimultaneousvigorous shak-ing at 37°C for 8-10 min, thedigestwaswashed andexplanted into culturedishesallowingcellular attachment(Nunclon90mm0;Nunc, Kamstrup, Denmark).During the first 24 h the culturemedium con-sistedofserum-free RPMI 1640 (I1.1 mMglucose; FlowLaboratories)
with addition of 10 mMnicotinamide (Sigma Chemical Co., St. Louis, MO). The cultures were then supplemented with 10% human serum (vol/vol) (The Blood Center, Huddinge Hospital, Huddinge,
Swe-den).The culturedishes were keptat37°Cinagasphaseconsistingof 5%CO2inhumidified air, and the culture medium was changed every secondday. On day4 of culture, mostof the islet-like cell clusters (ICC) werefree floating or could be easily detached by gentle flushing with culture medium. Free-floating fragments with a diameter < 0.7 mmwereconsidered to be ICC and were harvested without any further
purificationstep.
Atthetime of transplantation, the islets or ICC were implanted through an incision inthe leftrenalcapsuleof avertin-anesthetized normoglycemic B6 mice ( 14). In some experiments (groups F-H be-low), islets from different donor species or strains were carefully mixed inaculturedish before transplantation. Islets were injected by means ofaglass braking pipette, and the capsulotomy was left unsutured. A totalofeight recipientanimal groups were designated as follows: (a) B6
micetransplanted with 150 B6 islets; (b) B6 mice transplanted with 150 BKsislets;(c) B6 mice transplanted with 150 rat islets; (d) B6 mice transplanted with 150 porcine ICC; (e) B6 (nu/nu) mice transplanted with 150 porcine ICC; (f) B6 mice transplanted with 150 B6 islets
mixedwith 150 BKs islets; (g) B6 mice transplanted with 150 B6 islets
mixedwith 150 rat islets; and (h) B6 mice transplanted with 150 B6 islets mixed with150porcine ICC.
Lightmicroscopicalexaminations. Animalswerekilled 2or6 wk aftertransplantation.Thetransplant-bearing kidneywasdissected free from blood vessels and fat. The isletgraft, clearlyvisibleas awhitish spot under the renalcapsule,wasexcised withamarginof - 3mm,
fixed in Bouin'ssolution,and embedded inparaffin. 7-Mm thick
sec-tionswerestained withhematoxylin and eosin and examined in a light microscope. Some of the sectionswerestained for the presence of insu-linby theperoxidase-antiperoxidase technique( 15). Guinea pig anti-bodiesagainst insulinwereobtained from Bio-Yeda(Rehovot, Israel). The second antibody and peroxidase-antiperoxidase complex were
from Dakopatts(Glostrup,Denmark).Thespecificityofthe
immuno-stainingwastestedbyomission of theprimary antibodyandby
appli-cation of theprimary antibody blocked with insulin.
Insulin and DNA contentsofthe isletgrafts.Thesedeterminations were made onlyin animals killed 6 wkafter transplantation. Thegraft was removedfrom the kidney as previously described ( 14), and soni-cated in 200 Mlof redistilledwater.Duplicate25-MA sampleswere
ana-lyzed byfluorophotometry forDNAcontent(16, 17).Atotalof 375
;l
acid-ethanol (0.18 M HCI in 95% (vol/vol ethanol) was added to the remaining homogenate. The samples were extracted overnight at 41C, followed byradioimmunologicalassay ofinsulin( 18).Perfusion ofgraft-bearing kidneys.Thesestudieswereperformed2
or6 wkafter transplantation.Thetechniquehaspreviouslybeen de-scribed in detail ( 14). Briefly, thetransplant-bearing kidneywas
re-movedtogether with parts of theaortaandinferiorvena cava.Both the
ureterand renal vein werecut, while theaorta wascannulated and infused with a continuously gassed (02/CO2 = 95:5) Krebs-Ringer bicarbonatebuffer (19)supplemented with 2.0% (wt/vol) each of bo-vine serumalbumin (Fraction V; Miles Laboratories, Slough, United Kingdom), Dextran T70 (Pharmacia Fine Chemicals, Uppsala,
Swe-den),and2.8or16.7 mMD-glucose. Themediumwasadministeredat arate of1.0ml/minwithout recyclingfor65 minwith a perfusion pressureof - 40 mmHg.
