Changes in caecal microbiota and mucosal gene expression revealed new aspects of Epizootic rabbit Enteropathy

Changes in Cecal Mic obio a and Mucosal Gene Exp ession Re ealed New Aspec s o Epizoo ic Rabbi En e opa hy Ch is ine Ba ¨ue l 1. , M. Ca men Collado 1. , Manuel Zu ´n ˜iga 1 , En ique Blas 2 , Gaspa Pe ´ ez Ma ı ´nez 1 * 1Labo a o io de Bac e ias La ´c icas y P obio ´ icos, Depa amen o de Bio ecnologı ´a, Ins i u o de Ag oquı ´mica y Tecnologı ´a de Alimen os (IATA), Consejo Supe io de In es igaciones Cien ı ´ icas (Spanish Na ional Resea ch Council) (CSIC), Valencia, Spain, 2Animal Nu i ion Resea ch G oup, Ins i u e o Animal Science and Technology, Poly echnic Uni e si y o Valencia (UPV), Valencia, Spain Abs ac Epizoo ic Rabbi En e opa hy (ERE) is a se e e disease o unknown ae iology ha mainly a ec s pos -weaning animals. I s incidence can be p e en ed by an ibio ic ea men sugges ing ha bac e ial elemen s a e c ucial o he de elopmen o he disease. Mic obial dynamics and hos esponses du ing he disease we e s udied. Cecal mic obio a was cha ac e ized in h ee abbi g oups (ERE-a ec ed, heal hy and heal hy p e ea ed wi h an ibio ics), ollowed by ansc ip ional analysis o cy okines and mucins in he cecal mucosa and e mix by q- PCR. In heal hy animals, cecal mic obio a wi h o wi hou an ibio ic p e ea men was e y simila and domina ed by Alis ipes and Ruminococcus. P opo ions o bo h gene a dec eased in ERE abbi s whe eas Bac e oides,Akke mansia and Rikenella inc eased, as well as Clos idium,c-P o eobac e ia and o he oppo unis ic and pa hogenic species. The ERE g oup displayed ema kable dysbiosis and educed axonomic di e si y. T ansc ip ion a e o mucins and in lamma o y cy okines was e y high in ERE abbi s, excep IL-2, and i s analysis e ealed he exis ence o wo clea ly di e en gene exp ession pa e ns co esponding o In lamma o y and (mucin) Sec e o y P o iles. Fu he mo e, hese p o iles we e associa ed o di e en bac e ial species, sugges ing ha hey may co espond o di e en s ages o he disease. O he da a ob ained in his wo k ein o ced he no ion ha ERE mo bidi y and mo ali y is possibly caused by an o e g ow h o di e en pa hogens in he gu o animals whose immune de ence mechanisms seem no o be adequa ely esponding. Ci a ion: Ba ¨ue l C, Collado MC, Zu ´n ˜iga M, Blas E, Pe ´ ez Ma ı ´nez G (2014) Changes in Cecal Mic obio a and Mucosal Gene Exp ession Re ealed New Aspec s o Epizoo ic Rabbi En e opa hy. PLoS ONE 9(8): e105707. doi:10.1371/jou nal.pone.0105707 Edi o : Mo ´nica V. Cunha, INIAV, I.P.- Na ional Ins i u e o Ag icul u e and Ve e ina y Resea ch, Po ugal Recei ed Ap il 3, 2014; Accep ed July 22, 2014; Published Augus 22, 2014 Copy igh : ß2014 Ba ¨ue l e al. This is an open-access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal au ho and sou ce a e c edi ed. Da a A ailabili y: The au ho s con i m ha all da a unde lying he indings a e ully a ailable wi hou es ic ion. Da a a e a ailable om MG-RAST as he Rabbi -ERE-p ojec using he accession numbe 4543253.3. Funding: This wo k was suppo ed by he g an s AGL2010-22211, AGL2006-07596 and by he Fun-C-Food CSD2007-00063 p ojec om he Consolide -Ingenio p og amme, bo h om he Spanish Minis y o Science and Inno a ion. The unde s had no ole in s udy design, da a collec ion and analysis, decision o publish, o p epa a ion o he manusc ip . Compe ing In e es s: The au ho s ha e decla ed ha no compe ing in e es s exis . * Email: [email p o ec ed]ic.es .These au ho s con ibu ed equally o his wo k. In oduc ion The 19 h cen u y wi nessed he eme gence o animal ancy and abbi s (O yc olagus cuniculus) began o be aised also as pe s, and a p esen days hey a e among he mos impo an labo a o y models. Rabbi s a e hindgu e men e s, hei cecum has a la ge size and, wi h he s omach, i has a e y ele an ole in he diges i e unc ion. Cop ophagy (caeco ophy) is a e y cha ac e - is ic habi o his species ha aids o comple e he diges ion o ege able componen s, acili a es he assimila ion o p o eins and o he nu ien s syn hesized by cecal bac e ia and main ains gu bac e ial popula ions. In es inal heal h o domes ic abbi s is qui e delica e and any dis up ion o he diges i e p ocess esul s in gas oin es inal diseases, mos equen ly ela ed o die o s ess [1]. A high incidence o diges i e diseases o unknown ae iology in young mammals is o en ela ed o dis o ions in mic obio a composi ion [2]. The con ibu ion o he gas oin es inal ac mic obio a owa ds mammalian hos heal h and pe o mance is now widely accep ed. P e ious s udies on abbi cecal mic obio a pe o med using classical cul u e-based echniques [3] and ecen ly, by molecula echniques [4], showed ha bac e ial species a e mainly s ic anae obes wi h p edominance o he phylum Fi micu es o e Bac e oides-P e o ella. Epizoo ic abbi en e opa hy (ERE) is a se ious a m abbi disease ha eaches 30–95% mo ali y, independen o he b eed, wi h a pa icula incidence in he pos -weaning pe iod [5]. The main symp oms a e cecal impac ion, abdominal dis ension wi h gas and luid accumula ion in s omach and in es ines, le ha gy, hunched pos u e, subno mal empe a u e and copious clea mucus dia hoea some imes p eceded by cons ipa ion, o which i is also called Mucoid En e opa hy. In acu e o m, clinical signs p ecede dea h by 1–3 days, o 7–9 days in a longe e m condi ion. The mos ele an his ological ea u e is a ema