Circulating cytokines in Norwegian patients with psoriatic arthritis determined by a multiplex cytokine array system

Ci cula ing cy okines in No wegian pa ien s wi h pso ia ic a h i is de e mined by a mul iplex cy okine a ay sys em P. Szodo ay*, P. Alex 1, *, C. M. Chappell-Woodwa d 2, *, T. M. Madland 3 , N. Knowl on 1 , I. Dozmo o 1 , M. Zehe 4 , J. N. Ja is 5 , B. Nakken 6 , J. G. B un 3 and M. Cen ola 1 Objec i es. Se um cy okines play an impo an ole in he pa hogenesis o pso ia ic a h i is (PsA) by ini ia ing and pe pe ua ing a ious cellula and humo al au oimmune p ocesses. The aim o his s udy was o desc ibe a b oad spec um o T- and B-cell cy okines, g ow h ac o s and chemokines in pa ien s wi h PsA and heal hy indi iduals. Me hods. A no el p o ein a ay sys em, deno ed as mul iplex cy okine assay was u ilized o measu e simul aneously he le els o 23 ci cula ing cy okines o pa ien s wi h PsA and heal hy indi iduals. Addi ionally, co ela ional clus e ing and disc iminan unc ion analysis (DFA), wo mul i a ia e, supe ised analysis me hods we e employed o iden i y a subse o bioma ke s in o de o desc ibe po en ial unc ional in e - ela ionships among hese pa hological cy okines and iden i y bioma ke s wi h p ognos ic and diagnos ic u ili y. Resul s. Uni a ia e analysis demons a ed ha se um le els o a complex se o immune and in lamma o y modula ing cy okines a e signi ican ly up- egula ed in pa ien s wi h PsA ela i e o una ec ed con ols including in e leukin (IL)-10, IL-13, in e e en (IFN)-, epide mal g ow h ac o (EGF), ascula endo helial g ow h ac o (VEGF), ib oblas g ow h ac o [CCL3 mac ophage in lamma o y p o ein (MIP)-1], CCL4 (MIP-1) and CCL11 (Eo axin), while g anulocy e-colony s imula ing ac o was signi ican ly educed in PsA pa ien s. Co ela ional clus e ing was able o disc imina e among, and hence subclassi y, pa ien s wi h a ying le els o disease ac i i y, which may p o e use ul in guiding he apy in hese appa en ly pheno ypically dis inc disease subse s. DFA iden i ied EGF, IFN-, VEGF, CCL3 (MIP-1) and IL-12p40 as analy es wi h he s onges disc imina o y powe among a ious PsA pa ien s and con ols. Conclusions. Ou indings sugges ha hese ac o s modula e PsA pa hology and he a icula in ol emen in a syne gis ic manne . Iden i ying ac o s could be used in he de elopmen o clinical diagnos ic es s, which a e aluable o guide e idence- based diagnosis and disease managemen o PsA. KEY WORDS: Pso ia ic a h i is, Ci cula ing cy okines, Hie a chical clus e ing, Disc iminan unc ion analysis, Disease ac i i y, Mul iplex cy okine a ay sys em. In oduc ion A h i ides, associa ed wi h pso iasis is a he e ogeneous disease en i y consis ing o monoa icula , oligoa icula and/o polya icula ype o pe iphe al join in ol emen . The disease is also cha ac e ized by e osi e mani es a ions, bony pe ios eal eac ion, exube an p oli e a ion a si es o en hesis and he p esence o bony ankylosis [1]. Pso ia ic a h i is (PsA) has been diagnosed in 20% o hose who ha e pso iasis and mani es s in mos pa ien s be ween he ages o 20 and 50 y s [1]. In addi ion o he pe iphe al join syno i is and spon aneous join usion, pa ien s p esen wi h de ma ological mani es a ions o pso iasis, heuma oid ac o (RF) se onega i i y and human leucocy e an igen (HLA) associa ions [2]. The pa hological p ocess o skin and join lesions in PsA is p ima ily media ed by au oimmune in lamma o y eac ion and a pa hogenic in e play o gene ic and en i onmen al ac o s [3]. The p ominen in lamma o y lymphocy ic in il a e in he skin de mal papillae, he join s oma and a in lamma o y en hesis e lec s a complex cy okine p o ile in PsA [1]. Cy okines, as molecula media o s o he in lamma o y p ocess in PsA including umou nec osis ac o (TNF)-[4, 5], ha e p e iously been desc ibed as c ucial ac o s in he pa hogenesis o he disease. O he cy okines, such as in e leukin (IL)-10, IL-12, IL-13, IL-18 and ascula endo helial g ow h ac o (VEGF), ha e also been desc ibed o ha e a pa hogenic ole in PsA [6–9]. These s udies sugges ha cy okines sec e ed om ac i a ed T- and B- cells and o he immunocompe en cells induce p oli e a ion and ac i a ion o he syno ial and epide mal ib oblas s leading o clinical mani es a ions o a h i is B oegelmann Resea ch Labo a o y, The Gade Ins i u e, Uni e si y o Be gen, Be gen, No way, 1 Oklahoma Medical Resea ch Founda ion, Oklahoma Ci y, OK, USA, 2 Depa men o Pa hology, Uni e si y o Oklahoma Heal h Sciences Cen e , Oklahoma Ci y, OK, USA, 3 Depa men o Rheuma ology, Haukeland Uni e si y Hospi al, Uni e si y o Be gen, Be gen, No way, 4 Di ision o Clinical Immunology, 3 d Depa men o Medicine, Medical and Heal h Science Cen e , Uni e si y o Deb ecen, Deb ecen, Hunga y, 5 Depa men o Pedia ics, Uni e si y o Oklahoma College o Medicine, Oklahoma Ci y, OK, USA and 6 Depa men o Biochemis y and Molecula Biology, Medical and Heal h Science Cen e , Uni e si y o Deb ecen, Deb ecen, Hunga y. Submi ed 3 Feb ua y 2006; e ised e sion accep ed 21 June 2006. Co espondence o: P. Szodo ay, MD, PhD, B oegelmann Resea ch Labo a o y, Uni e si y o Be gen, A maue Hansen Building N-5021 Be gen, No way. E-mail: [email p o ec ed] *Con ibu ed equally o his wo k. Rheuma ology 2007;46:417–425 doi:10.1093/ heuma ology/kel306 Ad ance Access publica ion 27 Augus 2006 417 ß2006 The Au ho (s) This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion Non-Comme cial License (h p://c ea i ecommons.o