Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
1
Differentialexpressionofsolublereceptorforadvancedglycationend-products(sRAGE)in1
micesusceptibleorresistanttochroniccolitis2
3
Michael Bramhall1,2, Kevin Rich2, Ajanta Chakraborty2, Larisa Logunova2, Namshik Han3,4,4
JamesWilson5,CatherineBooth5,JohnMclaughlin3,6,AndyBrass3,SheenaM.Cruickshank25
6
AuthorAffiliations:7
1DepartmentofBiochemistryandMolecularBiology,SchoolofBiomedicalSciences,Monash8
University,Clayton,Victoria3800,Australia9
2Lydia Becker Institute of Immunology and Inflammation, Manchester Academic Health10
Science Centre, School of Biological Sciences, Faculty of Biology, Medicine, and11
Health,UniversityofManchester,ManchesterM139PT,UK12
3ManchesterAcademicHealthScienceCentre,SchoolofMedicalSciences,TheUniversityof13
Manchester,ManchesterM139PT,UK14
4TheGurdonInstitute,UniversityofCambridge,CambridgeCB21QN,UK15
5EpistemLtd.,ManchesterM139XX,UK16
6SalfordRoyalNHSFoundationTrust,SalfordM68HDUK17
Runningtitle:Markersofprotectiveresponsesincolitis18
19
CorrespondingAuthor20
S.M.Cruickshank,AVHillBuilding,UniversityofManchester,ManchesterM139PT,UK21
[email protected], Phone+44(0)1612751578 22
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
Abstract23
Aims: Identifying the factors that contribute to chronicity in inflamed colitic tissue is not24
trivial.However,inmousemodelsofcolitis,wecaninvestigateatpreclinicaltimepoints.We25
sought tovalidatemurineTrichurismuris infectionasamodel for identificationof factors26
thatpromotedevelopmentofchroniccolitis.27
Methods:WecomparedpreclinicalchangesinmicewitharesolvingimmuneresponsetoT.28
muris (resistant) versus mice that fail to expel the worms and develop chronic colitis29
(susceptible).Findingswerethenvalidatedinhealthycontrolsandpatientswithsuspected30
orconfirmedIBD.31
Results:TheReceptorforAdvancedGlycationEndProducts(Rage)washighlydysregulated32
between resistant and susceptible mice prior to the onset of any pathological signs.33
IncreasedsolubleRAGE(sRAGE) intheserumandfaecesofresistantmicecorrelatedwith34
reducedcolitisscores.Mousemodelfindingswerevalidatedinapreliminaryclinicalstudy:35
faecalsRAGEwasdifferentiallyexpressedinpatientswithactiveIBDcomparedwithIBDin36
remission,patientswithIBDexcludedorhealthycontrols.37
Conclusion: Pre-clinical changes in mouse models can identify early pathways in the38
developmentofchronicinflammationthathumanstudiescannot.Weidentifiedthedecoy39
receptor sRAGE as a potentialmechanism for protection against chronic inflammation in40
colitisinmiceandhumans.WeproposethattheRAGEpathwayisclinicallyrelevantinthe41
onset of chronic colitis, and that further study of sRAGE in IBD may provide a novel42
diagnosticandtherapeutictarget.43
44
Keywords:sRAGE,colitis,mouse45
46
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
3
Introduction47
Inflammatory bowel diseases (IBD) are a group of intestinal immune disorders, including48
Crohn’sdisease(CD)andulcerativecolitis(UC),thatcausechronicinflammationinthegut49
[1]. The cause of IBD is currently not known, but dysregulation of intestinal immunity,50
microbial dysbiosis, genetics and environmental factors contribute to disease onset.51
Unpredictablecyclesofremissionandrelapserequirecarefulmonitoringandthelong-term52
damage from inflammationoftenwarrantspotent immunomodulatory therapyor surgical53
intervention[2].54
Itisimpossibletoreliablypredictonset,relapseorremissionofIBD[3]andcurrently,only55
animalmodelsprovideameansofstudyingtheperturbationsinthegutthatprecedecolitis.56
Infecting susceptible mouse strains with the enteric nematode parasite Trichuris muris57
closelyparallelshumanCrohn’sdiseaseinboththepathologicalandtranscriptionalchanges58
induced[4]andhasbeenestablishedasamodelforthestudyoftheinitiationof immune59
responses in the colon [5]. T. muris resistant BALB/c and C57BL6 mice mount an early60
immune response against thewormswithin 24 hours of infection,with large numbers of61
dendritic cells (DCs)migrating to the lamina propria,whereasAKRmice or C57BL/6mice62
withalowdoseinfectionmountadelayedimmuneresponse,resultinginchronicintestinal63
inflammationandafailuretoexpeltheworms[5,6].Bothsusceptibleandresistantstrains64
showmildsignsofinflammationwithin24hours.However,inflammationinresistantmiceis65
controlled and ultimately resolves, whereas susceptible strains go on to develop clinical66
colitisinthesubsequentweekspostinfection.67
Exploiting these early differences in the host immune response to T. muris infection68
experimentallymayprovideinformationonthefactorsthatpromotetheonsetofchronic,69
ratherthanresolving,inflammationinthegut[7].Earlyfactorsareimpossibletodistinguish70
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
from the inflammatorymilieupresent in IBDpatients at thepointofdiagnosis as chronic71
inflammation is already well established. Identification and validation of early changes72
duringchroniccolitisonsetinmicecouldprovideausefulpipelinefordevelopingdiagnostic73