kable goble cells hype plasia o ileum, cecum, e mix, sacula ed colon and colon [6]. In p ac ice, exhaus i e hygienic measu es a e equi ed o cons ain his epizoosis and, in some cases, only sus ained an ibio ic medica ion p e en s ERE mo ali y. A g ea e o has been de o ed o s udy he ae iology o ERE and i is now accep ed PLOS ONE | www.plosone.o g 1 Augus 2014 | Volume 9 | Issue 8 | e105707 ha bac e ial elemen s a e in ol ed in he pa hogenesis, since an ibio ic ea men s a e usually e ec i e in he p e en ion o ERE, and also because he cecal con en il a e o ERE abbi s inocula ed o heal hy animals can ep oduce he disease [7]. Howe e , o he iden i ied and uniden i ied ac o s ela ed o he en i onmen , husband y condi ions and du a ion o he weaning pe iod ce ainly pa icipa e in he de elopmen o ERE [8]. Die is impo an , as he disease has been ep oduced p o iding eed wi h a educed ib e con en and liga ion o he cecum has been shown o be an in asi e, bu e ec i e o m o ep oduce he disease [6,8]. Exhaus i e bac e iological analyses ca ied ou in he pas could no esol e he mic oo ganisms in ol ed, al hough clos idial species and coli o ms ha e been mos equen ly ound in diseased abbi s [6,8–10]. A e y ele an ea u e o ERE is he abundan p oduc ion o mucus h oughou he small, bu ema kably, in he la ge in es ine and cecum [8]. Gas oin es inal mucins p oduced by goble cells a e he main s uc u al componen s o he mucus laye ha has a c ucial lub ica ing and p o ec i e ole. Mucins a e syn hesized as a i s line o de ence and accumula e di e en ial e ec o and egula o y esponses agains a ple ho a o commensal and pa hogen mic oo ganisms. Thei syn hesis is induced by p o- in lamma o y s imuli s emming om pa hogen in asion o chemical agg ession o he epi helia, as pa o he inna e de ence sys em, bu also mic oo ganisms and i uses e icien ly ecognize mucins as binding a ge s [11]. ERE his opa hological signs ha e also been desc ibed [7,8], bu u he biochemical, immunological o molecula echniques ha e no been implemen ed, and he e o e, a deepe pe spec i e o he disease is missing. In his s udy, qPCR and py osequencing o 16S DNA amplicons we e used o desc ibe and compa e he cecal mic obio a in heal hy abbi s and o analyse he impac o an ibio ic ea men and ERE on he mic obio a composi ion. In addi ion, gene exp ession o mucins and p o-in lamma o y cy okines was also de e mined and co ela ions wi h a ia ions in cecal mic obio a we e es ima ed. This s udy also e ealed new bac e ial g oups signi ican ly ela ed o ERE and new insigh s on he pa hology o his disease. Ma e ials and Me hods Animals, eeding and housing condi ions Animals used in he expe imen we e h ee-way c ossb eed abbi s om ma e nal lines A and V and pa e nal line R, de eloped by he Poly echnic Uni e si y o Valencia (Animal B eeding Uni ), Spain. They we e g own in an expe imen al a m ha kep s anda d comme cial handling condi ions a he Poly echnic Uni e si y o Valencia. Lac a ion las ed 30 days and animals we e sepa a ed in g oups o en pe cage in he a ening a ea. A e weaning a g oup o abbi s (an ibio ic ea men ) was ed ad libi um wi h comme cial medica ed pelle ed abbi eed (Nan a, Nu eco, Spain) con aining lyncomycin (29 ppm), spec i- nomycin (29 ppm), iamu in (40 ppm) and neomycin (121 ppm). A second g oup o abbi s was ed wi h a ba ch o he same eed wi hou an ibio ics. In o de o inc ease he incidence o ERE hese abbi s we e housed in a a ening sec ion o he a m ed wi hou an ibio ics, ba ch a e ba ch, du ing 4 mon hs p e ious o he expe imen . A he age o 40 days, en heal hy animals and en animals su e ing ypical ERE symp oms, like apa hy, yellowish pe ianal ecal s ain (mucus), cecal impac ion and wa e y sound in he gu we e sampled a andom and om di e en cages o he sec ion wi hou an ibio ics. Also en abbi s wi h an ibio ic ea men we e selec ed andomly in he a m. All abbi s we e collec ed on he same day. Heal hy o ERE animals (gas in he s omach and impac ion o he cecum) ha upon nec opsy could no be unequi ocally diagnosed we e elimina ed om he s udy ( o al o 2). The comple e clean cecum was ex ac ed a e liga ion a ileo-cecal al e and he comple e o gan, including he e mix, was immedia ely ozen a 280uC un il use. This s udy was ca ied ou in s ic acco dance wi h he Spanish na ional ules (RD223/1988 and RD1201/2005) ha p o ec animals used in expe imen a ion and o he scien i ic pu poses. The p o ocol was app o ed by he Commi ee on he E hics o he Poly echnic Uni e si y o Valencia (UPV). Due o hei small size, animals we e eu hanized by ce ical disloca ion and hen dissec ed. Nucleic acids ex ac ion om cecal con en , cecal mucosa and e mix The ozen cecum con en was agmen ed in pieces o app oxima ely 5 g and quickly de os ed. Then, o al DNA was ex ac ed om 200–300 mg samples o cecal con en using a Qiagen s ool DNA ex ac ion ki (QIAgen, Hilden, Ge many) acco ding o manu ac u e ’s ins uc ions wi h a p e ious dis up- ion wi h a bead bea ing s ep. RNA samples o q-RT-PCR we e ob ained as ollows. Cecal mucosa ci cula agmen s o 10–15 mm diame e and ans e sal e mix sec ions o 5 mm we e ca e ully excised and immedia ely subme ged in TRIzol Reagen (In i ogen). The issue was homogenized wi h a Poly on de ice and o al RNA was isola ed using he RNeasy Mini Ki (Qiagen) acco ding o he manu ac- u e s’ ins uc ions. Con amina ing genomic DNA was diges ed using