g/licenses/by-nc/2.0/uk/) which pe mi s un es ic ed non-comme cial 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. a Uni e si y o Deb ecen, Facul y o Medicine, Cen al Lib a y on May 5, 2012h p:// heuma ology.ox o djou nals.o g/Downloaded om in hese pa ien s. The impo ance o his class o disease media o s is unde sco ed by he highly e icacious ac ion o an i-cy okine-based biological he apies in PsA and o he in lamma o y a h i ides [10, 11]. Ci cula ing cy okines he e o e e lec he ac i a ion s a us o ongoing in lamma o y p ocesses and he e alua ion o se um cy okines is a good indica o and a eliable su oga e ma ke o disease ac i i y in in lamma o y a h i ides [12, 13]. Clinical s udies o cy okine unc ion in in lamma o y a h i is can be con o e sial and e en con adic o y. The oles o cy okines in heal h and disease can be blu ed due o he complexi y o cy okine mechanics, including kine ics o exp es- sion, mode o induc ion, egula ion o ecep o exp ession, compe i ion o occupancy, and syne gy o ac ion, all c i ical aspec s o hei ne e ec . Mo eo e , he e a e di e ences in pa hological mani es a ions, disease se e i y and d ug esponse among pa ien s ha mus be conside ed in coho s udies o cy okine unc ion o accu a ely de ine hei ole in heal h and disease [14]. B oad-based p o eomic sc eening me hods a e beginning o e eal he ich epe oi es o cy okines a play in a gi en o m o au oimmune diseases o in lamma o y a h i ides [15, 16]. Mo eo e , inc easingly sophis ica ed mul i- a ian analyses me hods ha e been de eloped o aid in ou unde s anding o he complex egula o y ne wo ks o hese disease media o s [17–19]. The aim o his s udy was o i s ca alogue a b oad spec um o lymphokines, monokines, chemokines, g ow h and angiogenic ac o s in he pe iphe y o PsA pa ien s, hen model pu a i e egula o y ne wo ks among hese ac o s, using mul i a ia e bios a is ical me hods and clinical co ela ions. In doing so, we ha e e i ied he p esence o p e iously cha ac e ized and iden i ied no el cy okines up- egula ed in PsA se a, media o s wi h bo h diagnos ic and mechanis ic po en ial. We ha e also cha ac e ized likely in e play among hese cy okines using mul i- a ia e analysis me hods ha model ne wo k-like beha iou s among biological a iables. Ma e ials and me hods Pa ien s and con ols The s udy popula ion consis ed o 43 No wegian pa ien s wi h PsA o polya icula ype who we e ec ui ed om he ou -pa ien clinic a he Depa men o Rheuma ology, Haukeland Uni e si y Hospi al, Be gen, No way (24 emales and 19 males; mean age 53.77 y s, ange 26–74) ul illing he diagnosis o PsA acco ding o he c i e ia o Moll and W igh [20]. Also 25 age- and sex-ma ched heal hy con ols we e en olled in he s udy. These con ols we e heal hy blood dono s, wi hou any sign o a h i is, pso iasis o ongoing in lamma ion. Con ols we e medica ion- ee o a leas 3 mon hs p io o his s udy. To assess he p esen disease ac i i y, clinical examina ion o 52 join s was pe o med ( he EULAR 44 join coun added o DIP join s o hands). Skin a ec ion by pso iasis was assessed using he PASI sco e [21]. Be o e inclusion in he s udy, in o med consen was signed by each pa ien . Gene al labo a o y and immunolabo a o y assessmen s included e y h ocy e sedimen a ion a e, C- eac i e p o ein, whi e blood cell coun , haemoglobin, an i-nuclea an ibody and RF. Clinical and labo a o y cha ac e is ics o he pa ien s included in he s udy a e summa ized in Table 1. Se um samples Blood samples we e ob ained om bo h pa ien s and una ec ed con ols a e in o med consen and ea ed anonymously h oughou he analysis. Blood was collec ed in endo oxin- ee silicone-coa ed ubes wi hou addi i e. The blood samples we e allowed o clo a oom empe a u e o 30 min be o e cen i uga- ion (3000 pm, 48C, 10 min), he se um was emo ed and s o ed a 808C un il analysed. Mul iplex cy okine assay Se um le els o cy okines and chemokines, including IL-1, IL-2, IL-4, IL-5, IL-6, CXCL8 (IL-8), IL-10, IL-12 (p40), IL-13, IL-15, IL-17, in e e on (IFN)-and IFN-, TNF-, epide mal g ow h ac o (EGF), VEGF, basic ib oblas g ow h ac o (FGF), g anulocy e-colony s imula ing ac o (G-CSF), g anulocy e- mac ophage colony s imula ing ac o (GM-CSF), CCL2 [monocy e chemoa ac an p o ein (MCP)-1]/(MCAF), CCL3 [mac ophage in lamma o y p o ein (MIP)-1], CCL4 (MIP-1) and CCL11 (Eo axin) we e measu ed using a bead-based immuno luo escence assay (Luminex Inc. Aus in, TX, USA) using mul iplex cy okine eagen s supplied by Biosou ce In e na ional, Cama illo, CA, USA as p e iously desc ibed [16, 22]. B ie ly, a sandwich immunoassay-based p o ein a ay sys em (Biosou ce In e na ional), which con ains dyed mic o- sphe es conjuga ed wi h a monoclonal an ibody speci ic o a a ge p o ein was used in his assay. Se um samples we e hawed and un in duplica es. An ibody-coupled beads we e incuba ed wi h he plasma sample (an igen) a e which hey we e incuba ed wi h bio inyla ed de ec ion an ibody be o e inally being incu- ba ed wi h s ep a idin–phycoe y h in. A b oad sensi i i y ange o s anda ds (Biosou ce In e na ional), anging be ween 1.95 and 32 000 pg/ml we e used o help enable he quan i a ion o a dynamic wide ange o cy okine concen a ions and p o ide he g ea es sensi i i y. This cap u ed bead-complexes we e hen ead by he Bio-Plex a ay eade (Bio-Rad Labo a o ies, He cules, CA), which uses Luminex luo escen bead-based echnology (Luminex Co po a ion Aus in, TX, USA) wi h a low-based