Theperfusionexperiments started with a 20-min period using a medium containing 2.8 mMglucose, followed by 30 min with 16.7
mMglucose andfinally 15 min with 2.8 mM glucose. A 1.0-ml sample of the effluent mediumwascollected every 5thmin,except forthe first 10minof perfusion after medium changes when samples were taken after 1-5, 7, and 10 min. Theinsulin concentrations of these samples were measured byradioimmunoassay. The rate ofinsulin secretion was calculated bymultiplyingtheinsulin concentration in the sample by theflow rate,givingvalues expressed as nanogramsof insulin per
min-ute.Thetotalstimulated insulin response to glucose of the grafts was obtained by planimetry of the areas for each individual perfusion
curve.During perfusionwith 16.7 mM glucose, areas were measured separatelyduringthefirst7min andduring the next 23 min to evaluate thefirst and second phases of insulin secretion separately.
Results
All animals tolerated the transplantation procedure without anysignsofinfirmity. The syngeneic grafts ingroup A could easilybeidentified macroscopically2 or 6 wk after transplanta-tion as anhomogenous whitishspot under the renal capsule. In groupsB,C, and D, onlyfibrousscar tissue could be seen at the site ofimplantation6 wkafter implantation,whereas in mixed isletgrafts(groupsF, G,and H), the islet tissue was scattered betweenstrandsoffibrotictissue.
Light
microscopicalexamination (Fig.1).
2wk aftertrans-plantation
endocrinecells were seen in all animals implanted withsyngeneic islets (groupA). ImmunohistochemicalJw
a~~~~~~~~~ak-'i~~~~~t2'-0-At i
AlA
ac
- *w>Sfi yv*i ,. Gdo ieNwt'g:'-V
*.
-.m
'#-'-w49 i
---A~~~~~~~~~~~~~~~~~~~~~~~~~i
~~~,
~~Ar
s 7 * @ ) + + ~~~~~~~~v..2.~~4r
WZ -'.two f f
;i'x
--d
x-d.*. . < t .
-Figure 1.Lightmicrographs of mixedsyngeneic-xenogeneic isletgrafts implantedunderthe renalcapsuleofC57BL/6mice. The sectionswere
immunocytochemicallystained for insulin (cells with dark cytoplasm). (a)Amixed transplantconsistingof 150syngeneicmouseislets and 150fetal isletlike cell clusters 2 wk afterimplantation. Scattered islets with preserved insulinimmunoreactivitycanbeseensurroundedbyadense mononuclearcell infiltrate. x100.(b)High power view of thesametransplantasina. Twoislets that containamajorityofinsulin-positive
cells andafew peripheral cells devoid of insulin immunoreactivitycanbeseen. Adense mononuclear cellinfiltrate,which doesnotpenetrate the isletcapsule, is present around the islets.x460.(c)Highpowermicrographofamixedtransplantconsistingof 150syngeneicmouseislets and 150ratislets 2 wk after implantation. Thesamemorphologicalappearanceasinbcanbeobserved.x460.(d)Amixedtransplantconsisting
of 150 syngeneic mouseislets and 150ratislets 6wkafterimplantation.Thegraftcontainsmainlyinsulin-immunoreactive islets with sparse fibrous tissue. The renalcortex canbeseenin the lower part of thefigure.x100.