anddisease-managementbiomarkersortherapeutictargetsinhumancolitis.74
In this study, we carried out a T. muris infection study, investigating preclinical75
transcriptionalchanges24hourspostinfection(PI).Weidentifiedthereceptorforadvanced76
glycation end-products (Rage) as highly upregulated in mice susceptible to T. muris77
infection.WefurtherinvestigatedthepresenceofRAGEandrelatedligandsincoliticmice78
andcarriedoutatranslationalvalidationstudy investigatingthepresenceofsolubleRAGE79
(sRAGE)inthefaecesofIBDpatientsandhealthycontrols.80
81
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
5
Materialsandmethods82
Mice83
All animal procedures used in this project were carried out in accordance with the UK84
Animals(ScientificProcedures)Act,1986.ForT.muris infectionexperiments,6-8weekold85
male BALB/c and AKR mice were used (Harlan UK, Bicester, UK). Mice were housed in86
individuallyventilatedcageswithnestingmaterialandweremaintainedunderconstant12h87
light–darkcycleat21-23°Cwithfreeaccesstowaterandstandardchow(BeekayRatand88
Mouse Diet, Bantin & Kingham, Hull, UK). Euthanasia was carried out by schedule 189
procedureofCO2asphyxiationfollowedbycervicaldislocationorexsanguination.3-6mice90
wereusedperstrain,pertimepointstudied.91
Parasitesandinfection92
Professor Kathryn Else, The University of Manchester, kindly provided eggs of T. muris93
Edinburgh (E) isolate for use in all infection studies. Egg infectivity and maintenance of94
parasite stocks were carried as described byWakelin, 1967 [8]. Experimental mice were95
infectedwith200embryonatedeggsin200μlofultra-puredistilledwaterviaoralgavage.96
Wormburdenwas assessed at day 21 PI. Caecum and proximal colonwere harvested at97
autopsytodetermineparasiteclearanceofeachmouseat theendofeachexperimentas98
describedbyElseetal.,1990[9].99
Humansamples100
Prior to thecommencementof theclinical study,NHSethicsapprovalwasobtained from101
BerkshireBResearchEthicsCommittee(RECreferencenumber:14/SC/1413;IRASreference102
number:157778) inordertoscreenclinical IBDsamplesof faecesandserum.Participants103
were recruited from the Salford Royal NHS Foundation Trust. Normal healthy volunteers104
wererecruited inaccordancewith theUniversityethicscommitteeandtheHumanTissue105
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
Act2004.Faecalsamplesweretakenfromhealthycontrols(n=10)withnopriorhistoryof106
IBDorgutproblems,orpatients (n=31)withsuspected IBDorclinicallyconfirmed IBD.All107
patientsamplesweretakenviaoutpatientclinics,returnedbypatientsaspartofstandard108
clinical practice to be assessed for Faecal Calprotectin (FCP). Colonoscopy/biopsy were109
undertakeninthosewithelevatedFCP.Ofthepatients,6patientshadIBDexcluded,mainly110
leadingtoaclinicaldiagnosisofirritablebowelsyndrome(IBS),19knownIBDpatientswere111
inremissionattheoftimetesting(10ulcerativecolitisand9Crohn’sdisease)and6patients112
hadactiveIBD(n=5CD,n=1UC)atthetimeoftesting.113
Statisticsandanalysis114
Where statistics arequoted, experimental groupswere comparedusing linear regression,115
Mann-WhitneyUtestortwo-wayanalysisofvariance(ANOVA)testfollowedbySidak’spost116
hoc multiple comparisons test, where appropriate. P values <0.05 were considered117
significant.Dataarepresentedasmean±SEMunlessotherwisestated.Statisticalanalyses118
were carried out using GraphPad Prism 7 (GraphPad Software, La Jolla, California, USA;119
www.graphpad.com). 120
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
7
Results121
EarlyimmuneresponseinformsresistancetoT.muris-inducedcolitis122
Following challengewithTrichurismuris, asexpectedBALB/cmiceexpelledmostorallof123
thewormsby21daysPI,whereasAKRmicewereunabletoexpelallwormsandremained124
infectedwith a significantly higherwormburden (P=0.016,Mann-WhitneyU test) (Figure125
1A).Colitisscoringrevealedincreasedhistologicalchangesassociatedwithinflammationin126
both AKR and BALB/c mice after infection (Figure 1B). These changes included influx of127
immunecells,presenceofimmunecellsinthesubmucosa,crypthyperplasiaandgobletcell128
loss.Inagreementwithpreviousdata,thecolitisscoresinBALB/cmicepeakedat21daysPI129
andhadbegun to return tonormalby31daysPI.Asexpected, colitis scores inAKRmice130
roseafterinfectionandpeakedat31daysPIwherethecolitisscorewassignificantlygreater131
than that of the BALB/c mice (P=0.046, ANOVA; Figure 1B). Representative images of132
haematoxylinandeosinstainedproximalcolonsectionsinnaïvemiceandat31daysPIare133
shown in Figure1C. Collectively, these results reproducepreviouslypublished research in134
the AKR/BALB/c infection model [4], where BALB/c mice initiate an acute, resolving135
inflammationafterT.murischallengeandAKRmice showdelayed immune response that136
resultsinachronicinflammatoryphenotypeduetoafailuretoexpelworms.137
Weinvestigatedimmunecellrecruitmenttothecoloniclaminapropriabyflowcytometryat138
1,7,14,21and31daysPI.BALB/cmicehadanacuteresolvingresponsewhereasAKRmice139
developedachronicinflammatoryresponse.At24hoursPIBALB/cmicerespondedrapidly140