dual lase de ec o wi h eal- ime digi al signal p ocessing o acili a e he analysis o up o 100 di e en amilies o colou -coded polys y ene beads and allow mul iple measu e- men s o he sample ensu ing in he e ec i e quan i ica ion o cy okines. Valida ion o he mul iplex assays was pe o med using single p o ein ELISAs (Biosou ce In e na ional). Values ob ained om mul iplex assay analy es we e highly co ela i e (Spea man’s ank co ela ion coe icien , 0.97 0.03) when compa ed wi h indi idual ELISAs o pa icula cy okines. Se um le els o 23 cy okines we e compa ed among PsA pa ien s and una ec ed con ol indi iduals, also among a ious subse s o pa ien s wi h PsA. The cy okines assayed included modula o s o se e al key aspec s o disease pa hology including egula ion o in lamma ion, cellula and humo al immuni y, leukocy e a icking, cell g ow h and angiogenesis. To acili a e unc ional in e p e a ion o esul s, cy okines we e so ed in o ou unc ional g oups in he g aphical ep esen a ions o hese analyses. This included a g oup deno ed ‘cellula cy okines’ which d i e, albei no exclusi ely, cy o oxic and an i- i al esponses (e.g. IL-1, IL-2, IL-12p40, IL-15, IL-17, TNF-, IFN-, IFN-), ‘humo al cy okines’ (e.g. IL-4, IL-5, IL-6, IL-10, IL-13), ‘g ow h and angiogenic ac o s’ (e.g. EGF, VEGF, FGF basic, G-CSF, GM-CSF) and ‘chemokines’ [e.g. CCL2 (MCP-1), CCL3 (MIP-1), CCL4 (MIP-1), CCL11 (Eo axin) and CXCL8 (IL-8)]. S a is ical analysis The concen a ions o analy es in hese assays we e quan i ied using a calib a ion o s anda d cu e. A 5-pa ame e logis ic eg ession analysis was pe o med o eg ess a known se ial dilu ion o analy e s median luo escen in ensi y ead om a low cy ome e . The esul ing equa ion was hen used o p edic he concen a ion o he unknown samples [23]. 418 P. Szodo ay e al. a Uni e si y o Deb ecen, Facul y o Medicine, Cen al Lib a y on May 5, 2012h p:// heuma ology.ox o djou nals.o g/Downloaded om S a is ical di e ences in measu ed alues we e analysed using a Mann–Whi ney U- es . P- alues <0.05 we e conside ed s a is ically signi ican . All P- alues epo ed a e nominal due o he explo a o y na u e o his analysis. Ma lab R13 (Na ick, MA, USA) and S a is ica 6 (Tulsa, OK, USA) we e used o pe o m all s a is ical analyses. Sample size G oup sample sizes o 25 and 43 achie e 97% powe o de ec a di e ence o 30.0 be ween he null-hypo hesis, whe e bo h g oup means a e 30.0 and he al e na i e hypo hesis, whe e he mean o cy okine X is 60.0 wi h es ima ed g oup SDs o 30.0 and 30.0, and wi h a signi icance le el () o 0.05000 using a wo-sided Mann–Whi ney U- es assuming ha he ac ual dis ibu ion is uni o m. Mul i a ia e analysis Hie a chical clus e ing using Pea son’s co ela ion coe icien as he dis ance me ic [24, 25] was used o assess cy okine simila i y. Nex , he esul ing ‘dis ance ma ix’ (he e ep esen ing co ela- ional s eng h) was so ed by cy okine o align he mos simila cy okines close o one ano he . The eso ed dis ance ma ix was hen plo ed like a hea map using SigmaPlo 2001 (SPSS Inc., Chicago, IL, USA). Disc iminan unc ion analysis (DFA) was used o selec ion o he se o analy es ha maximally disc imina e among he g oups s udied (¼0.10), buil in a s ep-wise manne , as desc ibed p e iously [18, 19]. A DFA c ea es a disc imina e unc ion, deno ed by a oo , which is an equa ion consis ing o a linea combina ion o analy es used o he p edic ion o g oup membe ship. Once he analy es a e selec ed by DFA, a lea e-one-ou c oss- alida ion is pe o med o p o ec agains o e - i ing. The bes c oss- alida ed model is hen epo ed. The inal disc imina o y TABLE 1. Clinical and labo a o y cha ac e is ics o pa ien s wi h pso ia ic a h i is Pa ien No. Age Sex Type o pso iasis The apy No. o swollen join s No. o ende join s WBC HgB ESR CRP RF ANA 1 45 F PV MTX 2 12 5.8 12.8 8 5 0 0 2 40 M PV COX-2 MTX 0 0 6.4 16.1 10 11 0 0 3 26 F PV – 0 11 6.6 13.9 4 5 0 0 4 58 F PV COX-2 MTX 1 2 9.2 14.9 26 20 0 0 5 57 F PV COX-2 3 15 4.9 12.7 39 5 1 1 6 73 M PV NSAID 4 22 9.3 16.1 2 5 0 0 7 33 F PV COX-2 MTX 2 26 10.4 12.2 10 5 0 0 8 69 F PP NSAID 3 20 6 11.7 7 5 0 0 9 72 M PV COX-2 4 9 4.9 13.8 19 5 0 0 10 58 F PV – 2 21 4.7 13.8 21 5 0 0 11 48 M PV COX-2 MTX 4 5 15.1 15.6 20 5 0 0 12 34 M PV NSAID 0 6 6.7 16 2 5 0 0 13 57 M PV COX-2 LEF 8 29 4.8 15.4 5 5 0 0 14 61 F PV NSAID MTX 1 7 701 14.5 3 5 0 0 15 51 M PV NSAID MTX SUL 10 14 7.3 13.8 10 16 1 0 16 53 M PV COX-2 5 16 6.6 14.6 15 23 0 0 17 67 M PV NSAID 9 37 9 15.1 5 5 0 0 18 55 F PP COX-2 MTX PRED 2 5 10.7 14.9 3 5 0 0 19 50 M PV NSAID HYDCL 2 10 8.1 14.1 24 25 0 0 20 55 F PV NSAID LEF 5 27 – – – – 0 0 21 50 F PV COX-2 SUL 2 13 5.4 14 3 5 0 0 22 46 M PV COX-2 0 15 4.8 14.4 10 5 0 0 23 56 F PV NSAID 8 23 5 15 15 5 0 0 24 65 F PV COX-2 SUL 1 24 5.9 12.7 11 5 0 0 25 32 F PV COX-2 MTX TNF 1 6 7.3 13.8 13 5 0 0 26 56 M PV COX-2 1 8 6.1 16.3 1 5 0 0 27 59 M PV NSAID MTX 1 7 8.6 15.2 1 5 0 0 28 70 M PV COX-2 0 27 6.1 16.6 1 5 0 0 29 61 F PP NSAID SUL 7 8 4.6 12.4 27 16 0 0 30 44 M PV NSAID 4 4 6.3 14.9 3 5 1 0 31 74 F PV – 4 7 10.2 13.8 27 34 0 0 32 54 F PV NSAID 8 13 6.4 13.1 14 18 0 1 33 47 M PV COX-2 MTX 0 23 8.5 14.4 3 5 0 0 34 35 M PV NSAID MTX 0 0 6.7 15.3 1 5 0 0 35 45 F PV NSAID LEF 0 9 8.4 12.6 31 15 0 0 36 46 F PV NSAID MTX 7 15 4.7 12.2 32 5 0 0 37 68 F PV NSAID 1 4 9.6 14.4 20 10 0 0 38 54 F PV COX-2 2 2 10.4 14 7 5 0 0 39 56 M PV – 0 2 7.6 15.1 3 5 0 0 40 74 M PV – 0 4 3.8 14.4 10 5 0 0 41 57 F PV NSAID MTX 2 5 7.2 