from animals receiving eitherallogeneicor concordant/discor-dantxenogeneic isletgrafts (groups B, C, and D) displayedan intense mononuclear cell infiltration, which completely cov-ered the transplant. Also necrotic areas were regularly seen
within thesegrafts, and only a fewinsulin-positive cellswere observed. The mixed islet
grafts
(groups F, G, andH)
werecomposed of scattered intact
islets,
inwhichnocellular infiltra-tion couldbeseen, surroundedbyalargenumberof mononu-clearcells, whichcovered theremaining
parts of thegraft
(Fig. 1 a-c). Insulin staining revealed a large numberof insulin-positivecellswithintheintactislets, butonlyfew intheregionscontaining
mononuclearcells.Grafts obtained 6wkafter
transplantation containing
synge-neicislets, either alone(groupA)ormixed with allogeneicor xenogeneic islets(groups F,G,andH)consistedof intact islettissue,
which stained intensely for insulin. The mixed islet graftsalsocontainedclustersofmononuclear cells intheregion between the basal partoftheisletgraft
and thekidney paren-chyma(Fig. 1d).
Fibrousscartissue withnovestiges ofendo-crinecells was observedingraftsremovedfromanimals trans-plantedwith only allogeneicorconcordant/discordant xeno-geneicislets (groupB, C, andD).
Insulin and DNA content ofthe islet graft (Figs. 2 and 3). Nosignificant differencesininsulin contents of theislet grafts were observedbetween the syngeneic islet grafts in group A and themixed islets graftsin groupsF, G,and H. Transplants re-movedfrom animals implanted either with ICC, BKs, or rat islets alone (groups B, C, and D) contained only minute amountsofinsulin,which were close tothedetectionlimit of the radioimmunoassay. Fetal porcine ICC transplanted to nude B6mice(group E)hadaninsulincontent thatwas two to three times that measured in group A.
transplantation. There was no
morphological
evidence ofa<-a1t0000 ~ n nonspecificdestructionofsyngeneic islets, despite the presence
Z 8000- ofamarked mononuclear cell infiltration in their close vicin-60
ity.
Similar results havepreviously
beenreported
withmixed6000
-syngeneic and allogeneic ratislets (
11).
Theseresults werein-4000 +
terpreted
tosupport
the view ofan infiltration withmainly
toT>C
cytotoxic
^ . T Tcells in theisletgraft,
whereas the datawereconsid-*' 2000
jLjJ_[[71
ered
tobeincompatible
withthepresenceof
anonspecific
im-0 - munedestruction, such as adelayedtype ofhypersensitivity A B C D E F G H reaction.Since,in thepresent study,the insulin content of the mixedsyngeneic-xenogeneic islet grafts did not differ from that Figure 2. Graft insulin contents of islet grafts 6 wk aftertransplanta- of syngeneic islets implanted alone, the rejection ofisolated
tion under the kidney capsule of athymic B6 (nu/nu) mice (group
xengeneic
isletsalso
tee
toocu
bymechanismqutea
E) or B6mice (other groups). Islet grafts were originally composed
of either (a) 150 B6 islets; (b) 150 BKs islets; (c) 150 rat islets; (d) specific as those mediating allograft rejections. There issofar
150 porcine ICC; (e) 150 porcineICC;(f) 150 B6 islets mixed with only limited knowledge on the cellular immune responses to-150 BKs islets; (g) to-150 B6 islets mixed with to-150 rat islets; and (h) wards a xenograft, although it is believed that
CD4+
T lympho-150 B6 islets mixed with lympho-150 porcine ICC. Values aremean±SEM cytesareofmajor importance
(7, 8). WhileallogeneicMHC for five to six animals.***P<0.001 vsgroupA.antigens
canberecognized
directly by
the Tlymphocytes
(
1),
therecognitionof xenogeneic MHC antigens seems to require thatpeptidesof thexenogeneicMHCarepresentedin associa-Perfusion experiments (Figs. 4 and
5).
2 wkafter trans- tion with the MHC complex ofthe responding species (7). plantation, a pronounced first peak of insulin release in re- Thisnotionisfurther supportedbythe identificationofspecific sponse to aglucose stimuluswasobserved fromthesyngeneic
oligopeptides of xenogeneic MHC class I antigen inassociation graftsin group A. This was followedbyaslightlysmallersec- withrecipientMHCmolecules(21, 22).However, this com-ondphase of insulinrelease,
whichpromptly
returnedtobasal binedantigen(xenopeptideand recipientMHC)
isnot present levels when theglucosestimuluswasremoved(Fig.
