interleukin inflammation induced - gut · murally with streptococcal...

Gut 1996; 39: 836-845

Interleukin 10 suppresses experimental chronic,granulomatous inflammation induced by bacterialcell wall polymers

H H Herfarth, S P Mohanty, H C Rath, S Tonkonogy, R B Sartor

AbstractBackground and Aims-Interleukin 10(IL10) inhibits monocyte/macrophage andT lymphocyte effector functions. Thisstudy examined the effect of systemicallyadministered IL10 on acute and chronicgranulomatous enterocolitis, hepatitis,and arthritis in a rat model.Methods-Lewis rats were injected intra-murally with streptococcal peptido-glycan-polysaccharide (PG-APS) polymers.Beginning 12 hours before PG-APSinjection, rats were treated daily withsubcutaneous murine recombinant IL10or vehicle for three or 17 days.Results-IL10 attenuated acute entero-colitis in a dose dependent fashion(p<0.01). Protective effects were moreprofound in the chronic granulomatousphase with decreased enterocolitis andmarkedly inhibited leucocytosis, hepaticgranulomas, and chronic erosive arthritis(p<0.001). IL10 downregulated tissue IL1,IL6, tumour necrosis factor a, and inter-feron y gene expression, consistent withthe in vitro effects of IL10 on PG-APS-stimulated splenocytes. Caecal IL1 proteinconcentrations and IL2 and interferon -Vsecretion by in vitro stimulated mesentericlymph nodes were downregulated in IL10treated animals.Conclusions-These results indicate thatexogenous IL10 can inhibit experimentalgranulomatous inflammatory responsesand suggest that IL10 treatment could bean effective new therapeutic approach inhuman disorders such as Crohn's disease,rheumatoid arthritis, and sarcoidosis.(Gut 1996; 39: 836-845)

Keywords: Interleukin 10, granuloma, inflammation,inflammatory bowel disease.

Inflammatory processes play a central part inmediating immune host defences to infectiouspathogens and in wound healing, but anunrestrained, overly aggressive inflammatoryresponse can induce inappropriate tissuedestruction. Recent observations in animalmodels show that the chronicity of inflam-mation is determined by a dysregulated balanceof pro and anti-inflammatory cytokines. 1-3Chronic immune mediated disorders likeinflammatory bowel diseases (IBD), rheu-matoid arthritis, and type I diabetes mellitus

seem to arise from a genetically determineddysregulated immune response to ubiquitousantigens.4`6 Therefore immunomodulatoryinterventions that utilise endogenous immuno-suppressive molecules provide an attractive,relatively non-toxic approach to treating thesechronic, idiopathic disorders.

Interleukin 10 (ILIO), originally namedcytokine synthesis inhibitor factor, is a cytokinethat inhibits many immune effector functions.IL10 is produced by a variety of cells includingT lymphocytes, B cells, and macrophages andsuppresses the in vitro production of interferon(IFN) y and IL2 by Thi lymphocytes and IL1,IL6, IL8, tumour necrosis factor (TNF) a,granulocyte-macrophage colony stimulatingfactor, and granulocyte colony stimulatingfactor by monocytes/macrophages.7 8

Several in vivo findings in experimentalanimal models suggest that ILl 0 is animportant anti-inflammatory agent. IncreasedIL10 gene expression correlates with recoveryin experimental autoimmune encephalomyelitis9and recombinant IL10 treatment attenuatesthe clinical course of this disease.'0 Spon-taneous intestinal inflammation occurs in IL10deficient (knockout) mice" and continuousadministration of IL10 attenuates experi-mental colitis in a lymphocyte transfer model.'2Furthermore, IL10 therapy prevents early deathand pancreatic necrosis in models of experi-mental endotoxaemia'3 and acute pancreatitis14respectively, probably by down regulatingTNFoa. Finally, blockade of endogenous IL10increases the severity of experimental arthritisin mice."

Peptidoglycan-polysaccharide (PG-PS) poly-mers, the basic structural component of the cellwalls of Gram positive and Gram negativebacteria, induce acute inflammation, which inthe susceptible host proceeds into a relapsing,granulomatous inflammatory reaction in severalorgans.'1'9 The granulomatous enterocolitismodel in susceptible Lewis rats is characterisedby chronic, transmural, spontaneously relapsingintestinal inflammation, accompanied by hepaticgranulomas, erosive arthritis, chronic anaemia,and leucocytosis.'7 Inflammation is induced bysubserosal (intramural) injection of PG-PSfrom group A Streptococcus pyogenes (PG-APS)into the terminal ileum and caecum. Inflam-mation in this model follows a biphasiccourse. The acute phase of focal intestinalinflammation peaks one to three days afterPG-APS injection, then gradually resolves, butis followed after 10-14 days by a spontaneous

Departments ofMedicine,Microbiology, andInununology andCenter forGastrointestinalBiology and Disease,University ofNorthCarolina,Chapel Hill, NC, USAH H HerfarthS P MohantyH C RathR B Sartor

College ofVeterinaryMedicine,North Carolina StateUniversity,Raleigh, NC, USAS TonkonogyCorrespondence to:Dr R B Sartor,Division of DigestiveDiseases,CB#7080,Bumett-Womack Bldg,University ofNorth Carolina,Chapel Hill,NC 27599-7080, USA.Accepted for publication22 July 1996

836

on March 9, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.39.6.836 on 1 D

ecember 1996. D

ownloaded from

http://gut.bmj.com/

Interleukin 10 suppresses experimental chronic, granulomatous inflammation induced by bacterial cell wall polymers

reactivation with extensive granulomatousintestinal and systemic inflammation. ILI isinvolved in the pathogenesis of acute andchronic phases of this model because caecalILI concentrations are increased and ILIreceptor antagonist (IL1 RA) treatmentattenuates inflammation.'7 T lymphocytes arenecessary for the chronic granulomatous phaseof PG-APS induced inflammation, as cyclo-sporin A completely prevents chronic inflam-mation and granuloma formation in the gut,liver and joints, and athymic animals fail todevelop chronic inflammation after PG-APSinjection.2022

Because T lymphocytes and pro-inflammatorycytokines such as ILI are involved in thepathogenesis of intestinal, liver, and jointinflammation in the PG-APS model, wepostulated that ILl 0 therapy should exert aprotective effect. In this study we examined theability of recombinant murine ILl0 to preventintestinal and systemic inflammation in thePG-APS enterocolitis model and investigatedthe in vivo and in vitro effects of IL10 on theregulation of key pro-inflammatory cytokines.

Methods

Animals and reagentsFemale, inbred specific pathogen free Lewisrats (145-155 g) were obtained from CharlesRiver Laboratories (Raleigh, NC). Rats werefed Agway Prolab rat chow (Agway, Syracuse,NY) ad libitum and were weighed daily. All ratexperiments were conducted in accord with thehighest standards of humane animal care asoutlined in the National Institutes of Health'sGuidefor the Care and Use ofLaboratory Animalsand approved by the University of NorthCarolina Institutional Animal Care and UseCommittee.Recombinant murine IL10 (Batch 3-mIL

10-3 specific activity 65 X 106 U/mg; Batch3-mIL 10-5 specific activity 45X 106 U/mg)was obtained from Schering-Plough ResearchInstitute (Kenilworth, NJ).