13.1 17 5 0 0 42 58 F PV COX-2 11 9 4.9 13.2 6 5 0 0 43 43 F PV MTX 1 7 6.5 13.3 6 5 0 0 ANA, an i-nuclea an ibody (0 ¼nega i e, 1 ¼posi i e) (no m 0.00–0.99); COX-2, selec i e cyclo-oxigenase-2 inhibi o ; CRP, C- eac i e p o ein (no m <10); ESR, e y h ocy e sedimen a ion a e (no m 0–20 mm/h ); HYDCL, hyd oxy-chlo oquine; F, emale; HgB, haemoglobin (no m: 13.2–16.6 g/ dl o men and 11.6–16.0 o women); M, male; MTX, me ho exa e, NSAID, non-s e oidal an i-in lamma o y d ugs (unselec i e), LEF, le lunomide; PP, pus ulosis palmoplan a is; PRED, p ednisolone (2.5 mg/day); PV, pso iasis ulga is; RF, heuma oid ac o (0 ¼nega i e, 1 ¼posi i e: Waale 128); SUL, sul asalazine, TNF, TNF-blocking agen (In liximab); WBC, whi e blood cell coun (no m 3.5–11.0 10 9 /l). Ci cula ing cy okines in pso ia ic a h i is 419 a Uni e si y o Deb ecen, Facul y o Medicine, Cen al Lib a y on May 5, 2012h p:// heuma ology.ox o djou nals.o g/Downloaded om powe o each analy e is cha ac e ized by a pa ial Wilk’s Lambda coe icien ; 1.0 (no disc imina o y powe ) o 0.0 (pe ec disc imina o y powe ). The disc imina o y po en ial o a DFA analysis is bes ealized h ough a simple sca e plo o he oo s. Simula ion s udies A o al o 500 Mon e Ca lo simula ion s udies we e pe o med on he hie a chal clus e analysis o es ima e he s eng h o andom co ela ions. The simula ions can be de ined as ollows: (i) Es ima e he a e age and SD o each analy e in each g oup (no mal o PsA). (ii) Gene a e 43 o 25 numbe s (depending on whe he he no mal o PsA g oup is being examined) om a no mal dis ibu ion using he es ima es ound in 1) o each analy e. (iii) C ea e a hie a chical clus e om he simula ed da a (see abo e o de ails). (i ) Repea 1–3 500 imes. ( ) De e mine he 95% pe cen ile o co ela ions and deno e i as . All alues g ea e han ha e <5% chance o being caused by andom luc ua ions in he da a. Resul s Uni a ia e compa ison o se um cy okines in pa ien s wi h PsA and una ec ed con ols Among he ‘cellula cy okines’, only IFN-(P¼0.039) was signi ican ly inc eased in PsA pa ien s compa ed wi h he le els ound in heal hy indi iduals (Fig. 1A). Among humo al cy okines, he le els o IL-10 (P¼0.0107) and IL-13 (P¼0.0006) we e signi ican ly ele a ed in pa ien s (Fig. 1B), as we e he chemokines CCL3 (MIP-1)(P<0.0001), CCL4 (MIP-1)(P<0.0001) and CCL11 (Eo axin) (P¼0.0101), and he g ow h and angiogenic ac o s EGF (P<0.0001), VEGF (P<0.0001), FGF (P<0.0001) (Fig. 1C and D). G-CSF was signi ican ly dec eased (P¼0.014) in he pa ien g oup compa ed wi h con ols (Fig. 1C). Compa ison o se um cy okine le els among PsA pa ien s wi h a ia ion in disease se e i y Pa ien s we e di ided based on swollen join coun s. G oup 1 consis ed o 8 pa ien s wi h coun s o 0; g oup 2 (12 pa ien s), 1–3, and g oup 3 (12 pa ien s), 4. Inc eases in he a e age le els IL-2, IL-12(p40), IL-15, IFN-and CCL3 appea ed ele a ed in pa ien s o g oup 3 ela i e o g oups 1 and 2 (Fig. 2A–D). These esul s sugges ha a mo e ac i e Th1/cellula immuni y may be a play in a subse o pa ien s wi h a mo e se e e disease. Acco dingly, se um le els o IL-2 appea o co ela e wi h inc easing disease se e i y among he h ee g oups (Fig. 2A–D). When he le els o hese di e en ial analy es we e e alua ed using disc iminan unc ional analysis, IFN-, IL-15 and IL-2 we e iden i ied as being signi ican ly disc imina o y be ween he h ee g oups wi h a P- alue o 0.012, 0.026 and 0.069, espec i ely (Table 2), sugges ing he po en ial ole o a T-cell signa u e in he ae iopa hogenesis o he disease. Cha ac e iza ion o coho p o iles by hie a chical clus e ing Hie a chical clus e ing was used o iden i y se s o cy okines whose exp ession changes a e linea ly ela ed wi hin a gi en popula ion, i.e. se s o cy okines whose exp ession le els among indi iduals wi hin a popula ion a e co ela ed. In his me hod, co ela ion coe icien s be ween pai s o cy okines a e posi i e o cy okines ha a e consis en ly a he same le el in indi iduals o a gi en popula ion unde s udy, nega i e o cy okines ha FIG. 1. Compa ison o a ious se um cy okine le els be ween PsA pa ien s (n¼43) and con ols (n¼25). Ba s show he mean in pg/ml and SEM. (A) Le els o ci cula ing cellula cy okines. (B) Le els o humo al cy okines. (C) Le els o g ow h ac o s. (D) Ci cula ing chemokine le els. *Signi ican ly di e en om compa ed alues by Mann–Whi ney U- es (P<0.05). 420 P. Szodo ay e al. a Uni e si y o Deb ecen, Facul y o Medicine, Cen al Lib a y on May 5, 2012h p:// heuma ology.ox o djou nals.o g/Downloaded om a e a opposi e le els (one is high, while one is low) and neu al when he e is a lack o a co ela ion in le els. Co egula ed se s o a iables will display simila ends in a iance ha can be iden i ied and used o clus e assignmen . The esul s o hese analyses we e ep esen ed in g aphical ou pu s, deno ed as hea maps. In a hea map, co ela ed cy okines a e g ouped oge he . Mo eo e , hea maps u ilize colou mapping o co ela ion coe icien s o acili a e isual assessmen o hese alues among a la ge se o a iables. Analy es wi h posi i e co ela ions a e ep esen ed in g aded shades o ed and nega i e co ela ions in g aded shades o blue, allowing hese g oups o co ela ed a iables o be eadily iden i ied by isual inspec ion as ela i ely monoch oma ic clus e s. Se e al g oss di e ences be ween he hea maps om heal hy indi iduals and PsA pa ien s we e obse ed (Fig. 3A and B). The mos p onounced was a la ge clus e o highly co