4a).
The on theparenchymal xenogeneiccells. Thus, the intense mono-total amountofsecreted insulinwassimilarduring
the first and nuclear cell infiltrate among the xenogeneic isletsislikelyto second phases of insulinrelease(Fig.
4b).
Perfusion curves represent a cytotoxic Tlymphocyte-independent mechanism. almostidenticaltothoseofgroupA wereobtainedfrom ani- Cytokines,especiallyIL-1, are known to impair if cell func-malsimplanted withacombination ofsyngeneic
andalloge-
tion in vitro (9). It should be noted in this context that the neic islets (group F,Fig.
4a).There was,however,
amarked effects ofIL-I
on islet function are dose dependent, and the decrease in both thefirst and secondphasesofinsulin release in glucose-stimulated insulin release is the first,3
cell-specific groupsGand H(Fig. 4,aandb,mixedsyngeneic
and xenogen- functionto be adversely affected already by low IL-1 concen-eicislets). trationsin vitro (23). IL-1 is released from activated macro-The first phase insulin release fromsyngeneic grafts
6 wk phages, and one of its functions is to activate Tlymphocytes after transplantation (groupA)
wassimilartothatobserved in that have bound to theantigen-MHCcomplex on an antigen-animalsperfused
2wkafterimplantation,
butthemagnitude
of presentingcell (3). The release of cytokines within a rejecting thesecondphase had increased almost 100%(Fig.
5,
aandb).
allograft wasrecently investigated in kidneys (24) and livers Theinsulin release from mixedgrafts
in groups Fand Gwere (25).Cytokinesthatare thought to contribute tononspecific similartothat ofgroup A(Fig.
5b). However,
mixedsynge- lymphoid infiltration, such as IL-1#,
IL-6, and TNF-a, were neic-discordantxenogeneic
grafts
(group H)hadanormalfirst regularlyobserved in the implants. In most rejecting kidney phase insulin release, but amarkedly
reduced secondphase
allografts,mRNA for both IL-2R, IL-4, andIL-S
were present, insulinrelease(Fig.
5,aandb).
implyingthatthese cytokinesplayan important rolein ongo-Inallperfusionexperiments
insulinreleasereturnedtoba- ing allograft rejections, whereas mRNA for IL-2, IL-7, and salvalues, whichwerecloseto orbelow thedetection limit ofIFN-,g
were detected only rarely. However, a transient expres-theradioimmunoassay
forinsulin,
whenever themediumglu-coseconcentrationwasloweredto2.8mM.
Discussion
In
normoglycemic
animals,
thereisacloseandhighly
signifi-cantcorrelationbetween
pancreatic
insulincontentand i3cell volume(20),
andit islikely
thatthis isthe casealsofortrans-planted
islets ofLangerhans
(11).
Inthe presentstudy,trans-plants
removed from miceimplanted
with onlyporcine ICC,
rat, orBKsislets contained only minuteamountsof
insulin,
whereas no
differences
in insulin contentcould be observed between the mixed isletgrafts
and the puresyngeneic
isletgrafts
6 wk aftertransplantation.
Thesefindings
suggestthat thesyngeneic
islets remainintact,
and that themechanismof bothallograft
andxenograft rejection
ishighly specific.
Further-more, therejection
processseems tobe completed6 wkafter1 4-X 12-.0 10-._Du
:1
8-om 6
-0
') 4
-z
a 2-0
4
A B C D E F G H
Figure3. GraftDNA contentsof isletgrafts6 wkafter transplantation under thekidneycapsuleofathymicB6(nu/nu) mice (group E)or
B6 mice(other groups). Groupsaredefinedasin Fig2. Valuesare
mean±SEMfor fivetosix animals. *P<0.05,**fP<0.01,and***P
<0.001vsgroupA.