toT.muris challenge,withanearly increase in theproportionsofDCs,macrophagesand141
neutrophils(P<0.05,ANOVA)incoloniclaminapropriatissuesandmesentericlymphnodes142
compared to naïve mice (Figure 1D and 1E, Supplementary data). BALB/c and AKR had143
differentresponsestoinfection,withBALB/ctendingtohavegreaterrecruitmentofinnate144
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
immune cells at D1 compared with susceptible AKR mice. This trend for a greater early145
magnitudeofresponseintheBALB/ccomparedwithAKRwasalsoseenatD7,D14andD21146
post-infectionwiththegreatestdifferencesbetweentheimmuneresponseofBALB/cmice147
to AKR at D14 PI (Figure 1B, Supplementary data). However, by D31, proportions of148
macrophages(P<0.001),inflammatorymonocytes(P<0.01)andneutrophils(P<0.001),were149
all significantlygreater inAKRmice,whereas theproportionsof thesecell types returned150
neartobaselinelevelsinBALB/cmice(Figure1D-G).Theincreaseinimmunecellsobserved151
inAKRmiceatD31correspondedwiththepeakcolitisscore,andlikewisethereductionin152
immunecellsinBALB/cmiceatD31wasparalleledwithareductionincolitisscore(Figure153
1B). Collectively thiswork indicated an altered dynamic of immune response in resistant154
versus susceptible mice, therefore we explored the early transcriptome in order to155
understandchangespresentbetweencolitis-susceptibleandcolitis-resistantmice.156
TranscriptionalchangesinducedbyT.murisinfectionidentifiedat24hourspostinfection157
Transcriptionalchangesintheproximalcolon(theprincipalsiteofT.muris infection)were158
investigatedat24hourspost-infectionviamicroarray,prior to theestablishmentofovert159
signsofinflammation.Geneswiththelargestdifferentialexpressionat24hourspostwere160
calculated(Figure2A).Ofthe77probesetsthatweresignificantlyupregulatedinAKRmice161
anddownregulated inBALB/cmice(1-IPPLR<0.05),65weresuccessfullymatchedtogene162
IDs in DAVID. The gene upregulated in colitis-susceptible AKR mice with the largest163
differentialexpressioncompared toBALB/cmicewas the receptor foradvancedglycation164
end-products(Rage)(Log2foldchange=2.0718;1-IPPLR=0.003)(Figure2B;Supplementary165
data).UpregulationofRageinsusceptiblemicewasconfirmedbyqPCRofproximalcolonat166
24hoursPIinanindependentexperiment(P<0.001,Mann-WhitneyUtest,Figure2C).167
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
9
DifferentialexpressionoftranscriptionfactorswasanalysedusingTIGERiforMATLAB[10].168
Themostnotablechangeintranscriptionfactorgeneexpressionwasthedownregulationof169
FoxO4 (Forkhead boxO4) followingT.muris infection in AKRmice (Supplementary data).170
FOXO4occursdownstreamof theRAGEactivationsignallingcascadeandserves to inhibit171
DNA binding and transcriptional activity of NF-κB (nuclear factor kappa-B), preventing172
inflammation[11].FoxO4isalsodownregulatedinthecolonicepithelialcellsofIBDpatients173
[11]. TheupregulationofproinflammatoryRageanddownregulationofanti-inflammatory174
FoxO4fromtheRAGEsignallingpathwayprovidecompellingevidencefortherelevanceof175
RAGEactivationincolitissusceptibilityinAKRmiceduringT.murisinfection.176
IdentifyingthecellularsourceofRAGE177
Over90%ofCD326+(EpCAM)epithelialcellsexpressedRAGE(Figure3A)andtheyfellinto178
twodistinctgroups,expressingeitherlow(RAGElo)orhigh(RAGEhi)levelsofRAGE.Innaïve179
mice, thetotalproportionofCD326+epithelialcells thatexpressedRAGEwassignificantly180
higherincolitis-susceptibleAKRmice(P<0.01,ANOVA).TheproportionofRAGElotoRAGEhi181
cellswassimilar inbothnaïveAKRandBALB/cmice,buttherewassignificantdrop inthe182
proportionofRAGEhiepithelialcellsobserved2and7daysPI (P<0.0001,ANOVA), inboth183
AKRandBALB/cmice(Figure3B).184
ImmunohistochemistrywasusedtoconfirmexpressionofRAGEintissue(Figure3C).Akinto185
the flow cytometry data, we saw high expression of RAGE throughout the colonic186
epithelium.Concurringwiththeflowcytometrydata,therewasonlyminimalfluorescence187
seen in the immune lamina propria cells indicating that epithelial cells indeed express188
greateramountsofmembrane-boundRAGEthanimmunecells.Thisdatasuggestthatthe189
epithelialcellsarelikelytobethesourceoftheobservedincreaseinRagemRNA.Therewas190
areductioninintensityofRAGEstainingat7daysPIbyimmunohistochemistry(Figure3C)191
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
compared tonaïvemice,which also correlatedwith themeasured shift in proportionsof192
epithelialcellsfromRAGEhitoRAGElocellsmeasuredbyflowcytometry.193
IsRAGEdifferentiallycleavedincolitis-susceptibleandcolitis-resistantmice?194
RAGE may be internalised after ligand binding, or released as soluble RAGE (sRAGE) via195
enzymaticcleavagebyADAM10orMMP9[12,13].ToinvestigatewhetherRAGEwasbeing196
cleavedweassessedsRAGElevelsinserumandthefaecesbyELISA.SerumsRAGElevelsin197
susceptible mice remained constant throughout the experiment. Resistant mice had198
significantly higher serum sRAGE than susceptiblemice both prior to infection and at 24199
hours PI (P<0.001, ANOVA; Figure 3D). sRAGEwas also detectable in faeces,where both200