ela i e cy okines iden i ied in pa ien s, which con ained IL-2, IL-12(p40), IL-15, TNF-, MIP-1, MIP-1, IFN-and FGF ha was no p esen in heal hy con ols. These esul s e lec a disease-speci ic change in he egula ion o hese cy okine le els in he pe iphe y. A subse o hese cy okines was iden i ied in he abo e uni a ia e analysis as being di e en ially exp essed. The high co ela ion alues o hese cy okines wi hin indi idual PsA pa ien s sugges s ha in addi ion o being ‘up- egula ed’ as iden i ied, hey pa icipa e in p ocesses ha esul om coo dina ion o hei le els. Gi en he collec ion o cy okines in his clus e , hese esul s unde sco e he likely unc ional ele ance o Th1/cellula immuni y in his disease, an immune esponse bias shown o be highly ele an in he skin-aspec s o PsA [26–28]. In o de o u he acili a e isual cha ac e iza ion o he di e ence in clus e pa e ns be ween pa ien and con ol clus e hea maps, an analysis, deno ed as ‘ex ac ion hea map’ me hod, was de eloped in which he hea maps be ween he con ol and PsA ma ices we e sub ac ed om each o he and he di e ences be ween he wo isually ep esen ed (Fig. 3C). Mul iple Mon e Ca lo simula ions we e used o de ine he h eshold below which signi ican co ela ions a e unlikely o appea by chance, hus p o iding a measu e o he s a is ical di e ences in he co ela ional coe icien s cha ac e ized. Using he esul s o hese simula ion s udies as a e e ence, he mos signi ican changes in co ela ion alues be ween pa ien s and con ols we e obse ed o CCL3 (MIP-1), CCL4 (MIP-1), FGF, CCL11 (Eo axin), IFN-, IL-2, TNF-, IL-15 and IL-12p40, sugges ing ha changes in egula ion o hese cy okines a e highly cha ac e is ic o PsA (Fig. 3D). These esul s could be used o de elop a mul i a ia e-based diagnos ic es o he disease, a es ha is likely o be signi ican ly mo e powe ul in e ms o disease disc imina ion han hose de eloped om simple uni a ia e s udies o up- and down- egula ion o pe iphe al cy okines. Iden i ica ion o se um cy okine le els co ela ed wi h disease ac i i y Mul i a ia e analyses can be e dis inguish a iables ha a e cha ac e is ic o disease subse s when he e ogeneous popula ions a e analysed han he uni a ia e analyses u ilized abo e. This is in pa because uni a ia e analyses a e based on iden i ying di e ences in he a e ages among popula ions, while hie a chical clus e ing dis inguishes pa e ns on an indi idual basis and a e hus una ec ed by he e ogenei y in he o e all cy okine le els FIG. 2. Compa ison o a ious se um cy okine le els among PsA pa ien s wi h a ious a ec ed join coun s. Ba s show he mean in pg/ml and SEM. (A) Le els o ci cula ing cellula cy okines. (B) Le els o humo al cy okines. (C) Le els o g ow h ac o s. (D) Ci cula ing chemokine le els. *Signi ican ly di e en om compa ed alues by Mann–Whi ney U- es (P<0.05). TABLE 2. Cy okines iden i ied by disc iminan unc ion analysis as being signi ican ly disc imina o y among PsA pa ien s wi h a ious a ec ed join coun s Wilk’s Lambda P- alue IFN-0.8385 0.0121 IL-15 0.7879 0.0263 IL-2 0.7295 0.0690 Ci cula ing cy okines in pso ia ic a h i is 421 a Uni e si y o Deb ecen, Facul y o Medicine, Cen al Lib a y on May 5, 2012h p:// heuma ology.ox o djou nals.o g/Downloaded om wi hin a gi en subg oup. Acco dingly, al hough di e ences in single cy okine alues desc ibed ea lie did no each s a is ical signi icance, di e ences in he coo dina ed egula ion o se um cy okine le els among he h ee disease ac i i y subse s desc ibed ea lie (swollen join coun s: 0, 1–3, 4) we e eadily obse ed using hie a chical clus e ing (Fig. 4A–C). These di e ences in cy okine pa e ns among disease se e i y subg oups can se e as a ‘cy okine inge p in ’ o he de elopmen o p ognos ic es s and also help guide an unde s anding o he mechanis ic di e ences among hese g oups. Fo example, he g oup 3 hea map (swollen join coun s: 4) was mos dis inc om mo e mild disease subg oups in ha i exhibi ed a high posi i e co ela ion o se e al p o-in lamma o y cy okines (Fig. 4C s Fig. 4A and B, la ge clus e s, lowe le co ne ). These esul s demons a e ha hese cy okines a e p esen in he pe iphe y as a uni in indi iduals wi h se e e disease, and may he e o e ha e a coope a i e unc ion. Disc iminan unc ion analysis (DFA) DFA is a mul i a ia e disc imina ion me hod ha uses obse ed changes in a iables o cha ac e ize he mos disc imina o y a iables amongs g oups. The diagnos ic powe o hese disc imina o y a iables is enhanced, as hey a e used addi i ely o c ea e he class disc imina ion algo i hms, deno ed oo s, de i ed om his me hod. A ‘ oo ’ is a linea combina ion o a iables wi h cons an coe icien s. DFA includes in he oo s a iables ha minimize g oup o e lap in a mul idimensional plo o oo alues (Fig. 5). Roo alues o indi iduals wi hin each o he h ee popula ions analysed o med non-o e lapping clus e s, sugges ing ha he esul s o he DFA can be used o de eloping cy okine-based se um diagnos ic c i e ia ha could dis inguish hese g oups h ough a clinical blood es . The cy okines EGF, IFN-, VEGF, CCL3 (MIP-1) and IL-12p40 we e iden i ied as ha ing he highes disc imina o y ac i i y. This is a subse o he cy okines iden i ied by uni a ia e me hods dis inguishing pa ien s om con ols. Also, he all bu one o hese cy okines we e cha ac e ized as disc imina o y by hie a chical clus e ing, highligh ing he po en ial signi icance o hese media o s in PsA. Discussion In lamma o y in il a e o lymphocy es in he de