Rheumatol Int (2007) 27:807–811

DOI 10.1007/s00296-006-0301-6

ORIGINAL ARTICLE

Serum proinXammatory cytokines directing T helper 1 polarization in patients with familial Mediterranean fever

Ismail Simsek · Salih Pay · Aysel Pekel · Ayhan Dinc · Ugur Musabak · Hakan Erdem · Ali Sengul

Received: 22 March 2006 / Accepted: 21 December 2006 / Published online: 17 January 2007© Springer-Verlag 2007

Abstract Th1 type polarization has been implicatedin the pathogenesis of familial Mediterranean fever(FMF). Interleukin-12 (IL-12) and IL-10 are proin-Xammatory cytokines, which play crucial role in Th1and Th2 type immune response, respectively. IL-18has a dual eVect on T cell response: it was recognizedas an IFN-�-inducing factor in T cells; acting in syn-ergy with IL-12, leading to the development of Th1type immune responses. But, in the absence of IL-12,IL-18 can promote the production of Th2 cytokinesand take part in allergic inXammation. The aim of thisstudy is to measure serum levels of IL-10, IL-12, andIL-18 in patients with FMF, and to investigate therelationship of their expressions with FMF attacks.Serum IL-10, IL-12, and IL-18 levels from patientswith FMF were investigated. Thirty-one FMF patientswith attack-free, 24 FMF patients with attack and 20healthy controls were enrolled in the study. The levelsof IL-10, IL-12p70 and IL-18 were measured byELISA. Serum IL-10 levels were not diVerent in FMFpatients with attack and attack-free, and healthy con-trols. Serum IL-12 levels in FMF patients both withattack and attack-free were signiWcantly higher thanhealthy controls (P = 0.002 and P = 0.047, respec-tively). There were no diVerences between FMF

patients with attack and attack-free with regard toserum IL-12 levels. Serum IL-18 levels in FMFpatients with attack and attack-free were signiWcantlyhigher than healthy controls (P < 0.001 for bothgroups). With respect to serum IL-18 levels, no diVer-ence was found between FMF patients with attack andattack-free. Our results suggest that IL-12 and IL-18contribute to the establishment of Th1 polarizationseen in FMF and play a part in its pathogenesis.Detection of increased levels of IL-12 and IL-18 inpatients with inactive disease implies that they seem toassist Th1 activation and subclinical inXammation per-sisting during the attack-free period of the disease.

Introduction

Interleukin-12 (IL-12) is mainly produced by the cellsof monocyte/macrophage lineage including, mono-cytes, macrophages, KupVer cells, mesangial cells, andglial cells in response to stimulation with variousmicroorganisms and as well as their products. Cells ofthe innate immune system such as the polymorphonu-clear leukocytes, mast cells, and keratinocytes have allbeen shown to produce IL-12. Furthermore, dendriticcells and Langerhans cells are the most important anti-gen-presenting cells that produce IL-12 [1–3]. IL-12 isthe main cytokine that regulates Th1 diVerentiationand has a number of important actions that serve topromote cell-mediated immunity. In addition, IL-12and IFN-� have been shown to antagonize Th2 diVer-entiation and the production of IL-4, IL-5, and IL-13[4].

Interleukin-10 is a pleiotropic cytokine, which playsa crucial role in regulating the immune response. It has

I. Simsek · S. Pay · A. Dinc · H. ErdemDivision of Rheumatology, Gulhane Military School of Medicine, Ankara, Turkey

A. Pekel · U. Musabak · A. SengulDivision of Immunology, Gulhane Military School of Medicine, Ankara, Turkey

I. Simsek (&)GATA Romatoloji BD, 06018 Etlik, Ankara, Turkeye-mail: isimsek@gata.edu.tr

123

808 Rheumatol Int (2007) 27:807–811

been shown to be expressed in various subsets of Tcells, macrophages, monocytes, dendritic cells, mastcells, B cells, eosinophils, and keratinocytes [5]. Itsmain biological functions seem to be to limit and termi-nate the inXammatory responses, block proinXamma-tory cytokine production and regulate thediVerentiation and proliferation of several immunecells, including T cells, B cells, natural killer cells, andmast cells [6, 7].