I
41
I
a
c1 4
-E 12 CD cn
1-10
-Xa 8
-c 4
'n
2-C s
-
0'-2.8 16.7
0 20 40
2.8
60
b
50
.-I
4
-0) 3 0
-U)
CIO
a
20-3
10-C,)
0-80
T
A F G H
Time (min)
Figure4.(a)Insulinconcentrations in the effluent medium collected from isletgraft-bearing kidneysof B6 mice.Kidneyswereperfused2 wk aftertransplantation with mediumcontaining2.8or16.7 mMglucoseasindicatedatthe top of thefigure.All animalsweretransplantedwith 150 syngeneic mouse islets, either alone (group A, opensquares), or mixed with 150allogeneicmouseislets(groupF, filled triangles), 150 rat islets(groupG,
filled
squares), or 150porcineICC(groupH,filled
diamonds).Valuesaremean±SEMfor fivetosix animals.(b)Total stimu-lated insulin secretion from isletgrafts2 wkafterimplantation.Insulin releasewasmeasuredseparatelyduringthe first 7 min(dotted bars),and during thenext23 min(hatched bars). GroupsaredefinedasinFig2. Valuesaremean±SEM for fivetosix animals. *P<0.05 and**P<0.01 vsgroupA.
sion ofIL-2andIFN-Mseems to precede the cellular infiltration in bothkidney (26)andislet (27) allograft rejection. Further-more, a local releaseof
IFN-A
and probably also TNF-a may be ofimportance toincrease the intrinsically low MHC antigen expression on islet tissue, thereby rendering islet cells more vulnerable to CD8 + T lymphocytes (28). Based on this, it can be assumedthatalargenumber ofcytokineswerelocally pro-duced within the allograft also in this model of mixed islet grafts, although no actual measurements of these substances wereperformed in the present study.Thus,although thesyngeneicmouseislets withinthegrafts areunlikelyto be aprimarytarget fortheimmunological
at-a
c C)
c
0
U) Cu
0 0
a)
Cn
c
2.8 16.7
12
-1 0
-
6-
4-2
-0
-0 20
2.8
C)
-0)
U1)
cnC
a)
CA
40
Time
(min)
60 80
tack, they may nevertheless be exposed to cytokines; e.g., IL-I
released during theallograft rejection in their close vicinity. However, no adverse effects on the syngeneic islets could be detected.The insulin secretion from the mixed syngeneic-allo-geneic grafts during the morphologically very intense allograft rejection (2wkafter implantation)orafter the completed re-jection process (6 wk after implantation) was not different when compared with thegrafts consisting only ofsyngeneic mouseislet.In contrast, a pronounced impairment of both the first and second phases of insulin release was observed 2 wk afterimplantationingrafts composed of mixed syngeneic-xen-ogeneic islets. When the mixed islet grafts wereperfused after
b
I
T
80
60-f
20
A B F G
I
HFigure5. (a) Insulin concentrations in theeffluent medium collected from islet graft-bearing kidneys of B6 mice. Kidneys were perfused 6wk
after transplantation with mediumcontaining2.8 or 16.7mMglucoseasindicatedatthe top of thefigure.Allanimals were transplanted with 150syngeneicmouseislets, either alone (group A, open squares), or mixed with 150 allogeneic mouse islets (group F, filled triangles), 150 rat islets (group G,filledsquares), or 150 porcine ICC (groupHfilled diamonds).Values aremean±SEM for animals. (b) Total stimulated insulin secretionfrom isletgrafts6wkafterimplantation. Insulin releaseweremeasuredseparatelyduring the first7 min(filled bars)andduring the
next23 min(hatched bars).Groupsaredefinedasin Fig 2. Valuesaremean±SEM for fivetosix animals. *P<0.05 and ***P<0.001 vs
groupA.
