Peripheral blood gene expression patterns discriminate among chronic inflammatory diseases and healthy controls and identify novel targets

Mesko e al. BMC Medical Genomics 2010, 3:15 h p://www.biomedcen al.com/1755-8794/3/15 Open Access RESEARCH ARTICLE BioMed Cen al © 2010 Mesko e al; licensee BioMed Cen al L d. 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 (h p://c ea i ecommons.o g/licenses/by/2.0), 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 wo k is p ope ly ci ed. Resea ch a icle Pe iphe al blood gene exp ession pa e ns disc imina e among ch onic in lamma o y diseases and heal hy con ols and iden i y no el a ge s Be alan Mesko †1 , Szila d Poliska1 †1,4 , And ea Szegedi 3 , Zol an Szekanecz 5 , Ka oly Pala ka 6 , Ma ia Papp 6 and Laszlo Nagy* 1,2,4 Abs ac Backg ound: Ch onic in lamma o y diseases including in lamma o y bowel disease (IBD; C ohn's disease and ulce a i e coli is), pso iasis and heuma oid a h i is (RA) a lic millions o people wo ldwide, bu hei pa hogenesis is s ill no well unde s ood. I is also no well known i dis inc changes in gene exp ession cha ac e ize hese diseases and i hese pa e ns can disc imina e be ween diseased and con ol pa ien s and/o s a i y he disease. The main ocus o ou wo k was he iden i ica ion o no el ma ke s ha o e lap among he 3 diseases o disc imina e hem om each o he . Me hods: Diseased (n = 13, n = 15 and n = 12 in IBD, pso iasis and RA espec i ely) and heal hy pa ien s (n = 18) we e ec ui ed based on s ic inclusion and exclusion c i e ia; pe iphe al blood samples we e collec ed by clinicians (30 ml) in Venous Blood Vacuum Collec ion Tubes con aining EDTA and pe iphe al blood mononuclea cells we e sepa a ed by Ficoll g adien cen i uga ion. RNA was ex ac ed using T izol eagen . Gene exp ession da a was ob ained using TaqMan Low Densi y A ay (TLDA) con aining 96 genes ha we e selec ed by an algo i hm and he s a is ical analyses we e pe o med in P ism by using non-pa ame ic Mann-Whi ney U es (P- alues < 0.05). Resul s: He e we show ha using a panel o 96 disease associa ed genes and measu ing mRNA exp ession le els in pe iphe al blood de i ed mononuclea cells; we could iden i y disease-speci ic gene panels ha sepa a e each disease om heal hy con ols. In addi ion, a panel o i e genes such as ADM, AQP9, CXCL2, IL10 and NAMPT disc imina es be ween all samples om pa ien s wi h ch onic in lamma ion and heal hy con ols. We also ound genes ha s a i y he diseases and sepa a e di e en sub ypes o di e en s a es o p ognosis in each condi ion. Conclusions: These indings and he iden i ica ion o i e uni e sal ma ke s o ch onic in lamma ion sugges ha hese diseases ha e a common backg ound in pa homechanism, bu s ill can be sepa a ed by pe iphe al blood gene exp ession. Impo an ly, he iden i ied genes can be associa ed wi h o e lapping biological p ocesses including changed in lamma o y esponse. Gene panels based on such ma ke s can play a majo ole in he de elopmen o pe sonalized medicine, in moni o ing disease p og ession and can lead o he iden i ica ion o new po en ial d ug a ge s in ch onic in lamma ion. Backg ound Ch onic in lamma o y diseases such as in lamma o y bowel disease (IBD; including C ohn's disease - CD and ulce a i e coli is - UC), pso iasis and heuma oid a h i is (RA) exis as a subs an ial bu den in social and economic e ms wo ldwide. Despi e he impo ance o hese dis- eases, i is s ill no clea i cha ac e is ic gene exp ession signa u es can disc imina e his g oup o diseases om heal hy con ols, he a ious diseases om each o he o whe he i is possible o s a i y he diseases based on gene exp ession changes. These ch onic condi ions ha e common ea u es such as he au oimmune o igin, he equen co-mo bidi y and * Co espondence: [email