Interleukin-18 is a recently described 18 kDa cyto-kine, derived by enzymatic cleavage of a 23 kDa pre-cursor, pro-IL 18, by at least caspase 1. Pro-IL 18expression is widely distributed and has been demon-strated in cells of the monocyte–macrophage system,dendritic cells, KupVer cells, keratinocytes, articularchondrocytes, synovial Wbroblasts and osteoclasts [8].IL-18, initially recognized as an IFN-�-inducing factorin T cells, acts in synergy with IL-12, leading to thedevelopment of Th1 type immune responses [9]. How-ever, later studies showed that IL-18 in the absence ofIL-12 can promote the production of Th2 cytokinesand takes part in allergic inXammation, suggesting apossible role for IL-18 both in Th1 and Th2 typeimmune responses [10]. Furthermore, involvement ofIL-18 in the pathogenesis of diseases characterized byTh1 type polarization such as RA, adult onset Still’sdisease, graft-versus-host disease and Behçet’s diseasehas been demonstrated [11–13, 14].

Familial Mediterranean fever (FMF) is character-ized by periodic attacks of fever accompanied byinXammation localized to serosal surfaces, synovium,or skin. The inheritance of the disease shows an auto-somal recessive pattern and the mutations in theMEFV gene has been shown to be the cause of the dis-order [15, 16]. Recently, it has been suggested thatthere seems to be an ongoing subclinical inXammationin FMF patients during an attack-free period [17–23,24]. Moreover, several studies have demonstrated aTh1 polarization in the inXammatory response of FMF[25–27]. Our group has previously demonstrated that Tcells are abnormally activated in patients with FMF,either when symptomatic or between the attacks, andTh1 diVerentiation is the hallmark of inXammations inthose patients [28]. The cause of Th1 polarization andits contribution to the pathogenesis and the clinical fea-tures of FMF have not been deWned yet. As a continua-tion of our previous work, we investigated the serumlevels of IL-10, IL-12, and IL-18, which are producedby antigen presenting cells (APC) and cells of mono-cyte/macrophage lineage and are suggested to beinvolved in the polarization of immune response inFMF patients and compared them with those of thehealthy controls.

Materials and methods

Patients

In this study, serum IL-10, IL-12, and IL-18 levels ofpatients with FMF were investigated. Thirty-one FMFpatients with attack-free, 24 FMF patients with attackand 20 healthy controls were enrolled in the study.Diagnosis of patients with FMF was made using crite-ria suggested by Livneh et al. [29]. Patients with FMFwere evaluated during both attack and attack-free peri-ods of the disease. Serum samples were obtained frompatients for diagnostic purposes at the time of admis-sion to outpatient clinic or clinic of rheumatology divi-sion. Informed consent of the patients was obtained,and all procedures were performed in accordance withthe principles of the Declaration of Helsinki. Ten milli-litres of venous blood was drawn and centrifuged at3,000 rpm for 30 min for the measurement of IL-10, IL-12 and IL-18 levels. The specimens were stored at¡70°C until analysis.

Measurements of IL-10, IL-12, IL-18

The levels of IL-10, IL-12p70 and IL-18, were mea-sured by human IL-10 ELISA kit and human IL-12p70ELISA kit (BioSource International, Inc., CA, USA),and RayBio Human IL-18 ELISA kit (Ray Biotech,Inc., Norcroos, GA, USA), respectively. Absorbancereadings were carried out on an EL 312 BiokineticsReader. Concentrations of unknown samples weredetermined from a curve obtained with the standards.

Statistical analysis

All of the statistical analysis was performed by usingSPSS (SPSS 10.0 FW, SPSS Inc., Chicago, USA) statis-tical package. Descriptive statistics were presented asarithmetic mean § standard deviation notation. Forthe tests of normality, we used Kolmogorov–Smirnovtest. For multiple groups, we used Kruskal–Wallis,Oneway ANOVA or Chi-square tests. Mann–Whitneyor Bonferroni tests were used as post hoc tests of Krus-kal–Wallis or Oneway ANOVA tests, respectively. Pvalues less than or equal to 0.05 were evaluated as sta-tistically signiWcant.