resistant and susceptiblemice had very low levels of sRAGE prior to infection. Over the201
courseofinfection,faecalsRAGEinBALB/cmiceincreasedtosignificantlyhigherlevelsthan202
susceptiblemiceat24hoursand21daysPI(P<0.01,ANOVA;Figure3E).FaecalsRAGEwas203
notdetectedatalluptoD21insusceptiblemice.204
CorrelationofserumandfaecalsRAGElevelstocolitisscoreshighlightsthechanginglevels205
of sRAGE in BALB/cmice during the course of T. muris infection relative to pathological206
changesinthecolon(Figure3F-G).TheincreasedcirculatingserumsRAGEatD0inBALB/c207
mice,where colitis scores are lowest, drops as colitis increases at D1 andD21 (R2=0.90).208
Faecal sRAGE in the BALB/c mice increases relative to colitis scores (R2=0.99). However,209
sRAGE levels in susceptiblemicedidnot changeduring thecourseof infection relative to210
increasingcolitisscoresfromD0toD21post-infection(serumR2=0.99,faecalR2=0.73).211
We then investigated levels of the RAGE ligand S100A8 (one part of the heterodimeric212
calprotectinprotein,currentlyusedasaclinicalbiomarkerfor IBD) inserumandfaecesas213
an indicator of whether sRAGE might be quenching the proinflammatory effects of214
circulating RAGE ligands by acting as a decoy receptor. Serum and faecal S100A8 did215
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
11
increase slightly during the course of infection in both AKR and BALB/c mice, but no216
statisticaldifferenceswereobservedbetweenthetwostrainsandtherewashighvariability217
betweenmice.At21daysPIBALB/cmicehadgreaterlevelsofserumS100A8thanAKRmice218
(not significant, ANOVA; Supplementary data). Faecal S100A8 remainedbroadly similar in219
naïveandT.muris infectedmiceofbothAKRandBALB/c strains.Aswith serumS100A8,220
faecalS100A8wasslightly raised inBALB/cmiceat21daysPIcomparedtoAKRmicebut221
thiswasnotsignificant(ANOVA;(Supplementarydata)).Inadditiontobeinghighlyvariable,222
S100A8correlatedpoorlywithcolitisscores(Supplementarydata).223
sRAGEisdifferentiallyexpressedinIBD224
As the differences in sRAGEweremost apparent and consistent in faeceswe focusedon225
analysis of faecal specimens from healthy volunteers and patientswith IBD or suspected226
IBD. sRAGEwas not detected in the faecal samples of healthy volunteers. In contrast, s-227
RAGEwasdetectable in thepatient cohort (Figure4A). Thehighest levels of sRAGEwere228
seen in patients with IBD excluded (largely IBS ascribed) and IBD in remission, although229
remission patients were more variable. Patients with active IBD characterised by severe230
inflammationhadlowlevelsofsRAGEandincreasedcalprotectin.231
WethentransformedthefaecalRAGEandcalprotectinELISAdataintopresent(1)orabsent232
(0),where protein is scored as present if ≥ 3*SD abovebaseline.Whenpatient datawas233
stratifiedintogroups(activeIBD,IBDinremission,IBDexcluded(IBS)andhealthycontrols)234
theratioofRAGEtocalprotectinclearlyidentifiedhealthycontrols(0,0)andactiveIBD(0,235
1) from IBS and IBD in remission (Figure 4B). In healthy individuals,we sawno sRAGEor236
calprotectinsignalinanysubject.InactiveIBDwehadaconsistentpatternofCalprotectin237
present,butnosRAGE.Forpatientswhosediseaseisresolvingwesawamorecomplicated238
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
picture,wherewehad a subset of patients undergoing routine test thatwere expressing239
bothsRAGEandCalprotectin. 240
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
13
Discussion241
Animalmodelsarecrucialforexaminingthecausativeeventsthatleadtodiseasessuchas242
chronic colitis. Genetic factors influence the likelihood of developing colitis, but it is243
impossible to continually monitor individuals with potential genetically susceptibility in244
order to identify the pathways that drive the development of chronic intestinal245
inflammation.Models that accurately simulate preclinical changes in the gut allow us to246
interrogatethepathwaysleadingtochroniccolitis,priortothedevelopmentofthecomplex247
inflammatoryenvironmentwhendisease isestablished.Levisonetal. [14]havepreviously248
shownthatT.muris infectioninAKRmicecausescolitisthatcorrelatesphenotypicallyand249
transcriptionallywiththeprofileofhumanCD.Here,weprovidenovelevidencereinforcing250
the use of the T. muris infection as a model for the discovery of preclinical intestinal251
inflammationmarkersthattranslateintohumanIBDpatients.252
In line with previous T. muris infection studies, we observed altered dynamics of the253
immune response between colitis resistant versus susceptible mice, characterised by an254
early influxofDCs [5].Wethen identifiedupregulationofRageasapotential indicatorof255
colitissusceptibilityinmice.RAGEactivityhasalreadybeenlinkedtoactiveIBD[15]aswell256
asotherinflammatorydiseasesincludingdiabetes,Alzheimer’s,airwayinflammation,cancer257
and haemorrhagic shock [16]. Additionally, several RAGE ligands have been identified as258
associatedwiththe inflammation in IBD, includingcalprotectin,EN-RAGEandHMGB1[17-259