I -- -- mdWA600-4I - - --,-I
I _ _ _ I
49
m
I I
thecompletedrejectionof the concordantxenogeneicratislets (6 wk afterimplantation),the insulin release from the remain-ingsyngeneic mouse islets were identical to that of the control graftsin group A.However,syngeneicmouseisletsexposedto therejection mechanism of the discordant xenogenic pig ICC were unable toachieve acomplete functional recovery. This was demonstrated by amarkedlyimpaired second-phase insu-lin release,despiteatotalgraftinsulin content that was similar to thatofthesyngeneicisletgraftin group A. If it is assumed thatthisfunctionalimpairmentis exertedbyIL-1,the present experimental results give supportto the concept ofacentral role of macrophages(i.e.,themajor producerof IL-1) in the xenograft rejection. This notionis furtherstrengthened byan
immunohistological study of the xenograft rejection after transplantationof fetalpig proisletstoCBA/Hmice(5).The intense cellular infiltrate consisted predominantly of eosino-phils,macrophages,andCD4+Tlymphocytes,withonlysmall numbersofCD8+ Tcellsbeingpresent.Likewise,the absence ofadetrimentaleffectonthesyngeneic islets,when mixed with allogeneic islets, isin accordancewithprevious immunohisto-logical characterization ofthe cellularinfiltrate in classic allo-graft rejections, which consist mainly of Thy-1+, CD4+, CD8+, and IL2-R+ T cells with a relativepaucity of macro-phages (29). It shouldalsobe noted thatduringa numberof clinical and experimental conditions, there is evidence that macrophages or products released from these cells are asso-ciated withj3 celldysfunction. First,invitro coculture of macro-phages and mousepancreatic isletsresulted ininhibitionofa cell functionand celldeath(30). It wassubsequently shown that IL-1 (31) ornitric oxide (32) released fromthe macro-phages could beassociated with thisdysfunction.Furthermore, afunctional impairment of the
fl
cellresponsivenesstoa glu-cosestimulushas been demonstratedin earlystagesoftype 1 diabetes mellitus in humans (33), aswell asinthe two most studied experimental models of autoimmune diabetes melli-tus,namelyBB rats(34)and NODmice (35).Inthe develop-mentof autoimmune diabetes mellitusthereisinitiallyan in-tenseinfiltration ofmononuclear cellswithintheislets of Lan-gerhans, one of the first to appear being the macrophage (36-38).Interestingly, itwasrecentlydemonstratedthatislet allograftprimary
nonfunctiondepends on the presenceof mac-rophages or macrophage products that inhibit islet function untilan ultimatedestruction byCD4+ and CD8 + T cells in a classicallograft rejectionthat takesplace (29).Thepermanently (6 wk)impaired insulinreleasefrom the remaining syngeneicmouseislets,when mixed with discordant xenogeneic porcineICC, but not whenmixedwith concordant xenogeneicratislets,suggests thatxenogeneic rejection
mecha-.;
nismsdepend also on the phylogenetic disparity between the donor andrecipient species. The underlyingmechanisms are unknown,butitmay be speculated that the xenograft rejection between closely related species share some of its features with thatofanallograft rejection(6).
Acknowledgments
Theskilled technicalassistanceofBirgitta Bodin and Astrid Nordin is
gratefullyacknowledged. The study was supported by grants from the
Swedish Medical Research Council(12X-109 and 12P-9287), the
Nor-dicInsulin Fund, Svenska BarndiabetesFonden, the Swedish Hoechst
DiabetesFund,theSwedish Diabetes
Association,
theSwedishSocietyfor MedicalResearch, the Ernfors Family Fund, and the Juvenile Dia-betes Foundation International.
References
1. Lechler,R.I., G.Lombardi,J.R.Batchelor,N.Reinsmoen,and F. H. Bach. 1990.Themolecular basis ofalloreactivity.Immunol.Today.11:83-88.
2. Stock,P.G., N.L. Ascher, S.Chen,J.Field,F. H.Bach,and D. E. R. Sutherland. 1991. Evidence for direct and indirectpathwaysin thegenerationof the alloimmuneresponse against pancreaticislets.Transplantation (Baltimore). 52:704-709.
3. Weaver,C.T.,andE. R. Unanue. 1990. Thecostimulatoryfunction of antigen presenting cells.Immunol.Today. 1 1:49-55.
4.Pierson, R. N., H. J. Winn,P.S.Russell, andH.J. Auchincloss. 1989. Xenogeneic skin graftrejection is especially dependentonCD4+T cells.J.Exp. Med. 170:991-996.