Results

Comparisons of age or sex in patient and control groupswere presented in Table 1. No diVerence was foundbetween the patients and controls with respect to their

123

Rheumatol Int (2007) 27:807–811 809

ages and sexes. No diVerence was detected among thestudy subgroups with respect to ESR. Serum CRP levelsin FMF patients with attack and attack-free were signiW-cantly higher than those of the healthy controls(P = 0.004 and P = 0.027, respectively). There were nodiVerence between FMF patients with attack andpatients without attack with regard to their serum CRPlevels (Table 2). Serum levels of IL-10, IL-12, and IL-18are shown in Table 2. Serum IL-10 levels were not foundto be diVerent among FMF patients with attack andattack-free and healthy controls. While a trend towardsdecreased serum IL-10 levels were noted in FMFpatients, it did not reach statistical signiWcance. SerumIL-12 levels in FMF patients with attack and attack-freewere signiWcantly higher than those of the healthy con-trols (P = 0.002 and P = 0.047, respectively). There wereno diVerences between FMF patients with attack andpatients without attack with regard to their serum IL-12levels. Serum IL-18 levels in FMF patients with attackand attack-free were signiWcantly higher than healthycontrols (P < 0.001 for both groups). However, no diVer-ence was noted between FMF patients with attack andattack-free regarding serum IL-18 levels. The ratio ofTh1/Th2 cytokines (IL-12/IL-10 and IL-18/IL-10) wasfound to be signiWcantly increased in FMF patients bothduring attack and attack-free periods as compared tohealthy controls (Table 2).

Discussion

Familial Mediterranean fever is an autosomal recessivedisorder characterized by acute attacks of fever accom-panied by serosal, synovial, or cutaneous inXammation.Although mutations in the MEFV gene has beenshown to be responsible for the disease, the precisepathophysiology of the disease seems to be more com-plex than previously appreciated [15, 16]. The MEFVgene encodes a protein, which is called pyrin or maren-ostrin that is expressed mainly in neutrophils, as well asin eosinophils and monocytes. In contrast to the com-mon expectation, neutrophils of the patients with FMFdo not diVer from those of the healthy controls bymeans of structure and function [30, 31]. It has beenrevealed that pyrin plays a regulatory role in theinXammatory pathway by leading inactivation of theactivated neutrophils [32, 33]. It has recently beendemonstrated that inXammation in FMF shows a Th1polarization. Aypar et al. [25] have found that the per-centage of IL-4 positive T cells was not signiWcantlydiVerent between the groups. However, the percentageof IFN-� positive T cells in FMF patients experiencingan attack was higher than asymptomatic FMF patients,and controls. Their study further disclosed anincreased IFN-� production in heterozygotes as com-pared to controls [25]. Detection of decreased preva-lence of atopic diseases, which are mainly mediated byTh2-type immune response, in FMF patients, indirectlyprovides evidence for the Th1 polarization in thesepatients [26]. Supporting the previous studies, Köklüet al. [27] found increased plasma IFN-� levels in FMFpatients both with and without attack as compared tohealthy controls. We investigated activation markersfor T lymphocytes in FMF patients in our previousstudy and showed that both sIL2R levels and percent-age of CD4 + CD69 T cells were signiWcantly higherboth in patients experiencing an attack and patients

Table 1 Demographic characteristic of patients with FMF andhealthy controls (mean § SD)

FMF familial Mediterranean fever, HC healthy controls, M male,F female* Comparisons of age or sex were done by using Oneway ANO-VA test or Chi-square test, respectively

FMF (attack-free) FMF (attack) HC P*

n 31 24 20Age 21 § 2 22 § 2 23 § 3 0.160Sex (M/F) 25/6 18/6 15/5 0.846

Table 2 Comparisons of serum IL-10, IL-12 and IL-18 levels in patients with FMF and HC (mean § SD)

The diVerences between two groups were evaluated by Mann–Whitney test. a vs. c P = 0.027, b vs. c P = 0.004, d vs. f P = 0.047, e vs.f P = 0.002, g vs. j P < 0.001, h vs. j P < 0.001, k vs. n P = 0.016, m vs. n P = 0.012, p vs. r P < 0.001, q vs. r P < 0.001

FMF familial Mediterranean fever, HC healthy controls, ESR erythrocyte sedimentation rate, CRP C-reactive protein, IL interleukin* Multiple comparisons were done by using Kruskal–Wallis test or Oneway ANOVA