protected] 1 Depa men o Biochemis y and Molecula Biology, Uni e si y o Deb ecen, Deb ecen, Hunga y † Con ibu ed equally Full lis o au ho in o ma ion is a ailable a he end o he a icle Mesko e al. BMC Medical Genomics 2010, 3:15 h p://www.biomedcen al.com/1755-8794/3/15 Page 2 o 13 a ew genes such as IL10, IL23R, SLC22A4 and SLC22A5 ha ha e been iden i ied as con ibu o s o hei gene ic backg ound [Table 1]. Howe e hei p e alence and he issues a ec ed a e clea ly di e en . RA is a sys emic au oimmune diso de , wi h a p e a- lence be ween 0.5-1.0% [1], ha causes in lamma ion and issue damage in join s and endon shea hs. Pso iasis which is a ch onic diso de o he skin and join s whe e he pso ia ic plaques a e a eas o in lamma ion and excessi e skin p oduc ion, a ec s app oxima ely 2% o he popula ion only in he USA [2]. 1-2% o Wes e n pop- ula ions su e om IBD [3] in which he common ea- u es a e he in lamma ion o he in es ines and he au oimmune o igin. Global and selec i e gene exp ession analyses ha e al eady been pe o med in o de o ga he hin s on he mechanisms o hese medical condi ions by using human biopsy samples such as colon issue in IBD [4]; skin issue in pso iasis [5] and syno ial issue biopsy in RA [6]. How- e e pe iphe al blood is a mo e accessible sou ce o cells and may be easie o use o sc eening p ocesses. Fu he - mo e as ci cula ing pe iphe al blood mononuclea cells (PBMCs) a e key cells o in lamma ion, i may also e lec disease mechanisms. In addi ion s udies o gene exp ession p o iling o PMBCs may p o ide a mo e cos e ec i e and less in a- si e al e na i e o biopsy o in asi e measu emen s [7]. Examples o he clinical implica ions o his app oach include he analysis o human b eas cance p og ession [8] and PBMC p o iles in RA, sys emic lupus e y hema o- sus, ype I diabe es and mul iple scle osis [9]. I appea s he e o e ha gene exp ession p o iling om PBMCs is a alida ed ool o disco e y and also may be used o explo e he pa hogene ic backg ound o hese medical condi ions [10-12]. Howe e compa a i e s udies on he exis ence o dis inc and o e lapping gene exp es- sion pa e ns a e lacking. We sough o ill his gap by ca - ying ou a compa a i e analysis o pe iphe al gene exp ession pa e ns o a panel con aining 96 genes in a - ious ch onic in lamma o y diseases and heal hy con ols. Me hods Pa ien ec ui men The Resea ch E hics Commi ee o Uni e si y o Deb e- cen Medical and Heal h Science Cen e app o ed he Table 1: Known SNP - disease associa ions Gene IBD Pso iasis Rheuma oid a h i is ADAM33 NA s512625 PMID: 18560587 NA IL10 s3024505 PMID: 18836448 NA s1800896 PMID: 18615156 IL13 NA s1800925 PMID: 19554022 NA IL23R s2201841 PMID: 18338763 s11209026 PMID: 18369459 NA IL4 s2243250 PMID: 18064451 NA NA IL8 NA NA PMID: 18799095 PADI4 NA NA s2240340 PMID: 12833157 PTGS2 s20432 PMID: 16273614 NA s5275 PMID: 18381795 PTPN22 NA s1217414 PMID: 18341666 s2476601 PMID: 18466513 SLC22A4 s3792876 PMID: 17476680 s11568506 PMID: 18614543 s3792876 PMID: 15107849 SLC22A5 s3792876 PMID: 17476680 s2631367 PMID: 16255050 s2631367 PMID: 15107849 Known associa ions be ween single nucleo ide polymo phisms and ch onic in lamma o y diseases a e shown in he able wi h SNP ID and PMID numbe e e ing o he publica ion ha documen s he associa ion. NA means he e is no documen ed associa ion on SNP le el. Bold cells mean we could de ec he associa ion in ou s udy be ween ha speci ic disease and heal hy con ols. Mesko e al. BMC Medical Genomics 2010, 3:15 h p://www.biomedcen al.com/1755-8794/3/15 Page 3 o 13 clinical p o ocol and he s udy ha we e in compliance wi h he Helsinki Decla a ion. Signed in o med consen was ob ained om all heal hy and diseased indi iduals who p o ided blood sample. Inclusion and exclusion c i- e ia we e de