Deoxy ibonuclease I (Sigma) and hen, RNA quan i y and quali y was e alua ed using he Agilen 2100 Bioanalyse (Agilen ). Re e se ansc ip ion eac ions we e pe o med using he T an- sc ip o Fi s S and cDNA Syn hesis Ki (Roche) acco ding o he ins uc ions o he manu ac u e . Quan i a i e eal- ime PCR (qPCR) analysis The qPCRs o speci ic bac e ial g oups we e conduc ed as p e iously desc ibed [12]. qPCR ampli ica ion and de ec ion was pe o med in a Ligh Cycle 480 Real-Time PCR Sys em (Roche). Each eac ion mix u e o 10 ml consis ed o SYBR G een PCR Mas e Mix (Roche), 0.5 ml o each o he speci ic p ime s [12] a a concen a ion o 0.25 mM, and 1 ml o empla e DNA. We also analysed me hanobac e ia using Me hanob e ibac e genus-speci - ic p ime s (MET-105 and MET-386 ) and M. smi hii ni H gene speci ic p ime s (Mni -342 and Mni -363 ) as desc ibed elsewhe e [13]. A mel ing cu e analysis was made a e ampli ica ion o assess he speci ici y o he ampli ica ion eac ion. The bac e ial concen a ion in each sample was calcula ed by in e pola ion o he ob ained C alues o s anda d cu es. These we e c ea ed using se ial 10- old dilu ion o pu e cul u e-speci ic DNA agmen s o known size co esponding o 10 o 10 9 numbe o agmen gene copies/ml. Tagged PCR ampli ica ion o bac e ial 16S ibosomal genes o py osequencing A ba coded p ime se based on uni e sal p ime s 27F (59- GAGTTTGATCMTGGCTCAG-39), and 518R (59- WTTACCGCGGCTGCTGG-39), we e used o ampli y a 500 bps o he 16S RNA genes encompassing he V3 egion. The PCR was ca ied ou using a high- ideli y KAPA-HiFi polyme ase (KappaBiosys ems, US) wi h an annealing empe a- u e o 52uC and 20 cycles o minimize PCR biases. Pu i ied PCR p oduc s we e pooled in equimola amoun s, as 454 Roche p o ocols desc ibe, and submi ed o py osequencing using he Rabbi Cecal Mic obio a and Gene Exp ession in ERE PLOS ONE | www.plosone.o g 2 Augus 2014 | Volume 9 | Issue 8 | e105707 Genome Sequence FLX Ti anium Se ies (454 Li e Science, B an o d, USA). All o he p ocedu es ollowed he manu ac u e ’s di ec ions (454 Li e Science) and we e conduc ed a Mac ogen (Seoul, Sou h Ko ea). Bo h chains o all amplicons we e sequenced o assu e high quali y da a. Sequence da a ob ained in his wo k a e a ailable in MG-RAST (h p://me agenomics.anl.go /) as Rabbi -ERE p ojec wi h accession Numbe 4543253.3. Gene exp ession analysis o abbi genes h ough eal ime qPCR Fo e e se ansc ip ion 5 mg o o al RNA we e used. The eac ions we e pe o med using he T ansc ip o Fi s S and cDNA Syn hesis Ki (Roche) acco ding o he ins uc ions o he manu ac u e . Quan i a i e eal- ime PCR was pe o med on a Ligh Cycle 480 Real Time PCR Sys em using SYBR G een I Mas e Mix (Roche). PCR cycling condi ions comp ised an ini ial polyme ase ac i a ion s ep a 95uC o 10 min ollowed by 40 cycles o 10 s a 95uC, 10 s a 60uC and 12 s a 72uC. qPCR p ime s we e designed using he P ime -Blas ool (h p://www. ncbi.nlm.nih.go /) and a e lis ed in Table S14. qPCR p ime s we e alida ed o con i m e iciency h ough se ial dilu ions. PCR p oduc s we e con i med by aga ose gel elec opho esis o yield a unique band and, addi ionally, a e each qPCR un a dissocia ion cu e was pe o med. S a is ical analysis Fo he analysis o qPCR da a, IBM-SPSS 19.0 so wa e (SPSS Inc., Chicago, IL, USA) was used. Due o non-no mal dis ibu- ion, mic obial da a a e exp essed as medians wi h in e qua ile anges (IQR). Compa isons among da a o mo e han wo g oups o abbi s we e done by applying he Wilcoxon/K uskal-Wallis es , and compa isons be ween da a o wo g oups we e done by applying he Mann-Whi ney U es . The Bon e oni adjus men es was also applied o co ec he signi icance o mul iple es compa isons among h ee g oups. The x-squa e es was used o es ablish di e ences in he bac e ial p e alence be ween he s udied g oups. A P,0.050 was conside ed s a is ically signi ican . The possible co ela ion be ween a iables was s udied by applying Pea son’s co ela ion coe icien , and signi icance was es ablished a 0.05%.Rega ding py osequencing, low quali y sequences we e il e ed ou o emo e sequences ha ing a leng h sho e han 100 nucleo ides om aw da a se s. Sequences we e aligned and classi ied agains he SILVA comp ehensi e RNA da abase (h p://www.a b-sil a.de/). A de eplica e eques on he pipeline was used o iden i ying he ep esen a i e sequences o each ope a ional axonomic uni (OTU) gene a ed om he comple e linkage clus e ing wi h a 97% simila i y. A e axono- mical assignmen o py osequencing da a, ela i e equencies o di e en axonomic ca ego ies ob ained we e calcula ed using he S a is ical Analysis o Me agenomic P o iles p og am (STAMP .2.0.0) [14]. S a is ical di e ences be ween expe imen al abbi g oups we e es ima ed by ANOVA analysis wi h he Games- Howell pos -hoc es and he mul iple es co ec ion o Benjamini- Hochbe g as implemen ed in STAMP. Ra e ac ion cu es we e calcula ed using he Ra e ac Win p og am (h p://www.uga.edu/ ,s a a/so wa e/So wa e.h ml) and also, alpha di e si y indexes we e de e mined wi h he QIIME pipeline om a e ied ables using he Shannon-Wiene index o di e si y, he Chao1 index o ichness and also, Obse ed Species (numbe o unique OTUs) and Phylogene ic Dis ance (PD_whole). A be a di e si y dis ance ma ix was compu ed om he p e iously cons uc ed OTU able using UniF ac analysis. Unweigh ed (p esence/absence me ix) and weigh ed (p esence/absence/abundance