5.Simeonovic, C. J.,R.Ceredig, and J.D.Wilson. 1990. Effect of GK 1.5 monoclonalantibody dosageonsurvival ofpig proislet xenograftsinCD4'T cell-depleted mice.Transplantation(Baltimore).49:849-856.
6.Auchincloss, H. J. 1988. Xenogeneictransplantation. Transplantation (Baltimore).46:1-20.
7.Auchincloss, H. 1990. Xenografting: a review. Transplant. Rev. 4:14-27. 8.Sachs,D.H., andF. H.Bach. 1990.Immunology of xenograft rejection. HumanImmunol. 28:245-251.
9. Sandler, S.,D. L. Eizirik, C.Svensson, E. Strandell,M.Welsh,and N. Welsh.1991.Biochemical and molecular actions of interleukin-1 onpancreatic
#-cells.Autoimmunity. 10:241-253.
10. Mandrup-Poulsen, T., S. Helqvist,J.Molvig, L.D. Wogensen, and J. Nerup. 1989.Cytokinesasimmuneeffectormolecules in autoimmune endocrine diseases withspecial referencetoinsulindependent diabetes mellitus. Autoimmu-nity.4:191-218.
I 1.Sutton, R.,D. W. R.Gray,P.McShane,M. J.Dallman, andP.J. Morris. 1989.Thespecificity of rejection and the absence ofsusceptibilityofpancreatic isletsj3cellstononspecific immune destruction in mixed strain isletsgrafted beneath the renalcapsule in therat.J.Exp.Med. 170:751-762.
12. Howell, S. L., andK. W.Taylor. 1968. Potassium ions and the secretion of insulin bytheislets of Langerhans incubated in vitro. BiochemJ.108:17-24.
13. Korsgren, O., S. Sandler,A.S. LandstrOm,L.Jansson,andA.Andersson. 1988. Large-scaleproduction of fetal porcine pancreatic isletlike cell clusters.An experimental tool for studies of islet cell differentiation andxenotransplantation.
Transplantation (Baltimore).45:509-514.
14.Korsgren, O.,L.Jansson, S.Sandler, andA.Andersson.1990. Hyperglyce-mia-inducedB celltoxicity.Thefate ofpancreatic islets transplanted into diabetic mice is dependentontheirgenetic background. J.Clin.Invest.75:2047-2055.
15.Sternberger, L.,P. H.Hardy,J. J.Cuculis,and H.G. Meyer. 1970. The labelledantibodyenzymemethodof immunohistochemistry. J. Histochem. C-tochem. 18:315-333.
16.Kissane,J.M., andE.Robins. 1958.Thefluorometricmeasurementsof deoxyribonucleicacid in animal tissues withspecial referencetothe central ner-vous system. J.Biol. Chem. 233:184-188.
17.Hinegardner,R. T.1971.AnimprovedfluorometricassayforDNA.Anal. Biochem. 39:197-201.
18.Heding, L. G. 1972. Determination oftotalseruminsulin (IRI) in insulin-treateddiabetic patients. Diabetologia. 8:260-268.
19. Krebs,H. A., and K.Henseleit. 1932.Untersuchungenuiberdie Harn-stofibildungimTierkorper. Hoppe-Seyler'sZ.Physiol.Chem. 210:33-66.
20. McEvoy,R.C. 1981.Changes in the volumes of theA-,B-, andD-cell populationsin thepancreaticislets during the postnatal development of therat. Diabetes. 30:813-817.
21.Maryanski,J.L.,P. Pala,G. Corradin,B. R. Jordan, and J.Cerottini. 1986. H-2-restrictedcytolyticTcellsspecific for HLAcanrecognizeasynthetic HLApeptide. Nature (Lond.). 324:578-580.
22.Gress,R.E.,S. G. Nathenson, andP. J. Lucas. 1989.Fine specificity of xenogeneic antigen recognition by human T cells. Transplantation (Baltimore). 48:93-98.