eloped using he bes e idence cu en ly a ailable. Online supplemen is p o ided o addi ional in o ma ion abou inclusion and exclusion c i e ia [Addi- ional File 1: Figu e S2]. The s udy included 13 pa ien s wi h IBD; 15 wi h pso i- asis and 12 wi h RA, all o whom had ac i e disease and we e medica ion- ee a he ime o blood d aw [Table 2]. Blood was also ob ained om a g oup o heal hy con ol indi iduals (18 pa ien s) ha did no show signi ican di - e ences compa ed o he diseased g oups ega ding age. A e he subjec s as ed o e nigh o 12 hou s, all o he blood samples we e ob ained locally be ween 8:00 AM and 9:00 AM; and we e p ocessed wi hin one hou a e sample collec ion. Pe iphe al blood mononuclea cell collec ion and RNA p ocessing Venous pe iphe al blood samples we e collec ed by clini- cians (30 ml) in Venous Blood Vacuum Collec ion Tubes con aining EDTA (BD Vacu aine K2E). PBMCs we e sepa a ed by Ficoll g adien cen i uga ion. To al RNA was ex ac ed om PBMCs using T izol eagen (In i ogen), acco ding o he manu ac u e 's p o ocol. RNA quali y and quan i y we e checked on NanoD op and Agilen Bioanalyse 2100 (Agilen Tech- nologies). TaqMan mRNA analysis by RT-QPCR Gene exp ession da a was ob ained using TaqMan Low Densi y A ay (TLDA) (Applied Biosys ems) which is a 384-well mic o luidic ca d ha enables o pe o m 384 simul aneous eal- ime PCR uns and which has been used o gene exp ession p o iling in se e al s udies [13,14]. This low- o medium- h oughpu mic o luidic ca d allows o 2 samples o be un in pa allel agains 96 TaqMan® Gene Exp ession Assay a ge s ha a e p e- loaded in o each o he wells on he ca d. cDNA was gen- e a ed wi h High Capaci y cDNA Re e se T ansc ip ion Ki acco ding o manu ac u e 's p o ocol. 2 mic og ams o RNA we e used pe sample in he RT-PCR uns. 400 ng (4 μl) cDNA was used in each sample. 196 μl nuclease- ee wa e and 200 μl 2× TaqMan Uni e sal PCR Mas e Mix (Applied Biosys ems) we e added o he Real-Time Quan i a i e PCR measu emen s. This mix u e was hen equally di ided o e ou sample-loading po s o he TLDA, each connec ed o one se o he 96 genes o in e - es . The a ays we e cen i uged once (1', 1300 RPM on oom empe a u e) o equally dis ibu e he sample o e he wells. Subsequen ly, he ca d was sealed o p e en an exchange be ween wells. RT-QPCR ampli ica ion was pe o med using an Applied Biosys ems P ism 7900HT sequence de ec ion sys em wi h he ollowing he mal Table 2: Pa ien pa ame e s Disease S a us Con ol IBD Pso iasis Rheuma oid a h i is n18131512 Sex (male/ emale) 8/10 4/9 5/10 2/10 Age (yea s) 40.07 ± 21.3 27.92 ± 8.49 30.53 ± 9.3 45.83 ± 18.24 Clinical sub ype NA CD/UC 8/4 A h i is posi i e/ nega i e 4/11 Bone e osion posi i e/nega i e 4/8 CDAI UCDAI PASI DAS28 Clinical se e i y NA 270.4 ± 67.18 8 ± 1.2 27.47 ± 9.62 6.12 ± 1.05 124-365 6-9 15-48 4.07-7.56 De ini ion o abb e ia ions: IBD = In lamma o y bowel disease, CD = C ohn's Disease, CDAI = C ohn's Disease Ac i i y Index, UC = Ulce a i e coli is, UC DAI = Ulce a i e Coli is Disease Ac i i y Index, PASI = Pso iasis A ea Se e i y Index, DAS28 = Disease Ac i i y Sco e 28, NA = non a ailable. Da a a e p esen ed as mean ± SD; also ange in Clinical Se e i y. Mesko e al. BMC Medical Genomics 2010, 3:15 h p://www.biomedcen al.com/1755-8794/3/15 Page 4 o 13 cycle condi ions: 2 min a 50°C and 10 min a 94.5°C, ol- lowed by 40 cycles o 30 s a 97°C and 1 min a 59.7°C. Gene Lis Selec ion P ocess A da abase con aining 400 genes which a e associa ed wi h he 3 ch onic in lamma o y diseases and in lamma- ion was cons uc ed by using he da a de i ed om he Human Genome Na iga o ha lis s all he genes ela ed o a speci ic disease and he e idence he ela ion is based on; he cu en medical li e a u e and in e na ional da a- bases (e.g. OMIM, En