FMF (attack-free) FMF (attack) HC P*

n 31 24 20ESR (mm/h) 24 § 22 27 § 22 20 § 15 0.539CRP (mg/l) 40 § 45a 83 § 73b 10 § 8c 0.006IL-10 (pg/ml) 4.6 § 0.8 4.9 § 0.7 5.1 § 0.4 0.072IL-12 (pg/ml) 1.0 § 0.6d 1.2 § 0.6e 0.6 § 0.1f 0.003IL-18 (pg/ml) 369 § 169 g 408 § 173 h 49 § 56 j <0.001IL-12/IL-10 0.25 § 0.17k 0.26 § 0.12m 0.13 § 0.04n 0.006IL-18/IL-10 86 § 49p 85 § 40q 9 § 10r <0.001

123

810 Rheumatol Int (2007) 27:807–811

without an attack as compared with those of thehealthy controls. However, percentages ofCD3 + HLADR + T cells to CD8 + CD25 + T cells toCD8 + CD69 + T cells were found to be signiWcantlyincreased only in those patients experiencing an attack.Taken together, these results suggest that T cell systemis activated abnormally in FMF patients and this acti-vation continues between the attacks [28]. Although allthese data provide evidence for the dominated Th1response in FMF patients, the factors responsible forthe Th1 type response in FMF have not been deWnedyet. Currently, APC and cytokines produced by thesecells have been known to play a key role in the polari-zation of the immune response. Cytokines secreted byAPC and cells of monocyte/macrophage lineage suchas IL-10, IL-12, and IL-18, play a critical role in thedevelopmental regulation of naïve T cells into eitherTh1 or Th2 cells. Considering the established roles ofthese cytokines in the polarization of immuneresponse, we investigated the serum levels of IL-10, IL-12, and IL-18 in FMF patients both having an attackand without an attack and compared with those of thehealthy controls.

In this study, we demonstrated that FMF patients,both in attack-free periods and with acute attacks, hadhigher serum IL-12 and IL-18 levels than healthy con-trols, while no such diVerence was detected withrespect to serum IL-10 levels. These results can betaken as an evidence of the possible roles of these cyto-kines in the inXammatory process and Th1 polarizationseen in FMF patients. However, factors inducing thesecretion of these cytokines have not been clariWed yet.Recently, in addition to its expression in neutrophils,MEFV expression has been observed in monocytes andeosinophils [34]. Monocytes are the member of innateimmune system and involved in the generation ofimmune response by producing several proinXamma-tory cytokines including, IL-12, IL-18, and TNF-�. Wesuggest that, in patients with FMF, cells of monocyte–macrophage lineage can be spontaneously stimulated,which may lead to proinXammatory cytokine activationresulting in Th1 type immune response.

Interleukin-12 is involved in the induction of severalcytokines from T and NK cells such as IFN-� and playsa key role in the regulation of adaptive immuneresponse toward Th1 [3]. IL-18 was initially recognizedas an IFN-�-inducing factor in T cells that acts in syn-ergy with IL-12, leading to the development of Th1type immune responses [9]. Besides its inXuential func-tions both in innate and acquired immunity, IL-18 con-tributes to the pathogenesis of chronic inXammatorydiseases such as RA, by directly activating monocytesand macrophages to produce IL-1� and TNF-� [11]. In

vitro stimulation of monocytes with Th1 type cytokines(IFN-�) or proinXammatory cytokines (TNF-� and IL-1�) has been shown to result in increased MEFVexpression, whereas treatment with the Th2 type cyto-kines (IL-4, IL-10, and TGF-�) reduced the expressionof MEFV levels [34]. Thus, it is quite possible thatMEFV gene may function as a negative feedback loopfor Th1 cytokines or proinXammatory mediators andthe molecular mechanisms underlying FMF are due toan imbalance in cytokine/MEFV expression levels [22].It is well documented that MEFV gene expression canbe substantially decreased in FMF patients as a resultof mutations [35]. Considering the proposed negativeregulatory role of MEFV gene, we suggest thatincreased expression of proinXammatory cytokines andTh1 type polarization in FMF patients might be due toinsuYcient suppression of monocytes/macrophages.