19].Thus,ourdatafromthemousemodelshowsaclearlinktoknownpathologyinIBD.Itis260
important to validate results from mouse model to human disease and therefore we261
conducted a small validation study to assess faecal sRAGE. Faecal sRAGE was readily262
detected in patient’s faecal samples. Akin to the mouse data we saw lower sRAGE in263
patients with active chronic inflammation. Surprisingly, symptomatic patients with IBD264
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
excluded on a basis of normal FCP and/or colonoscopy had higher levels of sRAGE than265
healthy volunteers. Largely, IBSwas clinically ascribed but thatwas not based on formal266
diagnostic criteria, and other diagnoses such as bile acid diarrhoea ormicroscopic colitis267
werenot formally excluded. The sample sizewas small andmoreprospective studies are268
neededtoconfirmthispreliminaryobservation.PatientswhoseIBDwasreportedtobe in269
remissionhadvariablelevelsofsRAGE.ItistemptingtospeculatethatlowerlevelsofsRAGE270
are associatedwith a risk of subsequent flare of inflammation but aswe only had single271
samplesfromthepatients,wecannotassessthis,howeveritwouldbeinterestingtotrack272
sRAGEovertimeinaprospectivestudyofpatientswithIBD.273
RAGEhasbeendescribedonseveralimmunecellswith,forexample,neutrophilsidentified274
asexpressing largeamountsofRAGE[20].Ourflowcytometryandimmunohistochemistry275
analysisofRAGEexpressioninmultiplecelltypespresentinandaroundthelaminapropria276
andcryptsofthecolon,however,didnotsuggestthatimmunecellswerethemainsources277
of cellularRAGE.Ourdata in fact showed thatamajor cellular sourceofRAGE in thegut278
were the gut epithelial cells. Previous studies have shown that epithelial cells not only279
express RAGE, but also upregulate RAGE expression during colonic inflammation [15].280
Changesweobserved in the levelsofRAGEexpressionsuggest that it is theepithelialcell281
response to T. muris infection that informs subsequent susceptibility to chronic282
inflammation.ThereductionintheamountofRAGEpresentatthecellmembranewesaw283
by flowcytometryandby immunohistochemistry immediately followingT.muris infection284
couldbecausedbyeither internalisationofactivatedRAGE-ligandcomplexesorADAM10-285
mediatedsheddingtoproducesRAGE[21,22].SplicevariantsofRagemayalsoresult ina286
truncatedRAGEmoleculeoramodifiedandactivelysecreteddecoyreceptor[23,24].The287
process by which epithelial cells may undergo RAGE shedding represents an important288
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
15
distinction in the course of gut immunity and homeostasis, and may be an essential289
component in dictating whether inflammation becomes chronic or resolves. Further290
investigation into the extent towhich splice variants, internalisation or sheddase activity291
form the mechanism for changes in sRAGE in both the Trichuris model and in human292
patientscouldprovidefurtherinsightfortheroleoftheRAGEpathwayincolitis.293
Activation of RAGE can have several outcomes including immune cell migration and294
transcriptionofpro-inflammatorycytokines.RAGE-mediatedleukocytemigrationviaCD11b295
(Mac-1)isinvolvedinmigrationofimmunecellstothesiteofinjuryandhomingofDCsto296
thelymphnodes[15,25].However,whethertheupregulationofRageiscrucialtofacilitate297
the early DC migration in the T. muris model remains unclear. While we observed298
differencesinDCmigrationasearlyasday1PI,wesawnodifferencesbetweencellsurface299
RAGE expression between AKR and BALB/c mice that might account for the altered300
dynamicsofDCrecruitment.Similarly,RAGEhasbeenlinkedtoneutrophilrecruitmentbut301
weonlyobservedmodestneutrophil infiltration in the first24hoursPIandnodifference302
betweencolitis-susceptibleorcolitis-resistantmice.Despiteminimaldifferencesinimmune303
cell presence during the early stages of T.muris infection, prolonged activation of RAGE304
resultsinactivationofinflammatorysignallingmoleculesincludingNF-κBandMAPkinases305
[16].Consequently,anenvironmentwhereRAGEligandssuchasHMGB1,andS100proteins306
are continually present results in perpetual NF-κB activation and subsequent chronic307
inflammatoryconditions.308
WeobservedstrikingdifferencesinthelevelsoffaecalandsystemicsRAGEbetweencolitis-309
resistantandsusceptiblemice,withBALB/cmicerapidlyproducingsRAGEinresponsetoT.310
murisinfection.sRAGEeffectivelyactsasadecoyreceptorforRAGEasitcanstillbindtothe311
samedamageinducedligandsasmembraneboundRAGEbut,asitlacksacytoplasmictail,it312
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
cannot initiate the pro-inflammatory signalling cascade [26]. Thus, higher levels of sRAGE313
mightbeexpected toblock inflammation and indeed reduced levels of sRAGEhavebeen314
found in mice with chronic inflammation as well as patients suffering from chronic315
inflammatory diseases [27]. The reduction in epithelial RAGE expression followed by316
increases in circulating sRAGE in colitis-resistantmice suggests sheddingof RAGE to form317