ez Gene). Genes we e also selec ed om genome-wide associa ion s udies as well as mic oa ay analyses ocusing on skin biopsy in pso iasis, colon biopsy in IBD and syno ial luid in RA. (Figu e 1) In he second s ep, genes we e pu in o de by using a sco e lis ha was based on he numbe o publica ions men ioning he disease-gene associa ion, ela ion o o he diseases and expec ed exp ession in PBMC acco d- ing o p e ious s udies and ou pilo expe imen in which we analyzed selec ed genes in PBMC samples h ough indi idual RT-QPCR assays. This selec ion p ocess esul ed in 150 genes. In he hi d s ep, genes ha did no show exp ession (missing signal in QPCR measu emen s) in he pilo s udy we e emo ed, leading o a inal lis o 96 genes Addi ional File 1: Figu e S1. S a is ical Analyses Rela i e gene exp ession le els o each gene we e calcu- la ed by compa a i e C me hod and esul s we e no mal- ized o glyce aldehyde-3-phospha e dehyd ogenase (GAPDH) exp ession o each sample. S a is ical analyses o he no malized gene exp ession da a we e pe o med in P ism (G aphPad). Due o he ac ha ou da a did no ollow no mal dis ibu ions, he gene exp ession in g oups wi h di e en numbe s o samples was compa ed sepa a ely using he non-pa ame ic Mann-Whi ney U es . P- alues < 0.05 we e conside ed o be s a is ically signi ican . This me hod ha does no include co ec ion o mul iple compa isons is widely used in he analysis o TLDA da a [15,13] and is explained in de ails in [16]. Gene in e ac ions we e analyzed by using he Di ec In e ac ions and Biological P ocesses unc ions o Gene- Sp ing GX (Agilen Technologies). P incipal componen analysis (PCA), a s anda d, non- pa ame ic ool ha educes a complex da a se o a lowe dimension, was used in o de o e eal he in e nal s uc- u e o he da a se s and o p ojec he di e ences be ween diseased and heal hy g oups based on each se o signi ican ly changing genes. Resul s Iden i ica ion o gene panels disc imina ing ch onic in lamma o y disease pa ien s om heal hy con ols A panel o 96 genes has been selec ed by he algo i hm depic ed on Figu e 1. Pa ien s we e ec ui ed and PBMC gene exp ession was de e mined as desc ibed in Ma e i- als and Me hods. When we compa ed he gene exp es- sion esul s we ha e ound 53 genes ha show signi ican di e ences be ween diseased and heal hy samples. Fi s , we looked a he gene panels ha di e en ia e each dis- ease om he se o con ol samples. Al oge he 25 genes show signi ican di e ences be ween IBD; 16 genes be ween pso iasis; 33 genes be ween RA and con ols (Figu e 2a-c). P incipal componen analysis o hese gene se s also sepa a es he di e en g oups o samples. In o de o ge hin s o he complex ansc ip ional basis o in lamma o y diseases and o ind po en ial a ge s ha migh play a ole in he pa hogenesis o he condi ions, he in e ac ion among mul iple genes needs o be e ealed. Gene in e ac ion analyses highligh ed IFNG, IL4, IL10, MMP9 and TIMP1 in IBD; IL10, IL13 and PTGS2 in pso iasis; IL8, IL10 and PTGS2 in RA. These genes had he mos di ec in e ac ions ha migh e eal hei key ole in he pa hogene ic backg ound o he dis- eases. Gene in e ac ion analysis on each se o genes is p o ided in he supplemen ma e ial Addi ional File 1: Figu e S3-5. Gene se s showing o e lapping o di e en ial exp ession In o de o analyze he gene exp ession pa e ns o hese condi ions, i s we c ea ed a Venn diag am highligh ing hose genes ha di e en ia e be ween a disease and he se o con ol samples (Figu e 3a). Figu e 1 Flowcha o gene selec ion p ocess. A low cha o he gene selec ion p ocess shows he s eps h ough which we chose he inal 96 genes ou o he da abase con aining 400 genes. Mesko e al. BMC Medical Genomics 2010, 3:15 h p://www.biomedcen al.com/1755-8794/3/15 Page 5 o 13 Pe o ming Gene In e ac ion