me ix) UniF ac dis ances we e used o cons uc wo- and h ee-dimensional P incipal Coo dina es Analyses (PCoA) plo s. Biplo s we e gene a ed as pa o he be a di e si y analysis in Qiime using genus le el OTU ables showing p inciple coo dina e clus e ing o samples alongside weigh ed axonomic g oup da a. Da a on assigned sequences a he genus le el sha ed be ween samples we e used o gene a e a Venn diag am. Rela i e gene-exp ession was quan i ied acco ding o he e iciency-co ec ed me hod [15] using he REST 2009 so wa e ool. Di e ences in inpu cDNA we e no malized o glyce alde- hyde-3-phospha e dehyd ogenase (GAPDH) and b-ac in (ACTB) exp ession. Finally, co ela ion analysis was pe o med wi h he S a - g aphics p og amme, as well as he P incipal Componen Analysis (PCA) o he sequences and exp ession da a, wi h ecip ocal p ojec ions, which we e hen d awn wi h SigmaPlo 10.0. Resul s Bac e ial popula ions in he cecum o heal hy abbi s A o al o 89,091 16S RNA double s anded sequence eads we e ob ained om 10 cecal samples o heal hy indi iduals ed wi h non-medica ed eed. The a e age ead leng h was 492.5 bp. Using he SILVA da abase o axonomic assignmen , he mos common phyla we e Fi micu es we e Fi micu es (78.25% o o al eads), Bac e oide es (15.75%), Ve ucomic obia (2.40%), and Tene icu es (2.39%). Wi hin Fi micu es, in he o de Clos idiales, he mos abundan amilies we e Ruminococcaceae (42.48% o o al eads) and Lachnospi aceae (34.85%) while in he o de Bac e oidales, Rikenellaceae (6.38%) was he mos abundan amily. We ound 90 di e en gene a in cecal samples o heal hy abbi s om which he p edominan ones we e Alis ipes (5.63% o o al eads), Ruminococcus (4.02%), Akke mansia (2.40%) and Subdolig anulum (2.28%). Howe e , he 20 mos abundan sequences (37.76% o o al eads) could no be classi ied a he species le el: 15 o hem belonged o he class Clos idia ( amilies Lachnospi aceae and Ruminococcaceae), including he mos abundan sequence in heal hy abbi s EF445173 al eady epo ed [3] (8.34% o o al eads) ( amily Lachnospi aceae); 4 o he class Bac e oidia ( amily Rikenellaceae); and one belonged o he class Ve ucomic obia (genus Akke mansia; Tables S1, S2, S3, S4). Impac o an ibio ic ea men and ERE on he cecal mic obio a Fu he 20 cecal samples o abbi s belonging o wo addi ional g oups, heal hy ea ed wi h an ibio ics (A) and ERE abbi s (E), we e collec ed in o de o be compa ed o he p e ious heal hy abbi s wi hou an ibio ic ea men , used as con ol g oup (C). One sample ailed in he an ibio ic g oup (A) so ha a o al o 19 samples we e analysed. A e quali y il e ing and leng h imming 199,217 16S DNA eads we e analysed, wi h an a e age numbe o axonomically assigned high-quali y double s anded eads o 10,485 pe abbi wi h an a e age size o 492 bp (Tables S1, S2, S3, S4). The domina ing Phyla in all abbi s we e Fi micu es ollowed by Bac e oide es. Class Clos idia (Fi micu es) was mo e abundan in heal hy g oups (A = 85.49%, C = 78.06%, E = 54.07%; co ec ed p,0.001) and class Bac e oidia in he ERE g oup (A = 8.83%, C = 15.73%, E = 22.18%; co ec ed p = 0.013). ERE abbi s showed a ema kable inc ease in P o eobac e ia, pa icula ly class c-P o eobac e ia (A = 0.016%, C = 0.023%, E = 10.03%; co ec - ed p = 0.003). The numbe o Ve ucomic obiae was also highe in ERE (A = 0.84%, C = 2.40% E = 8.40%, co ec ed p = 0.009) mos ly due o a g ea inc ease o eads o he genus Akke mansia (see below). O he axons wi h lowe o al numbe s showed Rabbi Cecal Mic obio a and Gene Exp ession in ERE PLOS ONE | www.plosone.o g 3 Augus 2014 | Volume 9 | Issue 8 | e105707 ema kably di e en p opo ions in he ERE g oup: class Bacilli was mo e abundan in he ERE g oup (A = 0.07%, C = 0.02%, E = 1.1%), and Tene icu es (class Mollicu es; A = 3.68%, C = 2.39%, E = 0.44%) less equen in diseased animals. A e he sc u iny o o al coun s o bac e ial amilies in he h ee g oups impo an di e ences we e de ec ed (Figu e 1). An ibio ic ea men (A) educed he numbe o eads in he amilies Clos idiaceae and Ruminococcaceae. A numbe o bac e ial amilies we e in high numbe s o speci ically p esen only in he ERE g oup, such as Ve ucomic obiaceae, En e obac e iaceae and Bac e oidaceae (Figu e 1). A a lowe axonomical le el, he mos equen gene a in heal hy g oups (A,C) we e Ruminococcus and Alis ipes, and in ERE abbi s Bac e oides (12,45%), Akke - mansia (8,40%), Esche ichia (8,25%), Rikenella (3.40%) as well as Clos idium (1.24%) showed a majo p esence (Figu e 2, Table S4). The compa ison o unclassi ied sequences showed ha wo Lachnospi aceae eads o unknown genus al eady epo ed obese humans DQ799912.1.1389 [16] and anae obic sludge CU919535.1.1344 [17] accoun ed o 8.96% and 4.87% espec- i ely o o al iden i ied sequences in ERE abbi s. In addi ion, po en ial oxin p oducing Fi micu es ha e also been ound, such as species o he genus Lysinibacillus, wi h 0.99% o o al eads in ERE cecal samples. The h ee g oups showed a ema kable di e si y, wi h a e ac ion cu es indica ing a numbe o Ope a ional Taxonomic Uni s (OTUs) o e 10,000 when eads we e clus e ed a 97% sequence iden i y ( he consensus alue o de e mining species bounda ies). Ra e ac ion cu es, a OTU le el, show ha he ERE g oup had lowe bac e ial di e si y han he con ol and an ibio ic g oups (Figu e 3). The slope a he end o he cu es o An ibio ic and Con ol g oups sugges s ha mic obio a is e y di e se