sRAGEasaprotectivefeedbackprocessagainstthedevelopmentofchronic inflammation.318
Colitis-susceptibleAKRmiceshowthesamereductioninepithelialRAGEexpressionbutdid319
notproducesRAGEinthesamequantities,suggestinginternalisationandactivationofRAGE320
andinitiationofsubsequentproinflammatorypathwaysafterT.murisinfection.Indeed,itis321
known thatT.muris excretory/secretory (E/S) products induceNF-κB signalling in colonic322
epithelial cells shortly after infection and the susceptible immune response toT.muris is323
associated with expression of the T helper (TH) 1 cytokines interferon (IFN)-γ, tumour324
necrosisfactor(TNF)-αandIL-12[28].Thisresponseparallelsthecytokinesexpressedafter325
RAGEactivation,whichincludeTH1andTH17cytokinesTNF-α,IL-1α,IL-6,IL-8andIL-12[16].326
DiagnosisofIBDusuallyinvolvesanassessmentofclinicalhistoryandphysicalexamination,327
with endoscopy and histology considered to be the gold standard tools [29]. Accurately328
assessingdiseaseactivityremainsdependentoncolonoscopyand/orsmallbowel imaging.329
The invasivenessofcurrentdiagnosticmethods isnot idealandrecentworkhasaimedto330
identify serum or faecal biomarkers that can reliably identify active disease [30]. The331
number of potential IBD biomarkers is high, but there remains a lack of reliable and332
reproduciblebiomarkersforuseinclinicalpractice[31].Efficacyofcurrenttherapiesisalso333
variable,withrisksofsometimesserioussideeffects,especially infectionmeaningthere is334
considerable interest fornewbiomarkersandnewtherapeutics [32].Calprotectinentered335
clinical practice as an IBD biomarker to aid clinical diagnosis non-invasively, but336
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
17
measurementsoffaecalcalprotectinarevariableandthereis littleagreementaboutwhat337
shouldbeconsideredanormalbaselinelevelinhealthypatients[33].Indeed,theconceptof338
a simple normal cut off is impossible to entertain given the enormous heterogeneity in339
faecalwatercontent,matrixcomposition,transittime,siteandextentofinflammationand340
the contact of faecal component sampled with the mucosa; composite measures are341
essential. Calprotectin is a product of tissue damage and binds to RAGE to promote342
inflammation,butthisactionwillbereducedinthepresenceofthedecoyreceptorsRAGE.343
OurpreliminaryobservationofalterationsinsRAGEexpressioninmouseandhumandisease344
suggests theremay bemerit in looking at both calprotectin and sRAGE to better predict345
whether calprotectin and other RAGE ligands are indeed able to drive pro-inflammatory346
signals.Thismayimprovethereliabilityofcalprotectinasabiomarker.347
Our pilot clinical study successfully validated the use of the T. muris infection model as348
highly translatable to human IBD states. TheT.murismodel is a useful tool in dissecting349
early pathways that are involved in the onset of colitis. By using a mouse model and350
focusing on early initiating events in the development of colitis, we have identified a351
potential role forRAGE inmediating thedevelopmentof inflammation. Furthermore, our352
observationofhighlevelsofsRAGEinacuteresolvinginflammationsuggestedtheremaybe353
utility inmonitoring of sRAGE tomonitor IBD. However, a larger clinical study would be354
requiredtoinvestigatethisfurther.355
Acknowledgements356
Funding for this researchwasprovidedby theEPSRC,EpistemLtd.and theBBSRC impact357
accelerator fund. NHS Salford Trust provided the human clinical samples for this project,358
andwearegratefultoallclinicalstaff,patientsandhealthyvolunteerswhotookpartinthe359
trial.360
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
The Bioimaging Facilitymicroscopes used in this study were purchasedwith grants from361
BBSRC,Wellcome and the University ofManchester Strategic Fund. Special thanks go to362
RogerMeadowsfrom the Bioimaging Facility for his helpwith themicroscopy and image363
acquisition. 364
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
19
References365
1. Baumgart, D.C. and S.R. Carding, Inflammatory bowel disease: cause and366immunobiology.Lancet,2007.369(9573):p.1627-40.367
2. Kaplan,G.G.,Theglobalburdenof IBD: from2015 to2025.NatRevGastroenterol368Hepatol,2015.12(12):p.720-7.369
3. Liverani, E., et al., How to predict clinical relapse in inflammatory bowel disease370patients.WorldJGastroenterol,2016.22(3):p.1017-33.371
4. Levison, S.E., et al.,Genetic analysis of the Trichurismuris-inducedmodel of colitis372revealsQTLoverlapandanovelgenecluster forestablishingcolonic inflammation.373BMCGenomics,2013.14:p.127.374
5. Cruickshank, S.M., et al., Rapid dendritic cell mobilization to the large intestinal375epithelium is associated with resistance to Trichuris muris infection. Journal of376Immunology,2009.182(5):p.3055-62.377
6. Bowcutt, R., et al.,A role for the pattern recognition receptor Nod2 in promoting378recruitment of CD103+ dendritic cells to the colon in response to Trichuris muris379infection.MucosalImmunol,2014.7(5):p.1094-105.380