Analysis esul ed in he iden i ica ion o a ne wo k ha highligh s he genes wi h he mos in e ac ions such as CCL5, IFNG, IL4, IL8, IL10, IL13, MMP9, PTGS2 and TIMP1 (Figu e 3b) clea ly showing ha no only indi idual genes bu en i e ne - wo ks a e impac ed and hence can be iden i ied. Al hough he e a e genes ha showed disease speci ic signa u es, all o hese genes showing signi ican di e - ences be ween diseased and heal hy samples o m a ne - wo k which migh ep esen he common backg ound o he pa hogenesis o his ype o in lamma ion. As he nex s ep in ou analysis all he signi ican ly changing genes we e assigned o unc ional ca ego ies ha we e c ea ed based on EASE, he Exp ession Analy- sis Sys ema ic Explo e ha p o ides s a is ical me hods o disco e ing biological hemes wi hin gene lis s [17]. Genes we e g ouped in o in lamma o y esponse; cell g ow h and main enance; p o eolysis; me abolism and unclassi ied (including genes in unique ca ego ies o genes wi hou ca ego ies) ca ego ies wi h 21, 11, 7, 9 and 7 genes espec i ely. In o de o illus a e he simila i ies in he pa hogene ic backg ound o hese diseases, he unc ional ca ego ies, he numbe o genes and he dis- ease gene panels a e isualized. I shows ha he e a e simila numbe o genes in he same unc ional ca ego ies in each o he diseases, hough he numbe o unique genes (11 in IBD, 6 in pso iasis and 21 in RA) sepa a ing only a disease om he con ol g oup is also high (Figu e 4). Diseases sub ype s a i ica ion by di e en ially exp essed genes We also sough o iden i y po en ial ma ke s ha di e - en ia e be ween dis inc sub ypes o s a es o p ognosis in each o he h ee diseases o s a i y he disease based on gene exp ession pa e ns (Figu e 5). Analyses o indi id- ual genes a e p o ided in he supplemen ma e ial Addi- ional File 1: Figu e S7-9. Figu e 2 Fold changes o genes di e en ia ing be ween diseased and con ol samples. Fold change alues o genes, showing s a is ically sig- ni ican (Mann-Whi ney U es ) di e en ial exp ession be ween diseased and con ol pa ien s, we e gene a ed om RT-QPCR measu emen s and ep- esen he di e ence o he means o he diseased and con ol g oups. P incipal componen analysis was pe o med and sepa a es he wo g oups o samples. 2a ep esen s he IBD, 2b he pso iasis and 2c he RA gene panel. Mesko e al. BMC Medical Genomics 2010, 3:15 h p://www.biomedcen al.com/1755-8794/3/15 Page 6 o 13 Rega ding IBD, ADAM10, CUGBP1, CUGBP2, DUSP1, ETS2, FOS, IL6, IL18 and PGRMC1 show signi ican ly highe exp ession le els in pa ien s wi h CD compa ed o pa ien s wi h UC (Figu e 5a). In pso iasis, CCL4, GZMK and KLRF1 show signi i- can ly highe , while ADM, ANXA3, IL4, MMP9 and OLR1 show signi ican ly lowe exp ession le els in pa ien s wi h a h i is nega i e pso iasis compa ed o pa ien s wi h symp oms o a h i is (Figu e 5b). HBB and PADI4 show lowe ; GZMK shows highe exp ession in RA pa ien s wi h bone e osions con i med wi h Magne ic Resonance Imaging compa ed o hose wi hou such symp oms (Figu e 5c). Pe iphe al blood de i ed uni e sal ma ke s o ch onic in lamma ion Impo an ly, we ha e iden i ied i e genes, ADM, AQP9, CXCL2, IL10 and NAMPT ha show signi ican di e - ences in exp ession le els be ween all he h ee condi- ions and con ol samples (Figu e 6). These genes migh be conside ed uni e sal ma ke s o ch onic in lamma ion in PBMCs. Rega ding pa hway analysis ha ea u es he biological p ocesses he genes a e associa ed wi h; IL10 has he highes numbe o ela ed biological p ocesses (o e 20) including lymphocy e, T and B cell p oli e a- ion, in e leukin p oduc ion and mac ophage ac i a ion. ADM is ela ed o 10 p ocesses such as ac i a ion o MAPK-ac i i y and in e leukin-6 p oduc ion; while AQP9, CXCL2 and NAMPT a e only associa ed wi h a single p ocess, u ea anspo , in e leukin-18 p oduc ion and neu ophil apop osis, espec i ely (Figu e 7). We also de e mined he numbe o diseased pa ien s wi h gene exp ession le els abo e he ange o he mean + SD o 2 SD o con ol samples. Addi ional File 1: Figu e S6. Discussion I is o c i ical impo ance o iden i y new bioma ke s and hin s o mechanisms o disease s a es and disease p o- g ession in illnesses a ec ing la ge popula ions such as ch onic in lamma o y diseases in o de o pe sonalize ea men and accu a ely moni o disease p og ession. Me hods based on genomic app oaches and amongs hem global gene exp ession p o iles play an inc easing ole in such e o s. The pa homechanism o hese ch onic in lamma o y diseases has been examined a di e en le els including cell cul u es, animal models and clinical samples. Impo - an ly, as he molecula signa u e o disease ac oss issues is o e all mo e p ominen han he signa u e o issue exp ession ac oss diseases [18], he e is good eason o belie e one can de ec disease-speci ic gene exp ession signa u es in PBMC samples which a e easy- o-access. The e ha e been s udies analyzing he gene exp ession pa e ns o hese diseases in PBMC samples which led o he iden i ica ion o genes, which we also ound o be sig- ni ican ly di e en be ween disease and con ol pa ien s. Examples include PTGS2 in pso iasis [10]; PTPN22 in RA [19] and GZMK in IBD [12]. O he g oups compa ed Figu e 3 Genes sepa a ing ch onic in lamma o y diseases om con ols. (a) Venn diag am shows all o hose genes ha show signi ican di e - ences be ween each disease g oup and con ol samples. Each se con ains genes ha sepa a e con ol samples only om he pa icula disease o diseases. Unde lined genes we e down- egula ed compa ed o heal hy con ols. (b) Gene in e ac ion analysis in GeneSp ing GX in Di ec In e ac ions mode highligh s 28 genes ha ha e di ec in e ac ions wi h each o he while 25 genes ha e no di ec in e ac ions. Genes wi h he highes numbe o in e ac ions a e shown in ex ended size. Cyclooxygenase 2 enzyme is loca ed in he middle o his pa hway ne wo k. The genes ha showed sig- ni ican di e ences be ween one o he diseased g oups and heal hy con ols ha e colo codes in which blue ep esen s RA, g een ep esen s IBD and ed codes o pso iasis- ela ed genes. Mesko e al. BMC Medical Genomics 2010, 3:15 h p://www.biomedcen al.com/1755-8794/3/15 Page 7 o 13 Figu e 4 Func ional ca ego iza ion o signi ican ly changing genes. We ound 53 genes ha show signi ican ly di e ences be ween ch onic in- lamma o y diseases and con ol samples. These genes we e g ouped in o unc ional ca ego ies made by EASE. Genes we e g ouped in he ca ego y ha had he highes ele ance acco ding o he EASE so wa e. Colo s ep esen disease g oups. The size o he diag ams co ela es wi h he numbe o genes in each se . O e lapping se s o genes a e shown as o e lapping diag ams. Figu e 5 Co ela ion be ween gene exp ession le els and clinical pa ame e s. No malized mRNA le els o genes showing signi ican di e ences be ween pa ien s wi h (a) C ohn's disease and Ulce a i e coli is; (b) pso iasis pa ien s wi h and wi hou a h i is; (c) RA pa ien s wi h and wi hou MR- con i med bone e osion we e gene a ed om RT-QPCR measu emen s and ep esen he means o he exp ession le els o he diseased and con ol g oups. Mesko e al. BMC Medical Genomics 2010, 3:15 h p://www.biomedcen al.com/1755-8794/3/15 Page 8 o 13 di e en ch onic diso de s o each o he s such as ju enile a h i is and spondyloa h opa hy [20] o pso ia ic a h i is, RA and spondyloa h opa hy [21], bu such an ex ended numbe o genes ha e no ye been ound in pe iphe al blood. Impo an ly he gene exp ession p o- iles o RA, IBD and pso iasis ha e no been compa ed p e iously. We ied o ill his gap wi h ou s udy and iden i y ma ke s o ch onic in lamma