Preparation ofPG-APSPurified, sterile PG-APS fragments from thecell walls of group A, type 3, strain D58streptococci (Streptococcus pyogenes) were pre-pared as described previously23 and providedby Roger Brown and Dr John Schwab(Department of Immunology and Micro-biology, University of North Carolina, ChapelHill). Sonicated cell wall fragments by thismethod have molecular weights ranging from5X 106 to 5X 108.23 The final PG-APS concen-tration was calculated based on rhamnosecontent.23

Cytokine expression in vitro by splenocytes fromLewis ratsSplenocytes (5 X 106 cells/ml) were cultured forsix hours in the presence of PG-APS (50,ug/ml) and IL10 (0, 1, 5, 50 or 250 ng/ml).The culture medium was RPMI 1640

supplemented with 5% heat inactivated fetalcalf serum, 2 mM glutamine, 5 X 10-6 M2-mercaptoethanol, 100 units/ml penicillin G,1000 ,ug/ml streptomycin sulphate, and 0.25,ug/ml amphotericin B. Total RNA was isolatedusing Trizol (Gibco, Grand Island, MY) andstored at -80°C until further processing.

Induction of enterocolitis by PG-APSLewis rats, randomised into treatmentgroups, were anaesthetised (20 gil/100 g byintramuscular injection Innovar-Vet; PitmanMoore, Washington Crossing, NJ), and intes-tines were exposed by laparotomy using aseptictechnique. Rats were injected subserosally withPG-APS using a total dose of 12.5 pug/g bodyweight rhamnose (equivalent to 37.5 ,ig dry wtPG-APS/g body weight) distributed amongseven injection sites in the ileal Peyer's patches,terminal ileum, and caecum as previouslydescribed.24 Control rats received identicalinjections of 37.5 ,ug/g/body weight humanserum albumin (HSA) (Baxter Health CareCorp, Glendale, CA).

ILIO treatment and experimental designRecombinant murine ILIO was administeredby subcutaneous injection in different concen-trations (100 ,ug or 250,g/kg/24 h in the acutephase experiment or 125 pug/kg/12 h in thechronic phase experiment) in 0.3 ml sterileNaCl. Control PG-APS and HSA injectedanimals received rat serum albumin (RSA)(Sigma, St Louis, MO) subcutaneously indifferent concentrations (250 ,ug/kg/24 h in theacute phase experiment or 125 pug/kg/12 h inthe chronic phase experiment) dissolved in0.3 ml sterile NaCl. In both studies treatmentwas started 12 hours before PG-APS or HSAintramural injection and given for the nextthree or 17 days respectively, in the abovedescribed scheme.

Clinical assessment ofinflammation andharvesting of tissuesAfter three (acute phase) or 17 (chronic phase)days all rats were killed by inhalation of 100%CO2. Cardiac blood was obtained for cellcounts, which were processed using anautomated cell counter. Liver and spleenweights were recorded and normalised for theindividual rat body weight. Gross intestinalinflammation was scored by a single blindedobserver using a previously validated method. 17Values of 0-4 (4 being the most severe) wereassigned to the (a) presence of caecal serosalnodules, (b) severity of contracted mesentery,(c) severity of adhesions, and (d) extent ofcaecal bowel wall thickening. The resulting'gross gut score' is the sum of these values, themaximum possible being 16. Additionally themacroscopically visible number of liver granu-lomas was assessed by assigning a value from0-4. Caecal, mesenteric lymph node and livertissues were snap frozen in isopentane and keptat -80°C for later mRNA and protein analysis.Samples of the liver, terminal ileum, caecum,

837



j.com/


ecember 1996. D

ownloaded from

http://gut.bmj.com/

Herfarth, Mohanty, Rath, Tonkonogy, Sartor

and joints from each animal were fixed informalin, embedded in paraffin wax, and sec-tioned for histochemical staining (haematoxylinand eosin).

Monitoring of arthritisArthritis was monitored clinically (erythema,swelling) and by measuring the diameters ofboth rear ankle joints with a Fowler Ultra-Cal.II caliper (Lux Scientific Instrument Corp,New York, NY) as described previously.23 25Joints were measured before PG-APS injectionand serially thereafter. To reduce bias, allgroups were coded, and the rats were selectedrandomly from the cages, so that no identi-fication was visible to the person making themeasurements. Data were expressed as theincrease of joint diameter in mm comparedwith the baseline measurement before PG-APSor HSA injection.

Histological and biochemical evaluation ofinflammationA histological inflammatory score was evaluatedfor each animal by a blinded observer asdescribed previously.'7 Briefly, values from 0-4(4 being the most severe) were assigned forboth acute and chronic inflammation in thedifferent layers of the intestine in coded crosssections of the terminal ileum, mid caecum,and caecal tip. The acute and chronic scoresof each animal were totalled, a score of 24representing the maximum possible histo-logical inflammatory score.

Measurement of myeloperoxidase activityCaecal myeloperoxidase (MPO) activity wasdetermined using a modified standard method.26Briefly, caecal tissue was homogenised in 4 ml20 mmol/l phosphate buffer (pH 7.4) andcentrifuged at 4000 rpm at 4°C for 20 minutes.The pellet was homogenised and sonicatedwith 2 ml 50 mmoIl acetic acid (pH 6) con-taining 05% (wt/vol) hexadecyltrimethyl-ammonium hydroxide. MPO activity of thesupernatant was determined by measuring theH202 dependent oxidation of 3,3',5,5'-tetra-methylbenzidine and expressed as units pergram of tissue.

TABLE I Oligonucleotide primers usedfor amplification ofdifferent cytokines by RT-PCR

5 --------Sense Primer--------3' Sample AnnealingmRNA 5'--------Antisense Primer-------3' cDNA temperature

IFNy 5'-AAG ACA ACC AGG CCA TCA GCA-3' 547 bp 600C5'-AGC CAC AGT GTG AGT TCA GTC-3'5'-ATG GAA GGA AAG AGC CTC CTC-3' 225 bp 600C5'-CT GGC GAT GCT CAT GAA TGC-3'

TNFa 5'-CAC CAT GAG CAC GGA AAG CA-3' 692 bp 580C5'-GCA ATG ACT CCA AAG TAG ACC-3'

IL6 5'-CTT CCA GCC AGT TGC CTT CT-3 504 bp 60`C5'-GAG AGC ATT GGA AGT TGG GG-3'

ILla 5'-TCC TGT GAC TCT CAA AGT CTC-3' 362 bp 580C5'-ATG CGA GTG ACT TAG GAC GA-3

IL13 5'-GCT ACC TAT GTC TTG CCC GT-3' 543 bp 580C5-GAC CAT TGC TGT TTC CTA GG-3'

IL1 RA 5'-CTG GGA TAC TAA CCA GAA GAC C-3' 367 bp 580C5'-TGG TGT GTT GGT GAG GCT CAC-3'

1B actin 5'-ACC ACA GCT GAG AGG GAA ATC G-3' 281 bp 60`C5-AGA GGT CTT TAC GGA TGT CAA CG-3'

RNA preparationTotal tissue RNA was prepared using astandard method with guanidine thiocyanate-cesium chloride as previously described.'7 TotalRNA was quantified by ultraviolet spectro-photometry (A260/A280). The integrity andquality of each RNA sample was checked byelectrophoresis on a 1% agarose gel containingethidium bromide.