7. Foell,D.,H.Wittkowski, and J.Roth,Monitoringdiseaseactivityby stoolanalyses:381fromoccultbloodtomolecularmarkersofintestinalinflammationanddamage.Gut,3822009.58(6):p.859-68.383
8. Wakelin, D.,Acquired immunity to Trichurismuris in the albino laboratorymouse.384Parasitology,1967.57(3):p.515-24.385
9. Else,K.J.,etal.,Theinfluenceofgenesmappingwithinthemajorhistocompatibility386complexonresistancetoTrichurismurisinfectionsinmice.Parasitology,1990.101Pt3871:p.61-7.388
10. Han,N.,H.A.Noyes,andA.Brass,TIGERi:modelingandvisualizingtheresponsesto389perturbationofa transcription factornetwork.BMCBioinformatics,2017.18(Suppl3907):p.260.391
11. Zhou,W.,etal.,FoxO4 inhibitsNF-kappaBandprotectsmiceagainstcolonic injury392andinflammation.Gastroenterology,2009.137(4):p.1403-14.393
12. Curran,C.S.andP.J.Bertics,HumaneosinophilsexpressRAGE,produceRAGEligands,394exhibit PKC-delta phosphorylation and enhanced viability in response to the RAGE395ligand,S100B.IntImmunol,2011.23(12):p.713-28.396
13. Zhang,L.,etal.,Receptorforadvancedglycationendproductsissubjectedtoprotein397ectodomain sheddingbymetalloproteinases. JBiolChem,2008.283(51):p.35507-39816.399
14. Levison,S.E.,etal.,Colonictranscriptionalprofilinginresistanceandsusceptibilityto400trichuriasis: phenotyping a chronic colitis and lessons for iatrogenic helminthosis.401InflammBowelDis,2010.16(12):p.:2065-79.402
15. Body-Malapel, M., et al., The RAGE signaling pathway is involved in intestinal403inflammationandrepresentsapromisingtherapeutictargetforInflammatoryBowel404Diseases.MucosalImmunol,2019.12(2):p.468-478.405
16. Sparvero, L.J., et al., RAGE (receptor for advanced glycation endproducts), RAGE406ligands,andtheirroleincancerandinflammation.JournalofTranslationalMedicine,4072009.7.408
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
17. Whitehead, S.J., et al., Effect of faecal calprotectin assay variability on the409managementofinflammatoryboweldiseaseandpotentialroleoffaecalS100A12.J410ClinPathol,2017.70(12):p.1049-1056.411
18. Chen,X.,etal.,HMGB1exacerbatesexperimentalmousecolitisbyenhancinginnate412lymphoid cells 3 inflammatory responses via promoted IL-23 production. Innate413Immun,2016.22(8):p.696-705.414
19. Aranda,C.J.,etal.,Calprotectinprotectsagainstexperimentalcolonic inflammation415inmice.BrJPharmacol,2018.175(19):p.3797-3812.416
20. Huebener, P., et al., The HMGB1/RAGE axis triggers neutrophil-mediated injury417amplificationfollowingnecrosis.JClinInvest,2019.130:p.1802.418
21. Sevillano, N., et al., Internalization of the receptor for advanced glycation end419products (RAGE) is required to mediate intracellular responses. J Biochem, 2009.420145(1):p.21-30.421
22. Yang,W.S.,etal.,EctodomainSheddingofRAGEandTLR4asaNegativeFeedback422Regulation in High-Mobility Group Box 1-Activated Aortic Endothelial Cells. Cell423PhysiolBiochem,2018.51(4):p.1632-1644.424
23. Hudson,B.I.,etal., Identification,classification,andexpressionofRAGEgenesplice425variants.FASEBJ,2008.22(5):p.1572-80.426
24. Jules, J., D.Maiguel, andB.I.Hudson,Alternative splicing of theRAGE cytoplasmic427domainregulatescellsignalingandfunction.PLoSOne,2013.8(11):p.e78267.428
25. Manfredi,A.A.,etal.,MaturingdendriticcellsdependonRAGEforinvivohomingto429lymphnodes.JImmunol,2008.180(4):p.2270-5.430
26. Basta,G.,etal.,Circulatingsolublereceptor foradvancedglycationendproducts is431inversely associated with glycemic control and S100A12 protein. J Clin Endocrinol432Metab,2006.91(11):p.4628-34.433
27. Maillard-Lefebvre,H.,etal.,Solublereceptorforadvancedglycationendproducts:a434new biomarker in diagnosis and prognosis of chronic inflammatory diseases.435Rheumatology,2009.48(10):p.1190-1196.436
28. deSchoolmeester,M.L., H.Manku, and K.J. Else, The innate immune responses of437colonicepithelialcellstoTrichurismurisaresimilar inmousestrainsthatdevelopa438type1ortype2adaptiveimmuneresponse.InfectImmun,2006.74(11):p.6280-6.439
29. Halfvarson, J.,Geneticepidemiologyof inflammatoryboweldisease,early twinand440familystudies,inMoleculargeneticsofinflammatoryboweldisease,M.D'amatoand441J.D.Rioux,Editors.2013,Springer:NewYork.p.23-43.442
30. Lehmann, F.S., E. Burri, and C. Beglinger, The role and utility of faecalmarkers in443inflammatoryboweldisease.TherapAdvGastroenterol,2015.8(1):p.23-36.444
31. Stevens, T.W., et al., Systematic review: predictive biomarkers of therapeutic445responseininflammatoryboweldisease-personalisedmedicineinitsinfancy.Aliment446PharmacolTher,2018.48(11-12):p.1213-1231.447
32. Nanau,R.M.,L.E.Cohen,andM.G.Neuman,Riskofinfectionsofbiologicaltherapies448withaccentoninflammatoryboweldisease.JPharmPharmSci,2014.17(4):p.485-449531.450
33. Dhaliwal,A.,etal.,Utilityoffaecalcalprotectinininflammatoryboweldisease(IBD):451whatcut-offsshouldweapply?FrontlineGastroenterol,2015.6(1):p.14-19. 452
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
21
Figurelegends453
Figure1:Colitis-susceptibleAKRmiceshowdelayedexpulsionofTrichurismuriswormsat454
21daysandincreasedevidenceofcolitisat31dayspostinfection.(A)Meanwormburden455