mal papillae o he skin and he s omal lining o he join s in PsA esul s in he collagenase clea age o ca ilage collagen ea ly in he disease and is achie ed h ough he elease o a ious chemokines and cy okines h ough cy okine-d i en p oduc ion o p o eases [2]. The aim o his s udy was o desc ibe subpopula ions o ci cula ing cy okines likely o d i e a b oad spec um o FIG. 3. Hie a chical clus e ing compa ing PsA pa ien s and con ols. A hea map ep esen a ion o he co ela ional coe icien s a e g aphed. Cy okines wi h posi i e co ela ions a e ep esen ed in g aded shades o ed and nega i e co ela ions in g aded shades o blue. The same o de o he analy es along axis is used o all he h ee hea maps. (A) Con ols, (B) PsA pa ien s and (C) ex ac ion hea map di e ences be ween he hea map ep esen a ions o pa ien s and con ols we e isola ed and depic ed in an ex ac ion hea map. Rep esen a ion o he le el o co ela ions is lis ed on he side o he g aph. (D) G aphical ep esen a ion o Mon e Ca lo simula ions ha we e used o de ine he h eshold abo e which signi ican co ela ions a e unlikely o appea . 422 P. Szodo ay e al. a Uni e si y o Deb ecen, Facul y o Medicine, Cen al Lib a y on May 5, 2012h p:// heuma ology.ox o djou nals.o g/Downloaded om au oimmune p ocesses in pa ien s wi h PsA. Besides he desc ip- ion o he di e ences in he ci cula ing cy okine ne wo k be ween heal hy indi iduals and pa ien s, we u he subca ego ized he pa ien s based on hei a ec ed join coun s as a e lec ion o he disease se e i y. We ha e iden i ied a subse o cy okines ha signi ican ly di e be ween con ols and PsA, as well as di e en ia e be ween a ious subse s o pa ien s wi h PsA. We ha e cha ac e ized di e en cy okine p o iles acco ding o he p esen disease ac i i y, compa ed hese disease g oups and pinpoin ed special subse s o cy okines ha a e a ibu ed o he pa icula disease subse . The key cy okines ha disc imina e be ween PsA and heal hy indi iduals we e IL-10, IL-13, IFN-, EGF, VEGF, FGF, CCL3 (MIP-1), CCL4 (MIP-1), CCL11 (Eo axin) and G-CSF. These indings suppo he hypo hesis ha a b oad spec um o impai ed immune unc ions a e in ol ed in he pa hogenesis o PsA, a ec ing cellula , humo al immune esponses and a ious leucocy e unc ions. Al hough a ew o he cy okines ela ed o PsA ha e been desc ibed p e iously [9–11], he majo i y o hese signi ican cy okines ha e no been iden i ied in PsA. Wi h he ad anced mul iplex cy okine a ay sys em, a no el and dis inc PsA- ela ed se um cy okine pa e n has been e ealed ha disc imina ed be ween pa ien s and heal hy indi iduals. Since limi ed sys em in o ma ion can be ob ained om uni a ia e analysis on he syne gis ic and/o an agonis ic FIG. 4. Hie a chical clus e ing compa ing PsA pa ien s wi h a ious a ec ed join coun s. A hea map ep esen a ion o he co ela ional coe icien s a e g aphed. Cy okines wi h posi i e co ela ions a e ep esen ed in g aded shades o ed and nega i e co ela ions in g aded shades o blue. Rep esen a ions o he le el o co ela ions a e lis ed on he side o he g aphs. (A) PsA pa ien s wi h 0 a ec ed join s, (B) PsA pa ien s wi h 1–3 a ec ed join s and (C) PsA pa ien s wi h 4 a ec ed join s. FIG. 5. A g aphical ep esen a ion o he disc imina o y po en ial o disc iminan unc ion analysis (DFA). The analysis was used o selec a iables ha maximally disc imina e among he coho , ep esen ed by a mul idimensional plo o oo alues. Ci cula ing cy okines in pso ia ic a h i is 423 a Uni e si y o Deb ecen, Facul y o Medicine, Cen al Lib a y on May 5, 2012h p:// heuma ology.ox o djou nals.o g/Downloaded om mul i ace ed ne wo k o cy okines d i ing he immune/ au oimmune p ocesses, a mo e complex app oach u ilizing a ious mul i a ia e s a is ical analyses was used in his s udy. Hie a chical clus e ing was u ilized o iden i y he changes in cy okine ne wo ks, as his me hod clus e s cy okines, whose ela i e alues a e co ela ed on indi idual basis wi hin a popula ion. In his analysis, CCL3 (MIP-1), CCL4 (MIP-1), FGF, CCL11 (Eo axin), IFN-, IL-2, TNF-, IL-15 and IL-12p40 we e ound o dis inguish be ween PsA pa ien s and heal hy indi iduals. Finally, DFA pinpoin ed EGF, IFN-, VEGF, CCL3 (MIP-1) and IL-12p40 o disc imina e among con ols and subse s o PsA pa ien s. The pa allel e alua ion o uni a ia e analysis, hie a chical clus e ing and DFA, besides o he s, e ealed CCL3 (MIP-1) o be he majo dis inc i e cy okine be ween PsA pa ien s and heal hy indi iduals. CCL3 (MIP-1) is in ol ed in he cell ac i a ion o human g anulocy es and appea s o be in ol ed in acu e neu ophilic in lamma ion. Also, i s imula es he p oduc ion o eac i e oxygen species in neu ophils and he elease o lysosomal enzymes. Fu he mo e, CCL3 (MIP-1) induces he syn hesis o o he p o-in lamma o y cy okines, such as IL-1, IL-6 and TNF-in ib oblas s and mac ophages [29–32]. These indings suppo he idea ha his chemokine can play a c ucial p o-in lamma o y ole in he pa hogenesis o PsA and s ongly associa ed wi h a h i is de elopmen . Fu he in es iga ions, u ilizing he mul iplex cy okine assay may be used o desc ibe he in si u cy okine milieu in PsA syno ial luid (SF)/syno ium. Recen ly, he exp ession o p o- in lamma o y cy okines in he syno ial biopsy o PsA pa ien s has been desc ibed [33]. PsA syno ium was cha ac e ized by a p edominan exp ession o IL-10, IL-15, IFN-, IL-1, and TNF-[33]. In he s udy by Pa sch e al., he concen a ions o T cell- de i ed cy okines in he syno ial luids (SFs) o pa ien s wi h PsA in compa ison wi h heuma oid a h i is (RA) and os eoa h i is (OA) was desc ibed [34]. They ound ha he pa e n o T cell- de i ed cy okines in PsA SFs was simila o ha o RA SFs. Howe e , bo h he equency and he concen a ions o cy okines we e lowe in PsA SFs han in RA SFs. The p esence o Th1 and Th2 cell-de i ed cy okines in PsA SFs sugges s he p esence o ac i a ed T cells in he in lamed join issues and hei pa icipa- ion in he immuno-in lamma o y e en s [34]. The assessmen o p o-in lamma o y cy okines in he SF o pa ien s wi h PsA in compa ison wi h RA and OA showed ha he le els o TNF-, IL-1, and IL-8 we e signi ican ly highe in PsA SF han in OA SF, al hough lowe han in RA SF [35]. The pa e n o exp ession o p o-in lamma o y cy okines seen in PsA is simila o ha in RA. Since PsA is also a des uc i e a h opa hy, cy okines, in pa icula TNF-and IL-1, may be he p inciple ac o s in join des uc ion [35]. Ea ly immune ascula emo phology and dys egula ed angio- genesis ha e been hypo hesized as p ominen ea u es linked o he speci ic up- egula ion o g ow h ac o s in PsA [2]. Ou da a is consis en wi h his hypo hesis wi h angiogenic g ow h ac o s like EGF and VEGF being iden i ied as signi ican ly disc imina o y be ween pa ien s and una ec ed con ols, u he e lec ing he ole o hese playe s in he molecula and cellula mechanisms esponsible o he ascula emo phology and he o ma ion o pa hological new bone in PsA. Ou esul s imply ha a complex diso de o sec e ed se um cy okine le els d i ing a ious immuno-compe en cell ypes can be ound in PsA pa ien s. Also, di e en pa e ns o ci cula ing cy okines we e de ec ed acco ding o di e en disease ac i i ies o PsA. We assume ha he u iliza ion o he mul iplex cy okine a ay sys em in PsA p o ides a powe ul ool o sub-ca ego ize he disease and, along wi h common clinical and labo a o y pa ame e s, help o e alua e disease ac i i y. In his s udy, we ocused on s a ic ime poin s o desc ibe he cy okine p o ile o he pa ien s, wi h simila disease ac i i y and ha e compa ed hem wi h una ec ed heal hy con ols. Se ial ollow-ups o hese pa ien s may esul in luc ua ions in cy okine le els ha some imes accompanies disease la e o emissions as is e iden om p o iles iden i ied in o he ongoing s udies (ou unpublished obse a ions). This alida es he u ili y o he mul iplex assay as an impo an applica ion o moni o he - apeu ic esponse o a ious ea men p o iles in pa ien s wi h PsA. A u u e impo an applica ion o his echnology will be o e alua e he ci cula ing cy okine pa e n and he eby help sub- ca ego izing pa ien s wi h PsA and de elop his echnology o become a powe ul diagnos ic ool. By he simul aneous moni o - ing o gene al labo a o y alues and se um mul iplex cy okine le els p ognos ic, he apeu ic assessmen s and moni o ing can be achie ed [e.g. an i-CCL3 (MIP-1) he apy]. Recen ly, o he biological agen s ha e been s udied in PsA. A g oup o d ugs is being de eloped, which inhibi T cells by blocking he ‘second signal’ o T-cell ac i a ion (e.g. ale acep , e alizumab, aba acep ) [36]. In hese s udies, 20–50% o he pa ien s wi h PsA appea ed o be esponde s o hese he apies [36]. Opposed o hese empi ical he apies, o he i s ime, ou esul s clea ly iden i ied a subse o PsA pa ien s wi h a s ong T cell-media ed pa hology; he e o e by u ilizing co ela ional clus e ing and DFA excellen candida es o an i-T cell he apy can be selec ed based on hei cy okine p o iles. Long, empi ical he apies can be eplaced by op imized combina ion he apies h ough pe sonalized p o-in lamma o y cy okine a ge ing and planned ad anced cos –bene i s a egies. The u iliza ion o he mul iplex cy okine a ay sys em will aid in he diagnosis and he apy design in PsA, and will p o ide an ad anced disease managemen in he u u e. Acknowledgemen s This wo k was unded by he Na ional Ins i u es o Heal h Na ional Cen e o Resea ch Resou ces (g an numbe s NIH 1 P20 RR15577 and NIH 1 P20 RR16478) and he Facul y o Medicine, Uni e si y o Be gen, Be gen, No way. Funding o pay he Open Access publica ion cha ges o his a icle was p o ided by D Michael Cen ola. The au ho s ha e decla ed no con lic s o in e es . Re e ences 1. Benne RM. Pso ia ic a h i is. In: Koopman WJ, ed. A h i is and allied condi ions: a ex book o heuma ology, 14 h edn. Philadelphia, Pennsyl ania: Lippinco Williams & Wilkins, 2001;1345–61. 2. Gladman DD, An oni C, Mease P, Clegg DO, Nash P. Pso ia ic a h i is: epidemiology, clinical ea u es, cou se, and ou come. Ann Rheum Dis 2005;64(Suppl 2):ii14–17. 3. Veale DJ, Ri chlin C, Fi zGe ald O. Immunopa hology o pso iasis and pso ia ic a h i is. Ann Rheum Dis 2005;64(Suppl 2):ii26–9. Rheuma ology Key message U ilizing a b oad spec um bead-based immunoassay and mul i a ia e analysis, we ha e iden i ied unique co ela ions among soluble immune modula o s in pa ien s wi h PsA. In doing so we ha e gene a ed dis inc p o iles o ac o s ha pa allels wi h a ious a ec ed join coun s, he e o e p o iding new insigh s in o he pa hophysiology o PsA. 424 P. Szodo ay e al. a Uni e si y o Deb ecen, Facul y o Medicine, Cen al Lib a y on May 5, 2012h p:// heuma ology.ox o djou nals.o g/Downloaded om 4. Mizu ani H, Ohmo o Y, Mizu ani T, Mu a a M, Shimizu M. Role o inc eased p oduc ion o monocy es TNF-alpha, IL-1be a and IL-6 in pso iasis: ela ion o ocal in ec ion, disease ac i i y and esponses o ea men s. J De ma ol Sci 1997;14:145–53. 