Polymerase chain reaction (PCR)For each sample 1 ,ug of total RNA was reversetranscribed in a total volume of 25 gl con-taining lXfirst strand buffer (Gibco, GrandIsland, NY) 125 U of Moloney murineleukaemia virus RT (Gibco, Grand Island, NY),15 U RNAse inhibitor (Promega, Madison,WI), 0.5 mM each of four dNTPs (Pharmacia,Piscataway, NJ), and 5 pmol of randomHexamers (Pharmacia, Piscataway, NJ). Thereaction was carried out at 39°C for one hour,followed by 93°C for seven minutes, 24°C forone minute, and finally cooled down to 4°C for30 minutes. The reaction mixture was storedat -20°C until further use.The amplification was carried out in a 9600

Perkin-Elmer cycler (Applied Biosystem, FosterCity, CA). One to 2 ,ul of cDNA sample wasamplified in 50 ,lI of a reaction mixturecontaining 1 XTaq buffer II (Perkin Elmer,Norwalk, Co), 1.5 mM MgCl2, 2 pLM each of5' and 3' primers and 1 U ofTAQ polymerase(Perkin Elmer, Norwalk, CO). Samples wereheated for four minutes at 94°C and thefollowing cycles were conducted at 94°C for45 seconds, 60°C for 45 seconds, 72°C for 1.30minutes. At the end the reaction mix was heldat 72°C for five minutes. Negative controlswithout cDNA were amplified with each PCRexperiment. To confirm that equal amounts ofRNA were added in each PCR within anexperiment and to verify a uniform amplifi-cation process P actin mRNA was reverse tran-scribed and amplified in each assay. Aliquots ofthe samples were analysed by electrophoresison a 2% agarose gel containing ethidiumbromide. The DNA products were visualisedby ultraviolet fluorescence and photographsof the gels were taken with a negative film(Polaroid 665, Polaroid Corp, Cambridge,MA). The negative was scanned with a Silver-scanner II PS v2.Aa using Adobe Photoshop2.5.1 software.

Primers specific for rat cytokines were con-structed according to published sequences27~33and are listed in Table I. The authenticity ofthe PCR product was confirmed by comparingactual and predicted size using a 100 bp DNAladder (Gibco) and by observation of frag-ments of predicted sizes after digestion withthe listed restriction enzymes: EcoR I (IFNy),TAQ I (TNFac, ILI RA), BgI II (IL6), Acc I(IL1 at), BamH I (IL>).

Cell cultures of mesenteric lymph nodes (MLN)Mesenteric lymph nodes from rats treated withIL10 or the vehicle were removed, gentlyminced, and filtered through a fine mesh.

838



j.com/


ecember 1996. D

ownloaded from

http://gut.bmj.com/


The cells were washed and resuspended(5X 106 in 1 ml) in RPMI and additives asdescribed above. The cells were cultured at37°C in 5% CO2 and 95% 02 in a 24 welltissue culture plate for 10 or 36 hours in thepresence or absence of concanavalin A(1 ,ug/ml), after which the supematants werecollected for the assays. Experiments wereperformed with cells from at least threeindividual animals.

Cytokine assaysILla: Caecal tissues were homogenised in icecold 20 mmol/l phosphate buffer (pH 7.4) andthe supernatant kept at -80°C until assay. Thehomogenates were measured in 1:1 dilutionsusing a radioimmunoassay kit for rat ILlcx(Cytokine Science, Boston, MA).IFNy activity was measured by a rat IFNy

ELISA (Biosource Int; Camarillo, CA).IL2 was measured in a bioassay using the

murine cell line designated NK. Briefly,titrated amounts of culture supematants wereadded to NK cells (5 X 103 cells/well in 100 ,lculture volume) and incubated for 48 hours.The cultures were pulsed with 3H-thymidinefor the last four hours. Cells were harvestedand incorporation of 3H-thymidine was de-termined by scintillation counting. Units perml of IL2 are defined as the reciprocal of thedilution of supematant that stimulates 50%of the maximal counts per minute of3H-thymidine incorporation, multiplied by 10as the bioassay volume is 0 1 ml.

Statistical analysisData were analysed using a statistical software(Statview II; Abacus Concept, Berkely, CA).Differences between the groups were com-pared using one way analysis of variance orusing x2 analysis. Differences were consideredsignificant if p values were <0 05.

BPL 1:1-1:10-1:100 1:1-1:10-1:100 1:1-1:10-1:100 1:1-1:10-1:100IFNy bp 225

ILloa bp 362

TNFa bp 692

IB Actin bp 281

PG-APS IL10 | _ 50 ng/mI 1 250 ng/mI

Figure 1: Semiquantitative RT-PCR analysis of the regulation ofcytokine gene expin isolated splenocytes by ILI 0. Splenocytes were isolatedfrom Lewis rats and stimuwith 50 ,g/ml PG-APS with or without different concentrations ofIL10 (50 or 250Total RNA was isolated after six hours, reverse transcribed, and the cDNA was dill10-fold steps (1:1, 1:10, 1:lOOfor ILIla, IFNy, TNFa or 1:10, 1:100, 1:1000 foxUsing specific oligonucleotide primers, amplification was performedfor 35 (Actin, I40 cycles (IFNy, TNFa) and aliquots of the products were visualised on a 2% agawith ethidium bromide. (BPL= 100 base pairDNA ladder). Representative gels ofindependent experiments are shown.

Results

In vitro inhibition of cytokines by ILIOTo demonstrate the potential anti-inflammatoryrole of recombinant murine IL10 on rat cellsstimulated with PG-APS, we isolated spleno-cytes from Lewis rats and added 50 ,ug/mlPG-APS and different concentrations of ILIOfor six hours. PG-APS markedly upregulatedIFNy, ILlox, and TNFo gene expression insplenocytes in vitro (Fig 1). Fifty and 250 ng/mlrecombinant murine ILl 0 inhibited geneexpression ofIFNy, ILlot, and TNFa in a dosedependent fashion (Fig 1), whereas lowerdoses (1 or 5 ng/ml) were ineffective (resultsnot shown).

Effects ofILIO treatment on the acuteinflammatory phaseTwo different doses of recombinant murineIL10, 100 or 250 jig/kg, were administereddaily by subcutaneous injection, starting 12hours before surgery. Three days after surgery,the rats were killed and the inflammatorychanges of the intestine were scored in ablinded fashion. An amelioration of the grossgut score (GGS) with a significant benefit bythe higher IL10 dose (250 ,ug/kg) was observed(Fig 2A). In the analysis of the differentcomponents of the GGS, the most significantimprovement was in decreased caecal thicken-ing (mean (SEM)) (1.9 (0.1) PG-APS/IL10250 ,ug/kg versus 2.3 (0.1) PG-APS/RSA;p<0. 01). Adhesions and mesenteric contractionsimproved with IL10 treatment, but werenot significantly better than control values.Surprisingly, caecal MPO values were signifi-cantly increased with the higher ILIO dosecompared with the control group (p<0 05)(PG-APS/RSA: 4.5 (04) units/g; PG-APS/IL10 100 jig/kg: 5.7 (0.6) units/g; PGPS/IL10250 jig/kg: 6.0 (0.4) units/g; HSA/RSA: 3.1(0.5) units/g) and histological examination ofselected animals showed no difference betweenthe IL10 treated and untreated PG-APSinjected groups (results not shown). ILl0treatment ameliorated the acute arthritis, butthe differences between the groups were notsignificant due to the low incidence of acutearthritis after intestinal intramural PG-APSinjection.17 Furthermore, there were no differ-ences in body weight, liver weight, packed cellvolume or white blood cell count between thegroups.