(±SD)at21dayspostinfection(PI).(B)Cumulativecolitisscore(0-20)basedonthegrading456
ofhistologicalchangesincludingcryptelongation(score0-4),depletionofgobletcells(score457
0-4), thickness of muscle wall (score 0-4), inflammatory cell infiltration (score 0-4) and458
destructionofarchitecture(score0or3-4).(C)Representativeimagesofhaematoxylinand459
eosinstainedproximalcolonsectionsfromnaïvemiceandat31daysPI;notethehighlevels460
of immunecell infiltrationand lossofgobletcells inthecolonictissuesofAKRmiceat31461
dayspost-infection.Bar=50μm.n=3-5micepertimepoint.AnalysisbyMann-WhitneyUtest462
ortwo-wayANOVAfollowedbySidak’smultiplecomparisonstestwhereappropriate.(D-G)463
Dendriticcell (CD45+MHCII+CD11c+F4/80-CD103+/-CD11b+/-),macrophage(CD45+MHCII+464
F4/80+ CD11c+/-), inflammatory monocyte (CD45+ Ly6G+ CD11b+ CD115+) and neutrophil465
(CD45+Ly6G+CD11b+CD115-)populationsasproportionofCD45+cells(±SEM)innaïvemice466
andduringT.muris challenge.n=3miceper timepoint.Analysisby two-wayANOVAwith467
Sidak’smultiplecomparisonsposthoctest.***P<0.001,**P<0.01,*P<0.05. 468
469
Figure2:Geneexpressionchangesinproximalcolon24hourspost-infectionwithTrichuris470
muris.(A)GenesmostsignificantlyupregulatedinAKRmiceanddownregulatedinBALB/c471
mice(red)ordownregulatedinAKRmiceandupregulatedinBALB/cmice(blue)following472
24hourTrichurismurisinfection.(B)SchematicofthestructureofRAGE,showingactivating473
ligands,downstreamNF-κBactivationandformationofsolubleRAGE(sRAGE)byADAM10474
orMMP9cleavage.(C)mRNAexpressionofRAGEintheproximalcolonatD1post-infection475
asmeasuredbyqPCR.DatageneratedusingAffymetrixMouse4302.0microarraysanalysed476
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
usingthepumaandTIGERi(TFAillustratorforglobalexplanationofregulatoryinteractions)477
packages for Bioconductor. n=4-5 mice per group. Analysis by Mann-Whitney U-test.478
**P<0.01. 479
480
Figure3:Receptorforadvancedglycationend-products(RAGE)expressioninthecolonof481
Trichurismuris infectedAKRandBALB/cmice (aged6-8weeks). (A) ProportionofRAGE482
expressing epithelial cells (CD326+, ±SEM) during early Trichuris muris infection. (B)483
ProportionofcolonicepithelialcellsexpressinghighlevelsofRAGEisreducedshortlyafter484
infection, as measured by flow cytometry. (C) Representative images of colon sections485
stained for RAGE (FITC; green) and nuclei (DAPI; blue) in naïvemice and at 7 days post-486
infection(Bar=100μm; insetbar=22μm). (D-E)sRAGEpresent inserumorfaecesduring487
TrichurismurisinfectionasmeasuredbyELISA.(F-G)CorrelationofserumandfaecalsRAGE488
versuscolitisscoreat0,1and21dayspost-infection.n=3-5micepertimepoint.Analysisby489
linearregression,two-wayANOVAwithSidak’sposthoctest.**P<0.01****P<0.001. 490
491
Figure 4: Soluble receptor for advanced glycation end-products (sRAGE) is detectable in492
thefaecesandserumofIBDpatients.(A)ScatterplotofsRAGE(pg/ml)versuscalprotectin493
(μg/ml)presentinthefaecesofpatientswithactiveIBD,IBDinremission,IBDexcluded(IBS)494
compared to healthy controls. (B) Relative levels of faecal sRAGE versus calprotectin in495
human IBD/IBS or healthy controls. Datawere transformed to arbitrary units (AU)where496
samplesgreaterthan3SDfrombaseline=1(present),otherwisetheyscored0(absent).497
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
AKR BALB/c
D0
D31
A
B
C
D E
F G
DCs
0 1 7 14 21 310
5
10
15
20
25
Time post-infection
% o
f CD
45+
cells
AKRBALB/c
***
Macrophages
0 1 7 14 21 3105
1015202530354045
AKR
BALB/c
*********
Time post-infection
% o
f CD
45+
cells
*
Inflammatory monocytes
0 1 7 14 21 310
10
20
30
40AKRBALB/c
*** ***
**
Time post-infection
% o
f CD
45+
cells
Neutrophils
0 1 7 14 21 310
10
20
30
40
Time post-infection
% o
f CD
45+
cells
AKRBALB/c *** *** ***
Figure 1
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
Rag
eG
m35
79C
dhr4
Am
hr2
Ldhc
Myc
nM
bd1
Pla
2g2a
Gra
spA
pobe
c2Fa
ds2
Cdc
26K
pnα1
Ttll1
0Iff
o1C
ops3
Rha
gU
pk1a
Ptg
esM
eox2
Art3
Ptg
er3
Slc
25a1
2M
roh2
aE
sco2
Soc
s7M
rap
Rab
14E
mb
Lep
Itsn1
A53
0053
G22
Rik
D12
Ertd
123e
Orm
1Iκ
bkβ
Hp
-3
-2
-1
0
1
2
3
Log2
Fol
d ch
ange
A
B C
Figure 2
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
0 2 4 6 80
20000
40000
60000
Colitis Score
Ser
um s
RA
GE
(pg/
ml)
Serum sRAGE vs colitis score
AKR BALB/c
D0
D7
A
B
C
D
0 2 4 6 80
5000
10000
15000
Colitis Score
Faec
al s
RA
GE
(pg/
ml)
Faecal sRAGE vs colitis score
AKR
BALB/c
E
F G
0 2 70
20
40
60
80
100
Timepoint post-infection
% o
f CD3
26+
cells
Epithelial cell RAGE expression
AKRBALB/c
**
0 2 70
20
40
60
80
100
Timepoint post-infection
% o
f RAG
E+CD
326+
cel
ls
RAGEhi epithelial cells
AKRBALB/c
Figure 3
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint
A B
sRAGE Calprotectin0.0
0.5
1.0
Aver
age
AU
Active IBDIBD remission IBD excludedHealthy
0 200 400 6000
500
1000
1500
sRAGE pg/ml
Calp
rote
ctin
μg/
ml
Active IBDIBD remissionIBD excludedHealthy
Figure 4
not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted August 5, 2019. ; https://doi.org/10.1101/719310doi: bioRxiv preprint