In conclusion, our results suggest that both IL-12and IL-18, which are mainly produced by APC or cellsof monocyte/macrophage lineage, involved in the cyto-kine network of the inXammatory cascade of FMF.Taken together, our study provides further evidencefor the presence of Th1 mediated inXammatoryresponse in these patients. Although mechanismsunderlying the increased levels of these cytokines inFMF patients remain to be established, we may specu-late that MEFV gene operates in a negative feedbackloop that is speciWc for Th1 and proinXammatory medi-ator activation of monocytes/macrophages.

References

1. Kang BY, Kim E, Kim TS (2005) Regulatory mechanisms andtheir therapeutic implications of interleukin-12 production inimmune cells. Cell Signal 17:665–673

2. Gately MK, Renzetti LM, Magram J, Stern AS, Adorini L,Gubler U, Presky DH (1998) The interleukin-12/interleukin-12-receptor system: role in normal and pathologic immune re-sponses. Annu Rev Immunol 16:495–521

3. Watford WT, Moriguchi M, Morinobu A, O’Shea JJ (2003)The biology of IL-12: coordinating innate and adaptive im-mune responses. Cytokine Growth Factor Rev 14(5):361–368

4. Hasko G, Szabo C (1999) IL-12 as a therapeutic target forpharmacological modulation in immune-mediated andinXammatory diseases: regulation of T helper 1/T helper 2 re-sponses. Br J Pharmacol 127:1295–1304

5. Williams LM, Ricchetti G, Sarma U, Smallie T, Foxwell BM(2004) Interleukin-10 suppression of myeloid cell activa-tion—a continuing puzzle. Immunology 113:281–292

6. Asadullah K, Sterry W, Volk HD (2003) Interleukin-10 ther-apy—review of a new approach. Pharmacol Rev 55:241–269

7. Moore KW, de Waal Malefyt R, CoVman RL, O’Garra A(2001) Interleukin-10 and the interleukin-10 receptor. AnnuRev Immunol 19:683–765

8. Liew FY, McInnes IB (2002) Role of interleukin 18 in inXam-matory response. Ann Rheum Dis 61:100–102

123

Rheumatol Int (2007) 27:807–811 811

9. Yamamura M, Kawashima M, Taniai M, Yamauchi H, Tan-imoto T, Kurimoto M, Morita Y, Ohmoto Y, Makino H(2001) Ãnterferon-gamma-inducing activity of interleukin-18in the joint with rheumatoid arthritis. Arthritis Rheum44:275–285

10. Nakanishi K, Yoshimoto T, Tsutsui H, Okamura H (2001)Interleukin-18 regulates both Th1 and Th2 responses. AnnuRev Immunol 19:423–474

11. Dai SM, Matsuno H, Nakamura H, Nishioka K, Yudoh K(2004) Interleukin-18 enhances monocyte tumor necrosis fac-tor alpha and, interleukin-1 beta production induced by con-tact with T lymphocytes: implications in rheumatoid arthritis.Arthritis Rheum 50:432–443

12. Fujimori Y, Takatsuka H, Takemoto Y, Hara H, Okamura H,Nakanishi K, Kakishita E (2000) Elevated interleukin (IL)-18levels during acute graft-versus-host disease after allogeneicbone marrow transplantation. Br J Haematol 109:652–657

13. Kawashima M, Yamamura M, Taniai M, Yamauchi H, Tan-imoto T, Kurimoto M, Miyawaki S, Amano T, Takeuchi T,Makino H (2001) Levels of interleukin-18 and its bindinginhibitors in the blood circulation of patients with adult-onsetStill’s disease. Arthritis Rheum 44:550–560

14. Musabak U, Pay S, Erdem H, Simsek I, Pekel A, Dinc A, Sen-gul A (2005) Serum interleukin-18 levels in patients with Be-hçet’s disease: is its expression associated with disease activityor clinical presentations? Rheumatol Int (online)

15. The International FMF Consortium (1997) Ancient missensemutations in a new member of the RoRet gene family arelikely to cause familial Mediterranean fever. Cell 90:797–807

16. The French FMF Consortium (1997) A candidate gene forfamilial Mediterranean fever. Nat Genet 17:25–31

17. Kiraz S, Ertenli I, Arici M, Calguneri M, Haznedaroglu I, Ce-lik I et al (1998) EVects of colchicine on inXammatory cyto-kines and selectins in familial Mediterranean fever. Clin ExpRheumatol 16:721–724