Effects ofIL1O on the chronic inflammatory** phase

Using the higher ILIO dose (250 ,ig/kg/24 h),which exhibited greater protective effects on

H20 gross intestinal inflammation in the acute phase,we investigated whether continuous IL10

ression administration for 17 days had a protectiveelated effect in the T lymphocyte dependent, chronicuted in phase of this model. Because a recentlyrActin). published study showed better effects of IL10Lla) or treatment when given at shorter time inter-roseegel vals,34 we administered divided doses of IL10three

(125 pig/kg/12 h).

839



j.com/


ecember 1996. D

ownloaded from

http://gut.bmj.com/


W

Ul)cto

0

E

(a

0)

(0

Co

00

HSAI PG-APS/ PG-APS/ PG-APS/RSA RSA IL10 IL10

(100 gg/kg) (250 fg/kg)(n=3) (n=8) (n=8) (n=8)

12.5r B

W

Ul)cCD

E

0

0

CO

cri

0

0

u0q-(D

10

7.5 _-

5

*

T

2.5 _-

0'~ ~ ~ ~ ~

PG-APS/ PG-APS/ HSA/RSA IL10 RSA

(n = 10) (n = 10)Figure 2: (A) In vivo administration ofIL10 (100 ,ug/kgor 250 ,ug/kg every 24 hours subcutaneously) shows a dosedependent effect on acute intestinal inflammation in thePG-APS model. ILI 0 or RSA treatment was started 12hours before subserosal PG-APS orHSA injection andthen was continuedfor three days. The gross gut score wasmade by a single blinded observer at necropsy on day 3.HSA=human serum albumin; RSA=rat serum albumin;PG-APS=peptidoglycan-polysaccharide polymersisolatedfrom groupA Streptococcus pyogenes. *p<0 01PG-APSIIL10 v PG-APS/RSA. (B) In vivo subcutaneousadministration ofIL10 (125 ,ug/kg every 12 hours)significantly reduces chronic macroscopic gut inflammation17 days after PG-APS orHSA injection. *p<0.003PG-APS/ILI 0 v PG-APSIRSA.

Chronic intestinal inflammationIL10 significantly reduced grossly detectableintestinal inflammation by more than 40% inthe chronic granulomatous phase of PG-APSinduced enterocolitis (p<0-003) (Fig 2B). In

contrast with the acute phase significantimprovement was observed in all parameters ofthe gross gut score, including thickening of thebowel wall, caecal nodules, intestinal adhesions,and mesenteric contractions. By histologicalassessment significant differences were seenbetween the treated and untreated PG-APSinjected groups in both the acute (neutrophiliccell infiltration and abscess formation) andchronic components (mononuclear cell infil-tration and granuloma formation) (Table II).IL10 treatment resulted in a 30% decrease inthe total (summed) histological inflammatoryscore (p<0-03). The improvement of the acutescore with IL10 treatment was more pro-nounced in the terminal ileum than in thecaecum, where only a trend could be observed.The small suppression of neutrophil infil-tration in the caecum by IL10 was alsoreflected by caecal MPO values, which showedonly a slight, non-significant improvementin the treatment group (PG-APS/RSA: 12-1(1.0) units/g; PG-APS/IL10 250 ,ug/kg: 11.8(1.7) units/g; HSAIRSA: 2.7 (0.5 units/g).

Chronic extraintestinal inflammationExtraintestinal inflammation was inhibited byIL10 to a greater extent than enterocolitis.Eighty per cent of the PG-APS/RSA group hadmacroscopically visible liver granulomas, com-pared with only two of 10 animals in thePG-APS/IL10 group (p<0-001). Histologicalanalysis showed that one additional IL10treated rat had scattered hepatic granulomas,which were not visible by macroscopicevaluation (Table III). However, these threeIL10 treated animals showed considerably lesshepatic inflammation than the PG-APS/RSAanimals in regard to histological distributionand destruction of liver parenchyma. Althoughparenchymal granulomas were only rarely seenin the liver ofIL10 treated rats, a mononuclearcell infiltrate was observed around bile ducts(Fig 3). The increase in liver weight ofPG-APS injected/ILl 0 treated rats comparedwith non-PG-APS injected controls wasdecreased by 45% (p<0-03) (Table III), con-firming a beneficial effect of IL10 on hepaticinflammation. Similar beneficial effects ofIL10 treatment were seen on haematologicalparameters, which are reliable indices ofintestinal and systemic inflammation in thismodel.'7 Peripheral leucocytosis and anaemiawere significantly attenuated by ILIO admin-istration (p<0-004) (Table III). No effect was

noted on body weights, which were notdifferent between PG-APS and HSA injectedgroups.

TABLE II ILI0 treatment suppressed inflammation in the gut 17 days after PG-APS injection as assessed by histologicalinflammatory scores of the intestine

Terminal ileum Caecum Histological inflammatory scoreGroupsGut injection/treatment acute chronic* acute chronic acutet chronic Histological scorest

PG-APA/RSA (n=10) 3-8 (0-1) 3-7 (0-8) 7-0 (0-2) 7-4 (0-2) 10-9 (0-2) 11-1 (0-2) 22-0 (0-5)PG-APS/IL10 (n=10) 2-4 (0-6) 2-6 (0-4) 5-1 (1.1) 5-2 (0-9) 7-6 (1-5) 7-8 (1-2) 15-4 (2-8)HSA/RSA (n=5) 0 0 0 0 0 0 0p Value§ <0-03 <0-02 <0-04 <0-05 <0-02 <0-03

*Mean (SEM). tSummation of distal ileal and caecal blinded histological scores (0-4), as described in the methods section.*Summation of acute and chronic score. §p Values: PG-APS/IL10 v PG-APS/RSA group.

840



j.com/


ecember 1996. D

ownloaded from

http://gut.bmj.com/


TABLE iII IL10 treatment suppresses extraintestinal inflammation and anaemia in the chronic phase (day 17) of thePG-APS model

841

Hepatic granulomastGroups WBC PCV Liver weightGut injection/treatment X 1O SI units/l % mg/g body weight macroscopic histological

PG-APS/RSA (n=10) 29-7 (2.6)* 37-8 (1.0) 70-5 (4.6) 80% 80%PG-APS/IL1O (n=10) 17-3 (1-7) 42-7 (1.1) 56-5 (3.6) 20% 30%HSA/RSA (n=5) 7-7 (0.4) 48-4 (0.7) 39-7 (1-2) 0% 0%p Value§: <0-001 <0 004 <0 03 <0-01 <0-02§

*Mean (SEM); tincidence of hepatic granulomas; tp values PG-APS/IL10 v PG-APS/RSA; §p values PG-APS/IL10 vPG-APS/RSA calculated by x2test.