and a new e o s would be needed o ep esen he comple e abbi cecal mic obiome. The 16S sequences o all indi iduals in each g oup we e pooled, wi h an ope a ional axonomic-uni de ini ion se a 97% sequence iden i y. An OTU was a clus e o 16S RNA sequences which we e o e 95% iden ical, a conse a i e es ima e o he bounda y be ween species, es ablished a 97% o ull- leng h 16S gene sequences (Each a e ac ion cu e is plo ed, along wi h i s 95% con idence in e al). As was seen in he p ima y analysis, he o e all di e si y in ERE g oup was signi ican ly lowe han ha in con ol subjec s and an ibio ic ea men . O e all di e si y o axa wi hin he samples was de e mined by alpha di e si y me ics (Chao1 me ic es ima es he species ichness; Shannon es ima es he di e si y; he Obse ed Species; Phyloge- ne ic Dis ance (PD_whole_ ee)) showed lowe di e si y, phyloge- ne ic ichness and e enness in ERE g oup compa ed o hose obse ed in he o he wo g oups: con ol and an ibio ic g oup (Figu e 3). Venn diag am shows a b oade pe spec i e, whe e Con ol and An ibio ic g oups sha e mos o he bac e ial amilies and gene a. A co e o 13 amilies and 24 gene a we e p esen in ERE, an ibio ic and con ol g oup (Figu e 4). P incipal Coodina es Analysis (PCoA) o he samples clea ly sepa a ed he ERE abbi s om he es o g oup samples in bo h, weigh ed (Figu e 5, panel A) and unweighed (Figu e 5, panel B) plo s (PC1 accoun ing o 14% o he a iance in unweigh and 50% in weigh Uni ac analysis), indica ing ha he ERE mic obio a is composi ionally dis inc . Heal hy indi iduals wi h o wi hou an ibio ic ea men (An ibio ic,Con ol) clus e ed oge he and sepa a ely om he ERE g oup o animals. This indica ed ha mic obio a om g oups An ibio ic and Con ol a e no signi ican ly di e en , bu a ema kable dysbiosis is a ec ing ERE abbi s. In o de o explo e axonomic d i ing ac o s o pa e ns in a PCoA bac e ial gene a we e plo ed in he same diag am o he unweigh ed PCoA o all samples (Biplo ), whe e ela i e abundance o hese axons we e also ep esen ed by he sphe e sizes. The genus Bac e oides clea ly clus e ed wi h ERE samples, clus e ing o heal hy samples was d i en by wo uncha ac e ized gene a o he Ruminococcaceae amily, one Figu e 1. To al eads o bac e ial amilies iden i ied by SILVA a e py osequencing o he 16S RNA o he cecal con en o abbi s. The g aph ep esen s he a e age o each o he h ee g oups: con ol (whi e), an ibio ic ea men (g ey) and ERE (black). doi:10.1371/jou nal.pone.0105707.g001 Figu e 2. P opo ional piled up column g aph ep esen ing he a e age abundance o he di e en bac e ial gene a o he con ol g oup (heal hy, no an ibo ics) (C), abbi s ea ed wi h an ibio ics (A) and ERE animals (E). Mos abundan gene a a e ma ked on he ba . doi:10.1371/jou nal.pone.0105707.g002 Rabbi Cecal Mic obio a and Gene Exp ession in ERE PLOS ONE | www.plosone.o g 4 Augus 2014 | Volume 9 | Issue 8 | e105707 genus o Lachnospi aceae and a less abundan genus o he Clos idia class (Figu e 5, panel C). Finally, qPCR was pe o med o analyse he p esence o ce ain species and o con i m signi ican di e ences in he p opo ion o speci ic axonomic g oups be ween ERE, an ibio ic and heal hy con ol g oup samples (Table S5). Signi ican ly highe le els o Akke mansia muciniphila,Bac e oides-P e o ella and Clos idium coccoides g oup we e obse ed in ERE g oups compa ed o hose obse ed in con ol (p = 0.001, p = 0.042 and p = 0.050, espec- i ely) and an ibio ic (p = 0.002, p = 0.002 and p = 0.049, espec- i ely). Fu he mo e, we de e mined he p esence o bac e ia om he genus Me hanob e ibac e as example o A chaea in abbi cecal samples. A signi ican ly highe le el o he bac e ia o he genus Me hanob e ibac e was obse ed in he ERE g oup compa ed o he heal hy con ol and an ibio ic g oups (p, 0.0001), al hough he species M. smi hii was de ec ed mo e o en in heal hy con ols (4/10, 40%) and an ibio ic g oup (3/10, 30%) han in ERE g oup (1/10, 10%). Howe e , o e all p opo ions o me anobac e ia (Me hanob e ibac e ) a e e y small, as ampli ied molecules o 16S DNA we e wo (C g oup) o i e (E g oup) o de s o magni ude below o al bac e ia. Mucin and cy okine exp ession in he cecum One o Epizoo ic Rabbi En e opa hy main physiological symp oms is he abundan sec e ion o mucins a he cecum and colon [18], o which moni o ing hei syn hesis du ing he cou se o he disease, compa ed o ha in heal hy animals, was conside ed a p io i y issue. Mucin sec e ion by he goble cells is known o be induced by in lamma o y signals, equen ly a e pa hogen challeng- es [11], hence we sea ched in he abbi genome p ojec (h p:// www.ncbi.nlm.nih.go /genome? e m = o yc olagus%20cuniculus) in o de o design sui able p ime s o q-RT-PCR o mucin and cy okine encoding genes. MUC1, MUC4 and MUC13 we e success ully es ed and also he SAM-poin ed domain-con aining E s-like ac o gene (SPDEF), an ac i a o o goble cells which is induced by p o-in lamma o y signals and by pa hogen s imuli, ha also egula es mucine exp ession [19]. Homologous gene o MUC2, he dominan in es inal mucin in humans, was no a ailable a he onse o his s udy. Also p ime s o abbi p o-in lamma o y cy okines and lymphocy e di e en ia ion p ecu so s we e designed (TNF-a,IL-4,IL-6,IL-8,IFN-cand IL-2). Gene exp ession o all he genes was de e mined in cecal mucosa samples o all 30 abbi s (Table S7), and exp ession da a could be success ully calcula ed wi h