ion ha o e lap among he 3 condi ions and also ones ha disc imina e hem om each o he . I should be no ed hough gene exp ession analyses o hese diseases p o ide li le insigh in o he pa homecha- nisms o he disease simply because i is impossible o so ou wha is cause and wha is consequence. Howe e i can s ill p o ide hin s o clues abou he po en ial pa h- ways a ec ed. The e o e such da ase s, al hough indi- ec ly, can o m he basis o mo e mechanis ic s udies. This is why we ha e ca ied ou and p esen ed he gene in e ac ion analyses o he genes iden i ied and linked o he di e en diseases. Signi ican ly he exis ence o o e - lapping gene se s sugges s ha he e a e common pa h- ways impac ed among he s udied diseases. A common cha ac e is ic o all diseases is an al e ed in lamma o y esponse. In addi ion, gene ic associa ion s udies ha e al eady iden i ied SNPs linked o hese diseases. A compa ison o ou da ase o hese can p o ide u he suppo o hose indings [Table 1]. Se e al o he now iden i ied di e en ially exp essed genes ha e been linked o he s udied diseases al hough ei he a a di e en le el o di e en model. Rega ding he IBD gene panel, OLR1/LOX1 [22] and CCR1 [23] bo h ha e immune esponse unc ion and a e exp essed in human in es inal cells. The homophilic liga ion o Figu e 6 Pe iphe al blood de i ed uni e sal ma ke s o ch onic in lamma ion. No malized mRNA le els o genes (wi h SD) showing signi ican di e ences be ween all he samples o ch onic in lamma o y diseases and heal hy con ols we e gene a ed om RT-QPCR measu emen s. Mesko e al. BMC Medical Genomics 2010, 3:15 h p://www.biomedcen al.com/1755-8794/3/15 Page 9 o 13 CEACAM1, which is a cell-su ace molecule, on T cells leads o he inhibi ion o a ange o T-cell unc ions he e- o e i migh be a new po en ial he apeu ic a ge in he ea men o IBD [24]. MMP9, a gela inase, ha showed a 7- old up- egula ion in diseased samples in ou s udy is an impo an media o o issue inju y in coli is and is up- egula ed in spu um samples o CD pa ien s [25]. Ga g e al. sugges ed ha de eloping s a egies o block MMP-9 ac i i y in he gu migh be o bene i o IBD [26]. TIMP1 is a na u al inhibi o o ma ix me allop o einases (MMPs) and migh be conside ed a po en ial d ug a ge . The DNA me hyla ion o IFNG ha egula es immune esponse was analyzed wi hin he mucosal compa men in bo h no mal and IBD popula ions [27]. PTGS2/COX2 is ela ed o IBD exp ession le el in colon biopsy [28]. ABCA1, BASP1, GZMK, HBB, SEC14L1, THEM2 and TNFAIP6 ha e ne e been associa ed wi h IBD; he e o e u he analyses should be ca ied ou . In he pso iasis panel, CCR1 ha plays a ole in in lam- ma o y esponse is associa ed wi h pso iasis ia he anal- ysis o skin lesion samples [29]. Koczan e al. iden i ied pai s o genes such as PTGS2/COX2 and NR1D2 which allowed an accu acy o disease s age p edic ion o 86%, based on gene exp ession pa e ns [10]. ADORA2B, IFRD1, ITGA2B, MAP4K1, PADI4, PDE4A, SEC14L1 and TREM1 ha e no documen ed associa ion wi h pso i- asis ye . The e o e hese a e candida es o u he SNP/ gene ic associa ion o mechanis ic s udies. Rega ding he signi ican ly changing genes in RA sam- ples, IL8 is ela ed o RA based on exp ession [30]. Van De Voo e al. demons a ed ha he exp ession o ADAM10 is s ongly enhanced in RA syno ia [31]. CCL4 was exp essed in PBMC [30] and CCL5 showed high se um le els [32]. Kehlen e al. ound up- egula ion o TNFAIP6 in ib oblas -like syno iocy es o pa ien s wi h RA [33]. HMGB1 [34] and PLCB1 [35] only showed di - Figu e 7 Pa hway analysis o uni e sal ma ke s o ch onic in lamma ion. Pa hway analysis was ca ied ou wi h GeneSp ing GX Biological P o- cesses mode ha e ealed all he biological p ocesses he genes a e ela ed o. Ci cles ep esen he 5 genes; del oids ep esen biological p ocesses.