EE._

a,0-.

.' XccW

Co

0 0._ E._

-1 4 8 9 1011121314151617Days after PG-APS injection

Figure 4: Effect ofmurine recombinant ILI 0 treatment onPG-APS induced arthritis. IL10 (125 pug/kg every 12hours) was given subcutaneously starting 12 hours beforePG-APS injection (12.5 ,ug/g subserosally in the intestine)and continuing throughout the duration of the experiment(n=10). Controls included rats injected intramurallywith PG-APS (n=10) orHSA (n=5) and treatedwith RSA (125 ,g/kg every 12 hours subcutaneously).J7oint measurements were carried out at indicated daysafter PG-APS injection. J7oint diameter changes inPG-APS/RSA treated rats were significantly greater(*p<0.01) than PG-APS/IL10 treated rats at eachinterval after day 10.

Figure 3: Histological demonstration of the liver architecture17 days after subserosal PG-APS injection and RSAtreatment (A) or continuous murine recombinant IL10treatment (B). Only infiltration ofmononuclear cells inthe area of the portal tract are visible in liver sections ofPG-APS/IL10 treated rats, whereas the livers ofPG-APSIRSA treated rats show destruction of thearchitecture with granulomas (haematoxylin and eosinstain, original magnification X 20).

Murine recombinant IL10 treatment sup-pressed the development of chronic arthritis ina highly significant fashion (Fig 4). On day 17after PG-APS injection joints of IL10 treatedrats were not significantly increased overHSA/RSA negative controls; the increase in themean (SEM) joint diameter from preinjectionvalues seen in the PG-APS/RSA group wasreduced by nearly 80% (1.85 (0.37) mm inthe PG-APS/RSA versus 0.38 (0.19) mm inthe PG-APS/IL10 treatment group, p<0-001)(Fig 4). Only two of 10 animals showed clinicalsigns of arthritis after IL10 treatment with veryattenuated arthritis in one of the two rats,whereas eight of 10 animals in the PG-APS/RSA group clinically showed profound arthritiswith marked swelling and erythema. The histo-pathology of the joints confirmed the clinicalpattern. Figure 5 shows that infiltration ofPMNs and mononuclear cells, pannus for-mation, cartilage erosion, bone degradation aswell as exudation of cells into the joint spacewere clearly inhibited by IL10 treatment.

A'.. -

Figure 5: Representative joint tissuesfrom rats treated withRSA (A) or recombinant murine IL10 (B) for 17 daysafter intestinal subserosal PG-APS injection. I1140treatment prevented clinical and histological (B) onset ofarthritis, whereas the histological section of only RSAtreated rats show pannus formation with infiltration ofneutrophils, and mononuclear cells, as well as cartilagedestruction and bone erosion (A) (haematoxylin and eosinstain, original magnification X 20).



j.com/


ecember 1996. D

ownloaded from

http://gut.bmj.com/


Effects ofILJO treatment on cytokine regulationIL10 reduced or inhibited the expression ofILla and P, TNFac, IL6, and IFNy genes inthe liver, mesenteric lymph nodes, and caecaltissues as measured by RT-PCR. The mostpronounced effects were seen in the liver con-sistent with the marked differences in hepaticgranuloma formation (Fig 6). Expression ofIL1 RA and ILIO mRNA (results not shown)did not differ between the IL10 treated andnot treated groups; these cytokines werealso constitutively expressed in non-inflamedHSA/RSA controls.

Earlier studies showed significantly increasedILlot protein concentrations in caecal tissues inthe chronic phase after PG-APS subserosalinjection.'7 IL10 treatment decreased ILlaprotein levels by 40% (Fig 7). Furthermore, weassessed the production of IFNy and IL2 byconcanavalin A stimulated MLN cells, isolated17 days after PG-APS injection. Whereas MLNcells showed no measurable spontaneous pro-duction ofIFNy or IL2, concanavalin A stimu-lation upregulated the production of IFNytwofold and IL2 25-fold in PG-APS/RSA

25r-

wU)'uc

0

'uE0.

-J

---

T20 F-

15i- T10 H-

PG-APS/RSA PG-APS/IL10 HSA/RSA

(n= 10) (n = 10) (n = 5)Figure 7: Caecal ILIa protein levels measured byradioimmunoassay in caecal tissues 17 days after injectionofPG-APS or HSA. ILIa protein concentrations areexpressed as mean (SEM). ILIO treatment inhibited ILIa,whereas PG-APS alone significantly increased caecal ILlaprotein levels compared with the HSA control group(*=p<0003 PG-APS/RSA v HSA/RSA).

treated rats compared with HSA/RSA controls. a greater effect on chronic, granulomatousHowever, concanavalin A stimulated MLN enterocolitis than on the acute phase of inflam-cells isolated from ILIO treated rats showed a mation. ILIO treatment significantly improved45% reduction of IFNy and IL2 production macroscopic, histological, and immunologicalfrom PG-APS/RSA treated animals. parameters of chronic intestinal inflammation

17 days after PG-APS injection. The improve-ment in neutrophilic infiltration was less pro-

Discussion nounced, as reflected by only minor changes ofThis study shows that murine recombinant caecal tissue myeloperoxidase levels, althoughILl0 treatment successfully attenuated acute the acute histological score in the ileum (butand chronic granulomatous inflammation not the caecum) was significantly decreasedinduced by bacterial cell wall polymers and with ILl0 treatment. These results suggestdownregulated cytokine expression in vivo and that ILl 0 treatment affects mononuclear cellin vitro. Continuous IL10 administration had (lymphocytes and macrophages) infiltration to

a greater extent than neutrophils, in agreement- I I with two previously published studies.'434

In murine acute pancreatitis IL10 did notinfluence cellular infiltration despite inhibitingtissue necrosis.'4 Furthermore, IL10 treatmentdid not improve the morphology or degree ofcellular infiltrate in footpads injected with ThIlymphocytes in a murine model of delayed type

I_- - hypersensitivity,34 albeit IL10 inhibited foot-pad swelling by decreasing vascular per-meability. Although we did not quantitate thedegree of oedema or vascular leakage in thisstudy, the significant improvement in acutemacroscopic inflammation was mainly due todecreased bowel wall thickening, correspond-ing with observations in the delayed typehypersensitivity model.The macroscopic and histological improve-

- I 1 1 ment of chronic inflammation by ILlO treat-ment was accompanied by a down regulationof gene expression of monokines as well as

- IFNy in tissues, consistent with the ability ofIL10 to inhibit expression of these cytokines byPG-APS stimulated splenocytes in vitro. These

HS4A/ PG-APS/ PG-APS/ IH0 results agree with the ability of IL10 to inhibitIRSA I IL10 RSA I 2 macrophage and lymphocyte activation7 and to

ession of different cytokines in the liver ofILI 0 treated PG-APS increase survival in a model of endotoxaemiapared with PG-APS/RSA and HSA/RSA animals. Shown are three by downregulating TNFot. "zls ofeach group. On day 17 RNA was isolatedfrom liver tissues and The effects ofIL10 treatment on in vivo gened using RT-PCRfor 23-40 cycles with rat cytokine specific primers.cts were visualised on a 2% agarose gel. To confirm a uniform expression of the monokines ILl, IL6, ands3 actin was coamplified. TNFa were most pronounced in the liver, in