23.Leiter,E. H.1987. Murine macrophagesandpancreatic, cell. Chemotac-ticproperties of insulinand
al-cytostatic
action ofinterleukin-1. J.Exp.Med. 166:1174-1179.24. Krams,S.M.,D. A. Falco, J.C. Villanueva,J. Rabkin, S. J.Tomlanovich, F.Vincenti,W. J.C.Amend, J. Melzer, M. R. Garovoy, J. P. Roberts, N. L. Ascher,and0.M.Martinez. 1992.Cytokineand T cell receptor geneexpression atthesite ofallograftrejection.Transplantation(Baltimore). 53:151-156.
25.Martinez,0.M., S. M. Krams,M.Sterneck,J. C.Villanueva,D. A. Falco, L. D.Ferrell, J. Lake, J. P. Roberts,and N. L. Ascher. 1992.Intragraft cytokine profile during humanliver allograft rejection. Transplantation (Baltimore). 53:449-456.
27.O'Connell,P.J.,A.Pacheco-Silva, P. W.Nickerson,R.A.Muggia,M. Bastos, V. R.Kelley and T. B. Strom. Unmodified pancreatic islet allograft rejec-tion results inpreferential expression of certainTcellactivation transcripts.J. Immunol. 150:1093-1104.
28. Hao,L., Y.Wang,R. G.Gill,F. G. Rosa, D. W. Talmage,and K. J. Lafferty. 1990. Roleof lymphokine inisletallograftrejection. Transplantation (Baltimore).49:609-614.
29.Kaufman,D.B.,J.L. Platt, F.L. Rabe, D. L. Dunn, F.H.Bach, and D. E.R.Sutherland. 1990.Differential roles ofMac- + cells,andCD4+ and CD8+Tlymphocytes in primary nonfunctionandclassic rejection of islet allo-grafts. J.Exp.Med. 172:291-302.
30.Schwizer,R.W., E. H.Leiter, andR. Evans.1984. Macrophage mediated cytotoxicity against cultured pancreatic islet cells.Transplantation(Baltimore). 37:539-544.
31. Mandrup-Poulsen,T., K.Bendtzen,J.Nerup,C. A.Dinarello,M. Sven-son,and J. H.Nielsen. 1986.Affinity-purifiedhumaninterleukinIis cytotoxicto islet cells of Langerhans. Diabetologia. 29:63-67.
32. Kroncke, K. D.,V.Kolb-Bachofen, B.Berschick, V. Burkart,and H. Kolb. 1991.Activated macrophageskill pancreatic syngeneic islet cells via argi-nine-dependent nitric oxide generation. Biochem. Biophys. Res. Commun. 175:752-758.
33.Ganda, 0. P., S. Srikanta,S. J.Brink,M. A.Morris,R.E.Gleason,J.S. Soeldner, andG. S.Eisenbarth. 1984. Differential sensitivitytof-cell secretago-gues in"early"type Idiabetes mellitus.Diabetes.33:516-521.
34.Grill, V.,and L.Herberg. 1983.Glucose- and arginine-induced insulin andglucagon responsesfromtheisolatedperfused pancreas of the BB-Wistar diabetic rat: evidence for selectiveimpairmentofglucoseregulation.Acta Endo-crinol. 102:561-566.
35. Kano, Y., T. Kanatsuna, N. Nakamura, Y. Kitagawa, H. Mori, S. Kajiyama, K. Nakano,andM.Kondo. 1986.Defectofthefirst-phaseinsulin secretion to glucosestimulationin the perfusedpancreasofthenonobesediabetic (NOD) mouse. Diabetes. 35:486-490.
36.Lee,K.U.,K.Amano,and J. W.Yoon.1988.Evidenceforinitial involve-ment of macrophages in development of insulitis in NOD mice. Diabetes. 37:989-991.
37. Hanenberg, H., V.Kolb-Bachofen,G.Kantwerk-Funke, andH. Kolb. 1989.Macrophage infiltration precedes and is aprerequisiteforlymphocytic insulitisinpancreaticislets of pre-diabetic BBrats.Diabetologia.32:126-134.