he excep ion o IL-4 whose C was oo high o allow eliable calcula ions in cecal mucosa. T ansc ip ion da a we e calcula ed as ela i e exp ession o ERE (E) and an ibio ic g oups (A) agains he mean alue o he heal hy con ol (C). Rabbi s su e ing ERE displayed an inc eased ansc ip ion le el o all mucins, especially MUC1 (Figu e 6, panel A). The a e age exp ession alues o SPDEF had no signi ican inc ease in any o he h ee g oups, howe e , pa icula di e ences in he ERE g oup will be dissec ed below. Exp ession o cy okines IL-8, TNF-aand IL-6 we e also signi ican ly up egula ed (Figu e 6, panel A). This esul showed a di ec linkage be ween he disease s a us and mucin syn hesis in he cecal epi helium o ERE abbi s, indica ing he p esence o a s ong p o-in lamma o y s imulus in he cecal con en s. IFN-cshowed a small induc ion, whe eas IL-2 had a mode a e dec ease in ansc ip ion. Howe e , a g ea a iabili y was no iced in he ansc ip ion a es o he cy okines wi hin he ERE g oup, o his eason we decided o u he analyse his g oup (see below). Li le di e ences we e ound be ween he ansc ip ion a es o abbi s ea ed wi h an ibio ics (A) and he con ol g oup (C), besides a sligh endency o o e p oduce mucins in he A g oup (Figu e 6, panel A). Figu e 3. Ra e ac ion cu es ep esen an app oxima ion o he numbe o species iden i ied o he o al numbe o sequences ob ained (A). The numbe o species was ep esen ed by he numbe o ope a ional axonomic uni s (OTUs). Cu es we e calcula ed wi h Ra e ac Win so wa e a 97% simila i y le el co esponding o species- le el phylo ypes. Ra e ac ion analysis also included he es ima ion o he di e si y o bac e ial axa in each sample wi h di e en me ics wi hin he QIIME pipeline, such as he alpha di e si y indices: Chao1 (B), Shannon (D), Phylogene ic dis ances (C) and Obse ed Species (E). doi:10.1371/jou nal.pone.0105707.g003 Figu e 4. Venn diag am showing he ela i e abundance a he genus le el based on he classi ica ion o he pa ial 16S DNA bac e ial sequences om cecal samples o an ibio ic, heal hy con ol and ERE g oups using he RDP classi ie so wa e o axonomy assignmen : con ol g oup (heal hy, no an ibio ics) (Con ol), abbi s ea ed wi h an ibio ics (An ibio ic) and ERE animals (ERE). doi:10.1371/jou nal.pone.0105707.g004 Rabbi Cecal Mic obio a and Gene Exp ession in ERE PLOS ONE | www.plosone.o g 5 Augus 2014 | Volume 9 | Issue 8 | e105707 Fu he mo e, gene exp ession da a in cecal mucosae o ERE was analysed in de ail. A ema kable and highly signi ican co ela ion was de ec ed be ween TNF-aand IL-6 and IL-8, as expec ed om hei di ec connec ion wi h canonical p o-in lamma o y signaling pa hways (i.e. NF-kB), and also connec ion o IFN-cwi h TNF-a and IL-6 and IL-8 could be con i med. Rega ding mucin exp ession, we ound a posi i e co ela ion (low signi icance, p# 0.1) o SPDEF exp ession wi h MUC13 sugges ing egula o y connec ions, and a nega i e co ela ion wi h IL-2 and INFc, sugges ing a e e se ( ep essing) egula o y connec ion, bu no ela ionship o SPDEF wi h he wo o he mucins MUC1 and MUC4. Howe e , MUC13 nega i e co ela ion o INFcand TNF- awas g ea e and mo e signi ican , and also ha o MUC1 wi h IL- 2 (Table S6). Then, PCA o gene exp ession da a in he cecal mucosa was pe o med and d awn (Figu e 7). Va iables clea ly g ouped in wo clus e s, one including cy okines (IFN-c, TNF-a, IL-6 and IL-8) and he second, mucins (MUC13, MUC1, MUC4 and he egula o SPDEF). They we e sepa a ed along he o dina e axis (second componen ) and IL-2 emained dis an om ei he g oup, unde lining an independen exp ession pa e n. In ac , exp ession o his cy okine showed e y small a ia ion be ween heal hy and ERE abbi s. When da a om all abbi s we e p ojec ed on he gene exp ession PCA, heal hy indi iduals (Con ol and An ibio ic g oups) clus e ed igh ly a he nega i e (le ) side o he abcise axis (PC1), while ERE animals we e all p ojec ed in he posi i es alues ( igh ). In e es ingly, abbi s su e ing he disease we e sepa a ed in wo subse s along he o dina e axis (PC2) wi h abbi s E5, E6, E8 (n = 3) and E1, E2, E3, E4, E7, E9, E10 (n = 7) close o he clus e Figu e 5. P incipal Coo dina es Analysis (PCoA) 3D plo s gene a ed wi h UniF ac showing clus e ing o OTUs bac e ial g oups o cecal samples om ERE, and Heal hy (Con ol and An ibio ic) g oups. PCoA is a Mul idimensional Scaling Me hod ha allows o explo e and o isualize dissimila i ies o phylogene ic da a om a dis ance ma ix impo ed om RDP, assigning a loca ion o each sample in a 3D g aphics. Panel A, unweigh ed plo ; panel B, weigh ed plo ( aking in o accoun he ela i e abundance o each axon in he samples); panel C, Biplo ep esen a ion o PCoA o unweigh ed, pai wise Uni ac dis ances showing clus e ing o bac e ial g oups cecal samples om ERE, Con ol and An ibio ic g oups. Blue sphe es a e Con ol abbi s, ed sphe es a e he An ibio ic g oup and ligh b own sphe es a e he ERE abbi s. In panel C, he Biplo shows he p ojec ion o axa posi ions (g ey sphe es) ha ep esen weigh ed a e ages o he coo dina es o all samples. The size o he sphe es is p opo ional o he ela i e abundance o he axon. doi:10.1371/jou nal.pone.0105707.g005 Rabbi Cecal Mic obio a and Gene Exp ession in ERE PLOS ONE | www.plosone.o g 6 Augus 2014 | Volume 9 | Issue 8 | e105707 o in lamma o y cy okines and mucins, espec i ely. In ac he exp ession p o iles o cy okines and mucins o hese