IFNybp 547

ILlabp362

IL1J bp 543

IL1RA bp 367

TNFa bp 692

1L6 bp 496

P Actin bp281

Figure 6: Gene expr6injected animals comrrepresentative animacDNA was amplifie4Aliquots of the produamplification process

842

5

a



j.com/


ecember 1996. D

ownloaded from

http://gut.bmj.com/


parallel with a marked inhibition of hepaticgranulomas. Kupffer cells and newly recruitedmonocytes are important in the genesis ofhepatic granulomas after intraperitoneal injec-tion of PG-APS.35 36 Kupffer cells rapidly clearthese bacterial polymers from the circulationand large amounts of PG-APS remain in thephagolysosomes of these cells after 14-21days.36 Our studies suggest that continuousIL10 administration to PG-APS treated ratsdownregulates monokine production by acti-vated Kupffer cells and mononuclear phago-cytes, thereby inhibiting formation of hepaticgranulomas. This hypothesis is supported byin vitro studies showing that isolated Kupffercells secrete ILI and TNFcL after stimulationwith PG-APS36 37 and that LPS or PG-APSstimulated TNFc. secretion by Kupffer cellscan be inhibited by ILl0 (38 and H Herfarth,S Lichtman unpublished results).IL10 nearly completely prevented chronic

arthritis in this model, probably by downregulation of IL1, IL6 or TNFox, as shown inthe liver. A role for these macrophage productsin the pathogenesis of experimental arthritisis demonstrated by reactivation of PG-APSinduced arthritis by systemic injections of IL1or TNFPc and by induction of acute, transientarthritis by intraarticular injection of ILlotor (.25 39 40 Furthermore, treatment withrecombinant ILl RA or anti-TNFa antibodyinhibits PG-APS induced arthritis,39 as doessystemic treatment with TGF.4' In vitroTGF(3 decreases LPS induced macrophageproduction of TNFax and IL1 to a similarextent as ILl0, but inhibits cytokinetranslation in contrast with the activity of IL10on transcriptional downregulation.42 IL1, IL6,and TNFot are also implicated in the patho-physiology ofhuman rheumatoid arthritis5 andblocking endogenously produced IL10 insynovial cultures markedly increased thesecytokines, suggesting an important immuno-regulatory role of IL10 in the joint.43 It is likelythat IL10 has a similar key role in maintainingmucosal homeostasis, as targeted deletion ofthis cytokine leads to spontaneous intestinalinflammation.`IL10 treatment more successfully inhibited

extraintestinal inflammation in liver and jointsthan in the intestine, similar to the therapeuticeffects of IL1 RA in this model.'7 The amountof PG-APS in various organs may be a criticalcomponent because concentrations of thispolymer are highest at intestinal injectionsites.24 Another variable is tissue concen-trations of the administered IL10 within thevarious inflamed organs, which could not bemeasured in this study. Given the ability ofPG-APS to activate multiple inflammatorypathways'9 and the redundancy of manyimmunological processes,4 44 45 treatment withIL10 alone might not be sufficient to com-pletely suppress aggressive enterocolitis. Anadditional anti-inflammatory agent may benecessary to provide synergistic benefit. Thisconcept is elegantly demonstrated by Powrieet al, who showed in a delayed type hyper-sensitivity model that the combination ofIL4 and IL10 is more effective than

monotherapy.46 In preliminary investigations,we have demonstrated additive activities ofIL10 and corticosteroids in the PG-APSmodel.47ILIO suppresses Thl cytokine production48

as well as Thl lymphocyte proliferationin vitro,49 if macrophages are used as antigenpresenting cells. IL4, however, seems to bemore important than IL10 in crossregulating aThl response towards a Th2 response.50 Thecrucial role of T lymphocytes in chronicintestinal and systemic inflammation in thePG-APS model is well documented.20-22 51 Inour model in vivo IL10 treatment down-regulated IFN-y and IL2 protein secretion ofmesenteric lymph nodes after Con A stimu-lation and inhibited tissue IFNy gene ex-pression. Furthermore in vitro ILIO decreasedIFN-y upregulation in PG-APS activated spleno-cytes. Negative regulation of IFNy by ILIO isthe suggested mechanism of disease preventionin the experimental allergic encephalomyelitismodel'0 and in the non-obese diabetic mousemodel.52As shown here, IL10 downregulated expres-

sion of several pro-inflammatory cytokinesin vivo. However, we cannot determine if theseprotective effects are direct or indirect throughinduction of other inhibitory cytokines, likeILI RA, or by inhibition of pro-inflammatorymediators like IL1 2. IL1 RA, which attenuatesPG-APS induced enterocolitis and systemicinflammation,'7 25 is upregulated in the pres-ence of IL10 in LPS activated neutrophils,53whereas IL10 had no effect on IL1 RA pro-duction by mononuclear cells isolated frompatients with rheumatoid arthritis54 or in vitroendotoxin stimulated whole blood from healthyvolunteers after intravenous ILIO admin-istration.55 In our study, tissue ILI RA mRNAexpression was not affected by IL10 treatment,but by downregulating IL1 production, IL10increased the ratio between IL1 RA and IL1towards a protective environment. A similarshift towards an immunosuppressive ratio ofIL1 RA/IL1 was recently demonstrated inisolated mononuclear cells from patients withCrohn's disease exposed to I10.`56 IL10 alsodownregulates the pro-inflammatory cytokineIL12,57 which is involved in the pathogenesisof intestinal inflammation, as recently shownby the successful treatment of experimentalcolitis by IL12 antibodies.58 IL12 is secreted byactivated macrophages and B cells and stronglyincreases IFNy production by T lymphocytes,thus shifting the immunological responsetowards a Thl repertoire, as recently shown ina parasitic experimental model.59The PG-APS model, which displays geneti-

cally determined host susceptibility, ileocaecalinflammation, extraintestinal manifestations ofanaemia, leucocytosis, hepatobiliary inflam-mation, arthritis, and exhibits spontaneousrelapses of inflammation, resembles manyfeatures of Crohn's disease. Furthermore,PG-APS is an environmentally relevant agent,which is present in high concentrations inthe lumen of the distal ileum and colon.Ubiquitous anaerobic bacteria and their cellwall components seem to be involved in IBD

843



j.com/


ecember 1996. D

ownloaded from

http://gut.bmj.com/

844 Herfarth, Mohanty, Rath, Tonkonogy, Sartor

and reactive arthritis.'9 60 Our results demon-strate that IL10 down regulates experimentalchronic granulomatous inflammation inducedby bacterial cell wall components in geneticallysusceptible hosts. In this model IL10 appearsto act by suppressing the expression of mono-kines and IFNy. Immunomodulating IL10treatment could be a new effective therapeuticapproach for chronic inflammatory diseasesin humans, particularly Crohn's disease andrheumatoid arthritis, which are charac-terised by increased tissue concentrations ofproducts of activated macrophages and Thllymphocytes.5 44 61

This research was supported by NIH grants DK 40249,DK 47700 and DK 34989, the Deutsche Forschungsgemein-schaft (He 2379/1-1; Ra 671/1-1), and partially supported bythe Schering Plough Research Institute. The authors thankRoger Brown for preparation of PG-APS and Diane Bender,Julie Mitchell and Lisa Holt for technical support.