18. Gang N, Drenth JP, Langevitz P, Zemer D, Brezniak N, PrasM et al (1999) Activation of the cytokine network in familialMediterranean fever. J Rheumatol 26:890–897

19. Tunca M, Kirkali G, Soyturk M, Akar S, Pepys MB, HawkinsPN (1999) Acute phase response and evolution of familialMediterranean fever. Lancet 353:1415

20. Direskeneli H, Ozdogan H, Korkmaz C, Akoglu T, Yazici H(1999) Serum soluble intercellular adhesion molecule 1 andinterleukin 8 levels in familial Mediterranean fever. J Rheu-matol 26:1983–1986

21. Korkmaz C, Ozdogan H, Kasapcopur O, Yazici H (2002)Acute phase response in familial Mediterranean fever. AnnRheum Dis 61:79–81

22. Notarnicola C, Didelot MN, Seguret F, Demaille J, Touitou I(2002) Enhanced cytokine mRNA levels in attack-free patientswith familial Mediterranean fever. Genes Immun 3:43–45

23. Bagci S, Toy B, Tuzun A, Ates Y, Aslan M, Inal A et al (2004)Continuity of cytokine activation in patients with familialMediterranean fever. Clin Rheumatol 23:333–337

24. Haznedaroglu S, Ozturk, Sancak B, Goker B, Onat AM, Bu-kan N et al (2005) Serum interleukin 17 and interleukin 18levels in familial Mediterranean fever. Clin Exp Rheumatol23(4 Suppl 38):S77–S80

25. Aypar E, Özen S, Okur H, Kutluk T, Besbas N, BakkalofluA (2003) Th1 Polarization in familial Mediterranean fever. JRheumatol 30:2011–2013

26. Sackesen C, Bakkaloglu A, Sekerel BE, ÖzaltÂn F, Besbas N,YÂlmaz E et al (2004) Decreased prevalence of atopy in pae-diatric patients with familial Mediterranean fever. AnnRheum Dis 63:187–190

27. Köklü S, Öztürk MA, Balc M, Yüksel O, Ertenli Ã, Kiraz S(2005) Interferon-gamma levels in familial Mediterranean fe-ver joint bone. Spine 72:38–40

28. Musabak U, Sengül A, Oktenli C, Pay S, Yesilova Z, Kenar Let al (2004) Does immun activation continue during an at-tack-free period in familial Mediterranean disease? Clin ExpImmunol 138:526–533

29. Livneh A, Langevitz P, Zemer D, Zaks N, Kees S, Lidar Tet al (1997) Criteria for the diagnosis of familial Mediterra-nean fever. Arthritis Rheum 40:1879–1885

30. Territo MC, Peters RS, Cline MJ (1976) Leukocyte func-tion in familial Mediterranean fever. Am J Hematol 1:307–311

31. Mege JL, Dilsen N, Sanguedolce V, Gul A, Bongrand P,Roux H et al (1993) Overproduction of monocyte derived tu-mor necrosis factor alpha, interleukin (IL) 6, IL-8 and in-creased neutrophil superoxide generation in Behcet’sdisease. A comparative study with familial Mediterranean fe-ver and healthy subjects. J Rheumatol 20:1544–1549

32. Baykal Y, Saglam K, Yilmaz MI, Taslipinar A, Akinci SB InalA (2003) Serum sIL-2r, IL-6, IL-10 and TNF-alpha level infamilial Mediterranean fever patients. Clin Rheumatol 22:99–101

33. Oktem S, Yavuzsen TU, Sengul B, Akhunlar H, Akar S, Tun-ca M (2004) Levels of interleukin-6 (IL-6) and its solublereceptor (sIL-6R) in familial Mediterranean fever (FMF) pa-tients and their Wrst degree relatives. Clin Exp Rheumatol22(4 Suppl 34):S34–S36

34. Centola M, Wood G, Frucht DM, Galon J, Aringer M, FarrellC, Kingma DW et al (2000) The gene for familial Mediterra-nean fever, MEFV, is expressed in early leukocyte develop-ment and is regulated in response to inXammatory mediators.Blood 95:3223–3231

35. Notarnicola C, Didelot MN, Kone-Paut I, Seguret F, Dema-ille J, Touitou I (2002) Reduced MEFV messenger RNAexpression in patients with familial Mediterranean fever.Arthritis Rheum 46:2785–2793

123