wo subg oups we e e y di e en . Samples o he i s g oup (E5, E6, E8) had e y high exp ession o p o-in lamma o y cy okines, IL-6x16, IL- 8x8, TNF-ax3 and IFN-cx11 old g ea e han he o he g oup, while samples in he second g oup (E1, E2, E3, E4, E7, E9, E10) had highe exp ession o mucins (Table 1, Table S8), o which hey we e assigned o In lamma o y P o ile and Sec e o y P o ile, espec i ely. Un o una ely he no he In lamma o y P o ile was low (n = 3) o u he compa isons. Mucin and cy okine exp ession in he e mix T ansc ip ion was also assayed wi h e mix samples, as hese samples could be en iched in lymphocy es (Figu e 6, panel B). In e es ingly, he e we e signi ican di e ences be ween ansc ip- ion le els o Con ol samples and ERE in all genes s udied, excep IFN-cand TNF-a, and mucin genes (MUC1, MUC13, MUC4) and cy okine genes IL-4, IL-6, IL-8 ollowed a simila endency o ha in he cecal mucosa, al hough IL-4 could no be quan i ied in he cecal mucosa. Induc ion o mucins mRNA was much lowe han in he cecum (Figu e 6, panel A). No signi ican di e ences we e ound be ween an ibio ic and he con ol g oups (Figu e 6, panel B). When compa ing only gene exp ession o e mix and cecal samples o ERE animals, we de ec ed no up egula ion o IFN-c and TNF-ain e mix, bu a mode a e o e exp ession o IL-2 ( wo Figu e 6. G aphs ep esen ing ela i e gene exp ession da a de e mined by q-RT-PCR. Values a e calcula ed ela i e o he Con ol g oup and no malized o endogenous ACTB and GAPDH exp ession. Box plo s show he median alue (solid line inside boxes), he limi s o box ep esen he 25 h and 75 h pe cen ile, and whiske s depic he 5 h and 95 h pe cen ile (*P,0.05 s. Con ol; **P,0.001 s. Con ol). doi:10.1371/jou nal.pone.0105707.g006 Figu e 7. PCA o cy okine gene exp ession, whe e axonomic da a om each abbi sample a e p ojec ed. Red ci cles indica e g oupings o gene exp ession da a. Blue ci cles indica e he posi ion o indi idual abbi s. doi:10.1371/jou nal.pone.0105707.g007 Rabbi Cecal Mic obio a and Gene Exp ession in ERE PLOS ONE | www.plosone.o g 7 Augus 2014 | Volume 9 | Issue 8 | e105707 Table 1. Mean alues and s anda d de ia ion o gene exp ession alues, ela i e o con ol abbi s, o he wo likely p o iles (In lamma o y and Sec e o y) ound in abbi s su e ing ERE. Cecum Ve mix In lamma o y p o ile (n = 3) Sec e o y p o ile (n = 7) In lamma o y p o ile (n = 3) Sec e o y p o ile (n = 7) Mean SDe Mean SDe P alue Mean SDe Mean SDe P alue MUC1 21.62 4.07 31.75** 11.66 0.039 6.84 1.95 8.80 4.46 0.180 MUC13 7.04 2.60 17.35** 5.96 0.003 7.90 2.55 8.47 3.39 0.390 IL2 1.07 0.48 0.62 0.38 0.120 3.88 2.56 1.81 1.10 0.146 IFN-c17.24** 4.02 1.54 1.14 0.009 1.84 1.65 1.11 0.66 0.263 MUC4 22.15 17.57 12.62 10.52 0.227 4.89 2.30 2.71 1.06 0.119 IL8 123.20 113.26 15.63 16.93 0.120 22.35 19.25 3.95 3.88 0.119 TNF-a20.81** 6.06 6.47 6.72 0.013 1.88* 0.87 0.94 0.32 0.097 IL6 341.28* 229.66 21.24 27.43 0.068 4.74 2.57 2.45 1.12 0.130 IL4 - - - - - 4.17** 0.88 2.74 1.08 0.040 SPDEF 0.74 0.32 1.71** 0.79 0.013 3.47 1.76 4.29 1.82 0.269 As e isks ep esen a e ages signi ican ly g ea e in one p o ile han he o he , wi h signi icance: (*)p#0.1; (**)p#0.05. doi:10.1371/jou nal.pone.0105707. 001 Rabbi Cecal Mic obio a and Gene Exp ession in ERE PLOS ONE | www.plosone.o g 8 Augus 2014 | Volume 9 | Issue 8 | e105707 old) and SPDEF ( ou old). Exp ession da a om e mix wi hin he ERE g oup we e u he analysed in o de o asce ain i he e could also be wo p o iles, as hose ound in he cecum samples. In e es ingly, IL-4 exp ession in e mix samples could be quan i ied and i was signi ican ly g ea e in he In lamma o y P o ile han in he Sec e o y P o ile. Cecal mic obio a and gene exp ession ela ionships Di e en s a is ical analysis we e un o de e mine likely co ela ions be ween indi idual gene exp ession da a and bac e ial equencies, wi h he aim o ela e possible o ganisms o bac e ial axons o he pa hogenesis o ERE. As expec ed, global co ela ion analysis o ela i e gene exp ession and mic obial g oups in all abbi s showed ha high exp ession o MUC1, MUC13, MUC4, IL-8 and IL-6 and TNF-awas in all cases signi ican ly bound o high eads o hose bac e ial axons mos abundan in ERE, such as o de s Ve ucomic obiales, En e obac e iales and Bac e oi- dales, and on he con a y, high exp ession o hose ma ke s we e co ela ed o low le els o Clos idiales eads belonging o he class Clos idia ( amilies Lachnospi aceae and Ruminococcaceae), desc ibed as associa ed o heal hy abbi samples (see abo e). Also a lowe axon le el (genus), wi h he excep ion o Alis ipes and Anae o uncus, all mos abundan gene a we e signi ican ly co ela ed o high o low mucins and p o-in lamma o y cy okines exp ession, depending whe he hey we e mos abundan in he ERE o Heal hy g oups, espec i ely (Tables S9, S10, S11, S12). Also we ied o ind, h ough co ela ion analysis, o ganisms signi ican ly bound o he gene exp ession P o iles iden i ied (In lamma o y and Sec e o y) wi hin he ERE g oup, bu signi ican in o ma ion could no be ob ained due o he low numbe o abbi s in he In lamma o y P o ile. Howe e , plain PCA a species (o ganisms) le el pe o med wi h mos abundan sequences in he ERE g oup, wi h p ojec ion o gene exp ession da a, indica ed ha uncul u ed and uncha ac e ized isola es o he genus Akke mansia,Alis ipes and Bac e iodes clus e ed wi h IL-6, IL-8, TNFaand IFNcand in he p oximi