Parts of this article were presented at the 46th AnnualMeeting of the American Association for the Study of LiverDisease, November 3-7, 1995 and were published in abstractform in Hepatology 1995; 22: 364A.

1 Locksley RM. Th2 cells: Help for helminths. Jf Exp Med1994; 179: 1405-7.

2 Herfarth HH, Sartor RB. Cytokine regulation of experi-mental intestinal inflammation. Current Opinion inGastroenterology 1994; 10: 625-32.

3 Sartor RB. Insights into the pathogenesis of inflammatorybowel diseases provided by new rodent models ofspontaneous colitis. Inflammatory Bowel Disease 1995; 1:64-75.

4 Sartor RB. Current concepts of the etiology and patho-genesis of ulcerative colitis and Crohn's disease.Gastroenterol Clin NAm 1995; 24: 475-507.

5 Sewell KL, Trentham DE. Pathogenesis of rheumatoidarthritis. Lancet 1993; 341: 283-6.

6 Atkinson MA, MacLaren NK. The pathogenesis of insulin-dependent diabetes mellitus. N Engl Jf Med 1994; 331:1428-36.

7 Mosmann TR. Properties and functions of interleukin-10.Adv Immunol 1994; 56: 1-16.

8 Bogdan C, Nathan C. Modulation of macrophage functionby transforming growth factor beta, interleukin-4, andinterleukin- 10. Ann NYAcad Sci 1993; 00: 713-39.

9 Kennedy MK, Torrance DS, Picha KS, Mohler KM.Analysis of cytokine mRNA expression in the centralnervous system of mice with experimental autoimmuneencephalomyelitis reveals that IL-10 mRNA expressioncorrelates with recovery. Jf Immunol 1992; 149:2496-505.

10 Rott 0, Fleischer B, Cash E. Interleukin-10 preventsexperimental allergic encephalomyelitis in rats. Eur JfImmunol 1994; 24: 1434-40.

11 Kuhn R, L6hler J, Rennick D, Rajewsky K, Muller W.Interleukin- 10 deficient mice develop chronicenterocolitis. Cell 1993; 75: 263-74.

12 Powrie F, Leach MW, Mauze S, Menon S, Caddle LB,Coffman RL. Inhibition of Thl responses preventsinflammatory bowel disease in scid mice reconstitutedwith CD45RBh CD4+ T cells. Immunity 1994; 1:553-62.

13 Gerard C, Bruyns C, Marchant A, Abramowicz D,Vandenabeele P, Delvaux A, et al. Interleukin-10 reducesthe release of tumor necrosis factor and prevents lethalityin experimental endotoxemia. Jf Exp Med 1993; 177:547-50.

14 Van Laethem JL, Marchant A, Delvaux A, Goldman M,Robberecht P, Velu T, et al. Interleukin 10 preventsnecrosis in murine experimental acute pancreatitis.Gastroenterology 1995; 108: 1917-22.

15 Kasama T, Strieter RM, Lukacs NW, Lincoln PM,Burdick MD, Kunkel SL. Interleukin-10 expression andchemokine regulation during the evolution of murinetype II collagen-induced arthritis. Jf Clin Invest 1995; 95:2868-76.

16 Wahl SM, Hunt DA, Allen JB, Wilder RL, Paglia L,Hand AR. Bacterial cell wall-induced hepaticgranulomas. An in vivo model of T cell-dependentfibrosis.JffExp Med 1986; 163: 884-902.

17 McCall RD, Haskill 5, Zimmermann EM, Lund PK,Thompson RC, Sartor RB. Tissue interleukin-1 andinterleukin-l receptor antagonist expression in entero-colitis in resistant and susceptible rats. Gastroenterology1994; 106: 960-72.

18 Cromartie WJ, Craddock JG, Schwab JH, Anderle SK,Yan CH. Arthritis in rats after systemic injection ofstreptococcal cells or cell walls. Jf Exp Med 1977; 146:1585-602.

19 Schwab JH. Phlogistic properties of peptidoglycan-polysaccharide polymers from cell walls ofpathogenic andnormal-flora bacteria which colonize humans. InfectImmun 1993; 61: 4535-9.

20 Sartor RB, Bender DE, Allen JB, Zimmermann EM,Holt LC, Pardo MS, et al. Chronic experimentalenterocolitis and extraintestinal inflammation areT-lymphocyte dependent. Gastroenterology 1993; 104:775A.

21 Allen JB, Malone DG, Wahl SM, Calandra GB, Wilder RL.Role of the thymus in streptococcal cell wall-inducedarthritis and hepatic granuloma formation. Comparativestudies of pathology and cell wall distribution in athymicand euthymic rats. J Clin Invest 1985; 76: 1042-56.

22 Yocum DE, Allen JB, Wahl SM, Calandra GB, Wilder RL.Inhibition by cyclosporin A of streptococcal cell wall-induced arthritis and hepatic granulomas in rats. ArthritisRheum 1986; 29: 262-73.

23 Stimpson SA, Schwab JH. Chronic remittent erosivearthritis induced by bacterial peptidoglycan poly-saccharide structures. In: Chang JY, Lewis AJ, eds.Pharmalogical methods in the control of inflammation. NewYork: Liss, 1989: 381-94.

24 Sartor RB, Cromartie WJ, Powell DW, Schwab JH.Granulomatous enterocolitis induced in rats by purifiedbacterial cell wall fragments. Gastroenterology 1985; 89:587-95.

25 Schwab JH, Anderle SK, Brown RR, Dalldorf FG,Thompson RC. Pro- and anti-inflammatory roles ofinterleukin-1 in recurrence of bacterial cell wall-inducedarthritis in rats. Infect Immun 1991; 59: 4436-42.

26 Grisham MB, Benoit JN, Granger DN. Assessment ofleukocyte involvement during ischemia and reperfusion ofintestine. Meth Enzymol 1990; 186: 729-42.

27 Dijkema R, Van der Meide PH, Dubbeld M, Caspers MWJ,Schellekens H. Cloning, expression, and purification ofrat IFN-gamma. Meth Enzymol 1986; 119: 453-63.

28 Eisenberg SP, Brewer MT, Verderber E, Heimdal P,Brandhuber BJ, Thompson RC. Interleukin-1 receptorantagonist is a member of the interleukin-1 gene family:evolution of a cytokine control mechanism. Proc NatlAcadSci USA 1991; 88: 5232-6.

29 Feeser W, Freimark BD. Nucleotide sequence of rat pro-interleukin-1 beta mRNA. GenBank 1992; Accession:M98820.

30 Kwon J, Chung I, Benveniste E. Cloning and sequenceanalysis of the rat tumor necrosis factor-encoding genes.Gene 1993; 132: 227-36.

31 Nishida T, Nishino N, Takano M, Sekiguchi Y, Kawai K,Mizuno KU, et al. Molecular cloning and expression ofrat interleukin-1 alpha cDNA. Jf Biochem 1989; 105:351-7.

32 Northemann W, Braciak TA, Hattori M, Lee F, Fey GH.Structure of the rat interleukin 6 gene and its expressionin macrophage-derived cells. Jf Biol Chem 1989; 264:16072-82.