5. Hohle T, K uge A, Schneide PM e al. A TNF-alpha p omo e polymo phism is associa ed wi h ju enile onse pso iasis and pso ia ic a h i is. J In es De ma ol 1997;109:562–5. 6. Elkayam O, Ya on I, Shi azi I, Ya on M, Caspi D. Se um le els o IL-10, IL-6, IL-1 a, and sIL-2R in pa ien s wi h pso ia ic a h i is. Rheuma ol In 2000;19:101–5. 7. Spada o A, Rinaldi T, Riccie i V, Valesini G, Tacca i E. In e leukin 13 in syno ial luid and se um o pa ien s wi h pso ia ic a h i is. Ann Rheum Dis 2002;61:174–6. 8. Rooney T, Mu phy E, Beni o M e al. Syno ial issue in e leukin-18 exp ession and he esponse o ea men in pa ien s wi h in lamma- o y a h i is. Ann Rheum Dis 2004;63:1393–8. 9. Balla a S, Taylo PC, Reusch P e al. Raised se um ascula endo helial g ow h ac o le els a e associa ed wi h des uc i e change in in lamma o y a h i is. A h i is Rheum 2001;44:2055–64. 10. Win e ield LS, Men e A, Go don K, Go lieb A. Pso iasis ea men : cu en and eme ging di ec ed he apies. Ann Rheum Dis 2005;64(Suppl 2):ii87–90. 11. Zagu y D, Gallo RC. An i-cy okine Ab immune he apy: p esen s a us and pe spec i es. D ug Disco Today 2004;9:72–81. 12. Zwe ina J, Redlich K, Sche G, Smolen JS. Pa hogenesis o heuma oid a h i is: a ge ing cy okines. Ann N Y Acad Sci 2005; 1051:716–29. 13. Fi es ein GS. Pa hogenesis o heuma oid a h i is: how ea ly is ea ly? A h i is Res The 2005;7:157–9. 14. Yada D, Sa e nick N. Cy okines and au oimmuni y: edundancy de ines hei complex na u e. Cu Opin Immunol 2003;15:697–703. 15. Hi chon CA, Alex P, E dile LB e al. A dis inc mul icy okine p o ile is associa ed wi h an i-cyclical ci ullina ed pep ide an ibodies in pa ien s wi h ea ly un ea ed in lamma o y a h i is. J Rheuma ol 2004;31:2336–46. 16. Szodo ay P, Alex P, B un JG, Cen ola M, Jonsson R. Ci cula ing cy okines in p ima y sjog en’s synd ome de e mined by a mul iplex cy okine a ay sys em. Scand J Immunol 2004;59:592–9. 17. Dozmo o IM, Cen ola M, Knowl on N, Tang Y. Mobile classi ica ion in mic oa ay expe imen s. Scand J Immunol 2005;62(Suppl 1):84–91. 18. Ja is JN, Dozmo o I, Jiang K e al. No el app oaches o gene exp ession analysis o ac i e polya icula ju enile heuma oid a h i is. A h i is Res The 2004;6:R15–32. 19. Szodo ay P, Alex P, Jonsson MV e al. Dis inc p o iles o Sjog en’s synd ome pa ien s wi h ec opic sali a y gland ge minal cen e s e ealed by se um cy okines and BAFF. Clin Immunol 2005. 20. Moll JM, W igh V. Pso ia ic a h i is. Semin A h i is Rheum 1973;3:55–78. 21. F ed iksson T, Pe e sson U. Se e e pso iasis–o al he apy wi h a new e inoid. De ma ologica 1978;157:238–44. 22. Szodo ay P, Alex P, Dandapani V e al. Apop o ic e ec o i uximab on pe iphe al blood B cells in heuma oid a h i is. Scand J Immunol 2004;60:209–18. 23. Go schalk PG, Dunn JR. De e mining he e o o dose es ima es and minimum and maximum accep able concen a ions om assays wi h nonlinea dose- esponse cu es. Compu Me hods P og ams Biomed 2005;80:204–15. 24. Jo gensen ED, Dozmo o I, F ank MB, Cen ola M, Albino AP. Global gene exp ession analysis o human b onchial epi helial cells ea ed wi h obacco condensa es. Cell Cycle 2004;3:1154–68. 25. Dozmo o I, Saban MR, Knowl on N, Cen ola M, Saban R. Connec i e molecula pa hways o expe imen al bladde in lamma- ion. Physiol Genomics 2003;15:209–22. 26. Tassiulas I, Duncan SR, Cen ola M, Theo ilopoulos AN, Boumpas DT. Clonal cha ac e is ics o T cell in il a es in skin and syno ium o pa ien s wi h pso ia ic a h i is. Human Immunology 1999;60:479–91. 27. Biede mann T, Rocken M, Ca ballido JM. TH1 and TH2 lymphocy e de elopmen and egula ion o TH cell-media ed immune esponses o he skin. J In es ig De ma ol Symp P oc 2004;9:5–14. 28. Jacob SE, Nassi i M, Ke del FA, Vincek V. Simul aneous measu emen o mul iple Th1 and Th2 se um cy okines in pso iasis and co ela ion wi h disease se e i y. Media o s In lamm 2003; 12:309–13. 29. Appelbe g R. Mac ophage in lamma o y p o eins MIP-1 and MIP-2 a e in ol ed in T cell-media ed neu ophil ec ui men . J Leukoc Biol 1992;52:303–6. 30. Appelbe g R. In e e on-gamma (IFN-gamma) and mac o- phage in lamma o y p o eins (MIP)-1 and -2 a e in ol ed in he egula ion o he T cell-dependen ch onic pe i oneal neu ophilia o mice in ec ed wi h mycobac e ia. Clin Exp Immunol 1992;89:269–73. 31. Bischo SC, K iege M, B unne T e al. RANTES and ela ed chemokines ac i a e human basophil g anulocy es h ough di e en G p o ein-coupled ecep o s. Eu J Immunol 1993;23:761–7. 32. Fahey TJ,3 d, T acey KJ, Tekamp-Olson P e al. Mac ophage in lamma o y p o ein 1 modula es mac ophage unc ion. J Immunol 1992;148:2764–9. 33. Kane D, Goga y M, O’Lea y J e al. Reduc ion o syno ial sublining laye in lamma ion and p oin lamma o y cy okine exp ession in pso ia ic a h i is ea ed wi h me ho exa e. A h i is Rheum 2004; 50:3286–95. 34. Pa sch G, Wagne E, Leeb BF, B oll H, Dunky A, Smolen JS. T cell de i ed cy okines in pso ia ic a h i is syno ial luids. Ann Rheum Dis 1998;57:691–3. 35. Pa sch G, S eine G, Leeb BF, Dunky A, B oll H, Smolen JS. Highly inc eased le els o umo nec osis ac o -alpha and o he p oin lamma o y cy okines in pso ia ic a h i is syno ial luid. J Rheuma ol 1997;24:518–23. 36. Mease PJ, An oni CE. Pso ia ic a h i is ea men : biological esponse modi ie s. Ann Rheum Dis 2005;64(Suppl 2):ii78–82. Ci cula ing cy okines in pso ia ic a h i is 425 a Uni e si y o Deb ecen, Facul y o Medicine, Cen al Lib a y on May 5, 2012h p:// heuma ology.ox o djou nals.o g/Downloaded om