y o abbi samples 5, 6 and 8 (In lamma o y P o ile) (Figu e S1). Di e se sequences we e posi ioned nex o MUC1 and MUC13 in he PCA plo wi h ep esen a i es o Akke mansia,Esche ichia ( ou sequences), Bac e oides ( wo sequences), Clos idium and h ee uniden i ied species and genus o he amily Lachnospi aceae. Howe e , he ou sequences o Esche ichia and one Lachnospi aceae we e close o he co e o MUC1 and MUC13 p ojec ion. This s esses he di e ences be ween he In lamma o y and Sec e o y P o iles and sugges s hey may be bound o speci ic mic obial en i onmen / elemen s. Discussion S udies using molecula me hods o he analysis o he gu mic obio a in abbi s a e no abundan . Capilla y elec opho esis single-s anded con o ma ion polymo phism (CE-SSCP) and dena u ing g adien gel elec opho esis (DGGE) we e use ul o de e mine he simila i y be ween bac e ial popula ions o he cecal con en and so aeces [20] on die s wi h di e en ib e con en [21] and o d aw a gene al quan i a i e map o he e olu ion o he mic obio a along he abbi ’s li e [4]. Cloning o 16S RNA genes in Esche ichia coli o e ed a mo e p ecise iden i ica ion o he bac e ial inhabi an s o abbi ’s gu , howe e , he p ocedu e in oduced quali a i e and quan i a i e biases. In a i s in en o y, Abecia e al. (2005) [21] showed he ele ance in abbi ’s cecal con en o Clos idia (Lachnospi aceae, Ruminococcaceae) and Ve ucomic obia, and epo ed a numbe o new uniden i ied sequences. Au ho s desc ibed he absence o sequences o he P e o ella–Bac e oides g oup, howe e , nowadays iden i ica ion da abases ha e been en iched and we could classi y all hose sequences (G oup 5, in Abecia e al, 2005), in he Bac e oide es phylum (Table S13). In a la e wo k, a la ge numbe o sequences (228) we e analysed and dis ibu ed in 70 OTUs in nine clus e s, whe e he majo i y o sequences belonged o he phylum Fi micu es (94%), only h ee we e Bac e oide es (4%) and one Ve ucomic obia [3]. Py osequencing has been ecen ly used o in es iga e he mic obial composi ion o a a ie y o cecal il a es and ac ions ha ep oduced ERE in abbi s, bu no s a is ical di e ences we e ound be ween samples in ha wo k [22]. Ou wo k showed ha cecal mic obio a in heal hy abbi s is domina ed by Fi micu es (78.25% o o al OTUs), Bac e oide es (15.75%), Ve ucomi obia (2.40%) and Tene icu es (2.39%) and his quan i a i ely simila o ha ound in o he monogas ic he bi o es (hindgu e men e s) [23] and in mouse, as he closes model, in which also Fi micu es domina e o e Bac e oide es [24,25]. In his wo k, we ob ained 289,000 good quali y sequences (bo h s ands) om 29 samples o acili a e he compa a i e analysis o heal hy abbi s ea ed o no wi h an ibio ics and abbi s su e ing ERE. The i s su p ising esul was ha he e we e no signi ican di e ences due o he an ibio ic ea men , despi e ha i was epo ed ha an ibio ics educed he mic obial di e si y in abbi s [22,26], as hey did in mice [27] and humans [28]. The ac ha we ound no signi ican di e ences be ween con ol abbi s and he an ibio ic g oup migh be due o he p olonged p e en i e medica ion in he expe imen al a m, a si ua ion ha could gene a e a uni o m en i onmen al popula ion ins alled in he a m ha ul ima ely colonize abbi s. Howe e , his uni o m mic obi- ological en i onmen could acili a e he de ec ion o bac e ial pa hogens in ol ed in ERE, o a leas o es ablish he mic obiological scena io in he cecum o abbi s su e ing ERE. The mos ele an ea u e o ERE cecal mic obio a is a ema kable dysbiosis and educed axonomic di e si y. An ex eme en i onmen mus be gene a ed du ing ea ly pa hogen- esis, as occu s in a numbe o o he pa hologies wi h a ma ked in lamma o y p o ile, whe e pa icula ly c-P o eobac e ia p oli - e a e, such as human in lamma o y bowel disease (IBD), nec o izing en e ocoli is in human in an s [29] o pigle sho bowel synd ome and mu ine model o C ohn’s disease [30–33]. Quan i a i e and quali a i e di e ences be ween heal hy and ERE abbi s ha e been ound, such as he abundance o Ve ucomi- c obiae, pa icula ly he genus Akke mansia.A. muciniphila is a ypical in es inal anae obe equen ly associa ed o human heal hy mucosa [34]. I s high p e alence in ERE cecal samples could be ela ed o hei mucin sca enge cha ac e [35,36], howe e , A. muciniphila can agg a a e gu in lamma ion when he e is p e alen a Salmonella in ec ion [37]. Ruminococcus is he mos ele an genus o he Fi micu es phylum dominan in heal hy indi iduals, ha no ably dec eases in ERE, despi e he ac ha his is gene ally an e icien mucin deg ading genus [30,31]. Ano he componen o he heal hy mic obio a is he genus Alis ipes. Bo h play an impo an ole in he deg ada ion o ege able eed componen s and he p oduc ion o sho chain a y acids [38,39], hence con olling Bac e oide es numbe s [40], and he e o e, could cons i u e essen ial componen s o abbi cecal mic obio a and candida es as p obio ics. Me hanobac e ia (A - chaea) we e epo ed in low numbe s in he cecal con en o abbi s, only p esen a e day 7 o li e [4]. Ou da a ob ained by qPCR con i med i . P oduc ion o me hane has been shown o be in e sely p opo ional o he size o animals. Ca bon exc e ed as me hane is ep esen s a me abolic loss, hence in he case o small he bi o es like abbi s, his may ep esen a deg ee o e men a i e Rabbi Cecal Mic obio a and Gene Exp ession in ERE PLOS ONE | www.plosone.o g 9 Augus 2014 | Volume 9 | Issue 8 | e105707