33 Nudel U, Zakut R, Shani M, Neuman S, Levy Z, Vaffe D.The nucleotide sequence of the rat cytosplasmatic beta-actin gene. Nucleic Acids Res 1983; 11: 1759-71.

34 Li L, Elliott JF, Mosmann TR. IL-10 inhibits cytokineproduction, vascular leakage, and swelling during Thelper 1 cell-induced delayed-type hypersensitivity.JImmunol 1994; 153: 3967-78.

35 Kossmann T, Manthey CL, Brandes ME, Morganti KMC,Ohura K, AllenJB, et al. Kupffer cells express type I TGF-beta receptors, migrate to TGF-beta and participate instreptococcal cell wall induced hepatic granulomaformation. Growth Factors 1992; 7: 73-83.

36 Manthey CL, Kossmann T, Allen JB, Corcoran ML,Brandes ME, Wahl SM. Role of Kupffer cells indeveloping streptococcal cell wall granulomas.Streptococcal cell wall induction of inflammatorycytokines and mediators. Am Jf Pathol 1992; 140:1205-14.

37 Lichtman SN, Wang J, Schwab JH, Lemasters JJ.Comparison of peptidoglycan-polysaccharide and lipo-polysaccharide stimulation of Kupffer cells to producetumor necrosis factor and interleukin-l. Hepatology 1994;19: 1013-22.

38 Grewe M, Gausling R, Gyufko K, Hoffmann R, Decker K.Regulation of the mRNA expression for tumor necrosisfactor-a in rat liver macrophages. Jf Hepatol 1994; 20:811-8.

39 Schwab JH. Bacterial cell-wall induced arthritis: models ofchronic recurrent polyarthritis and reactivation ofmonoarticular arthritis. In: Henderson B, Pettifer R,Edwards J, eds. Mechanisms and models in rheumatoidarthritis. London: Academic Press, 1995: 439-54.

40 Stimpson SA, Dalldorf FG, Otterness IG, Schwab JH.Exacerbation of arthritis by IL-1 in rat joints previouslyinjured by peptidoglycan-polysaccharide. J Immunol1988; 140: 2964-9.

41 Brandes ME, Allen JB, Ogawa Y, Wahl SM. Transforminggrowth factor beta 1 suppresses acute and chronic arthritisin experimental animals. JClin Invest 1991; 87: 1108-13.

42 Bogdan C, Paik J, Vodovotz Y, Nathan C. Contrastingmechanisms for suppression of macrophage cytokinerelease by transforming growth factor-beta and inter-leukin-10.JffBiolChem 1992; 267: 23301-8.

43 Katsikis PD, Chu CQ, Brennan FM, Maini RN,Feldmann M. Immunoregulatory role of interleukin 10 inrheumatoid arthritis. JfExp Med 1994; 179: 15 17-27.

44 Sartor RB. Cytokines in intestinal inflammation: patho-physiological and clinical considerations. Gascroenterology1994; 106: 533-9.



j.com/


ecember 1996. D

ownloaded from

http://gut.bmj.com/

Interleukin 10 suppresses experimental chronic, granulomatous inflammation induced by bacterial cell wall polymers 845

45 Paul WE. Pleiotropy and redundancy: T cell-derivedlymphokines in the immune response. Cell 1989; 57:521-4.

46 Powrie F, Menon S, Coffman RL. Interleukin-4 and inter-leukin-10 synergize to inhibit cell-mediated immunityin vivo. EurJImmunol 1993; 23: 2223-9.

47 Herfarth HH, Mohanty S, Rath HC, Sartor RB. In vivoIL-10 treatment suppresses chronic granulomatousinflammation and proinflammatory cytokine expression inexperimental enterocolitis, arthritis and hepatitis.Gastroenterology 1996; 110: A924.

48 Fiorentino DF, Bond MW, Mosmann TR. Two types ofmouse T helper cell. IV: Th2 clones secrete a factor thatinhibits cytokine production by Thl clones. Jf Exp Med1989; 170: 2081-95.

49 Ding L, Shevach EM. IL-10 inhibits mitogen-inducedT cell proliferation by selectively inhibiting macrophagecostimulatory function. JlImmunol 1992; 148: 3133-9

50 Paul WE, Seder RA. Lymphocyte responses and cytokines.Cell 1994; 76: 241-51.

51 Van Den Broeck MF, Van De Langerijt LG, VanBruggen MC, Billingham ME, Van Den Berg WB. Treat-ment of rats with monoclonal anti-CD4 induceslong-term resistance to streptococcal cell wall-inducedarthritis. EurJImmunol 1992; 22: 57-61.

52 Pennline KJ, Roque GE, Monahan M. Recombinanthuman IL-10 prevents the onset of diabetes in thenonobese diabetic mouse. Clin Immunol Immunopathol1994; 71: 169-75.

53 Cassatella MA, Meda L, Gasperini S, Calzetti F, Bonora S.Interleukin 10 (IL-10) upregulates IL-1 receptor antagonistproduction from lipopolysaccharide-stimulated humanpolymorphonuclear leukocytes by delaying mRNAdegradation. _Exp Med 1994; 179: 1695-9.

54 Hart PH, Ahem MJ, Smith MD, Finlay-Jones JJ.Comparison of suppressive effects of interleukin-10 andinterleukin-4 on synovial fluid macrophages and bloodmonocytes from patients with inflammatory arthritis.Immunology 1995; 84: 536-42.

55 Chernoff AE, Granowitz EV, Shapiro L, Vannier E,Lonnemann G, Angel JB, et al. A randomized, controlledtrial of IL-10 in humans. Inhibition of inflammatorycytokine production and immune responses. J Immunol1995; 154: 5492-9.

56 Schreiber S, Heinig T, Thiele HG, Raedler A. Immuno-regulatory role of interleukin 10 in patients withinflammatory bowel disease. Gastroenterology 1995; 108:1434-44.

57 Cassatella MA, Meda L, Gasperini S, D'Andrea A, Ma X,Trinchieri G. Interleukin-12 production by human poly-morphonuclear leukocytes. EurJtImmunol 1995; 25: 1-5.

58 Neurath MF, Fuss I, Kelsall B, Stuber E, Strober W.Antibodies to interleukin 12 abrogate establishedexperimental colitis in mice. J Exp Med 1995; 182:1281-90.

59 Finkelman FD, Madden KB, Cheever AW, Katona IM,Morris SC, Gately MK, et al. Effects of interleukin 12 onimmune responses and host protection in mice infectedwith intestinal nematode parasites. J Exp Med 1994; 179:1563-72.

60 Sartor RB. Microbial factors in the pathogenesis of Crohn'sdisease, ulcerative colitis, and experimental intestinalinflammation. In: Kirsner JB, Shorter RG, eds.Inflammatory bowel disease. Baltimore: Williams andWilkins, 1995: 96-124.

61 Simon AK, Seipelt E, Sieper J. Divergent T-cell cytokinepatterns in inflammatory arthritis. Proc NatlAcad Sci USA1994; 91: 8562-6.



j.com/


ecember 1996. D

ownloaded from

http://gut.bmj.com/

interleukin inflammation induced - gut · murally with streptococcal...

Documents