Id

SNa

b

c

d

a

ARRA

KAPJT

I

s(t

m

CT

BS

0h

Immunobiology 218 (2013) 285– 291

Contents lists available at SciVerse ScienceDirect

Immunobiology

jou rna l h om epage: www.elsev ier .com/ locate / imbio

L-1� and TSH disturb thyroid epithelium integrity in autoimmune thyroidiseases

andra A. Rebuffata,c,1, Maha Kammoun-Krichenb,c, Ilhem Charfeddined, Hammadi Ayadib,c,oura Bougacha-Elleuchb,c, Sylvie Peraldi-Rouxa,c,∗

CNRS FRE 3400, CPID, Montpellier, FranceUnité Cibles pour Diagnostic et Thérapie, Centre de Biotechnologie de Sfax, TunisiaLaboratoire International Associé LIA-135, CNRS, FranceService ORL, CHU Habib Bourguiba, Sfax, Tunisia

r t i c l e i n f o

rticle history:eceived 17 November 2011eceived in revised form 2 May 2012ccepted 16 May 2012

eywords:utoimmune thyroid diseasesro-inflammatory cytokinesunction complexight junction

a b s t r a c t

Pro-inflammatory cytokines such as IL-1� and TNF� are known to affect thyroid function. They stimulateIL-6 secretion and modify epithelium integrity by altering junction proteins. To study the role of cytokineson thyroid epithelia tightness in autoimmune thyroid diseases (AITD), we analyzed the expression profilesof junction proteins (ZO-1, Claudin, JAM-A) and cytokines in human thyroid slices and also investigatedthe effect of IL-1� on the epithelium integrity in primary cultures of human thyrocytes. Junction proteinsexpression (ZO-1, Claudin, JAM-A) has been analyzed by immunohistochemistry on thyroid slices andby Western blot on membrane proteins extracted from thyrocytes of patients suffering from Gravesand Hashimoto diseases. The high expression of junction proteins we found on Graves’ disease thyroidslices as well as in cell membrane extracts acknowledges the tightness of thyroid follicular cells in thisAITD. In contrast, the reduced expression of JAM and ZO-1 in thyroid cells from patients suffering fromHashimoto thyroiditis is in agreement with the loss of thyroid follicular cell integrity that occurs in thispathology. Concerning the effects on epithelium integrity of TSH and of the pro-inflammatory cytokineIL-1� in primary cultures of human thyroid cells, TSH appeared able to modify JAM-A localization butwithout any change in the expression levels of JAM-A, Claudin and ZO-1. Inversely, IL-1� provoked a

decrease in the expression of- and a redistribution of both, Claudin and ZO-1 without modifying theexpression and sub-cellular distribution patterns of JAM-A in thyroid cells. These results demonstrate (i)that Hashimoto’s- and Graves’ diseases display different junction proteins expression patterns with a lossof epithelium integrity in the former and (ii) that IL-1� modifies thyroid epithelial tightness of humanthyrocytes by altering the expression and localization of junction proteins. Therefore, IL-1� could play arole in the pathogenesis of thyroid autoimmunity.

ntroduction

Thyroid auto-immunity can be clinically divided into two major

ubtypes, with over functioning (hyperthyroidism: Graves’ diseaseGD)) or under functioning (hypothyroidism: Hashimoto thyroidi-is (HT)) of the thyroid gland. In both GD and HT, auto-antibodies

Abbreviations: GD, Graves’ disease; HT, Hashimoto thyroiditis; AITD, autoim-une thyroid diseases; TJ, tight junctions.∗ Corresponding author at: CNRS FRE 3400, CPID, Faculté de Pharmacie, 15 avenueharles Flahault, BP 14491, 34093 Montpellier Cedex 5, France.el.: +33 4 11 75 95 02; fax: +33 4 11 75 95 47.

E-mail address: sylvie.roux@univ-montp1.fr (S. Peraldi-Roux).1 Present addresses: Diabetes and Obesity Laboratory, Institut d’Investigationsiomediques August Pi I Sunyer (IDIBAPS), Hospital Clinic de Barcelona, Barcelona,pain. CIBERDEM, CIBER Diabetes y Enfermedades Metabolicas Asociadas, Spain.

171-2985/$ – see front matter © 2012 Elsevier GmbH. All rights reserved.ttp://dx.doi.org/10.1016/j.imbio.2012.05.016

© 2012 Elsevier GmbH. All rights reserved.

(aAbs) specific for thyroid auto-antigens, such as thyroglobulin (Tg),thyroperoxydase (TPO) and/or the TSH receptor (TSHr), have beencharacterized. Besides their function as AITD markers, in vitro stud-ies have demonstrated that in the two pathologies they operate aspathogenic factors via CDC (Chiovato et al. 1993; Parkes et al. 1994)and ADCC (Guo et al. 1997; Metcalfe et al. 1997; Rebuffat et al. 2008;Rodien et al. 1996) mechanisms. However, if it is easy to understand(i) that in vivo the accessibility to immune cells of TSHr, located atthe basolateral membrane of the thyrocytes, makes it possible foranti-TSHr aAbs to be produced and (ii) the role of TSHr aAbs in thy-roid cell destruction through cytotoxic mechanisms mediated byeffector cells and/or complement activation, the process should be

more complex for TPO or thyroglobuline, present in the follicularlumen. Indeed, TPO is a transmembrane molecule located at theapical pole of thyroid epithelial cell in contact with the lumen andnon-accessible to immuno-competent cells, due to the tightness of

2 nobio

tomacTaolt

ciltmseObg1eiiTwmnT

aeda1acatacacFlm1tic

ooicsgpadiisrv

86 S.A. Rebuffat et al. / Immu

hyroid epithelia in non-pathological conditions. The mechanismf TPO presentation to the immune system necessary for the for-ation of anti-TPO aAbs, as well as the problems of how anti-TPO

Abs could be in contact with the TPO molecule and how effectorells exert their cytotoxic activity, remain to be clearly addressed.o explain the presence of anti-TPO aAbs and their vivo cytotoxicctivity in vivo, one hypothesis could be based on the alterationf follicular tightness through modifications in the expression andocalization of junction proteins, enabling immuno-competent cellso join the apical pole of the thyrocyte.

The junctional complex between thyroid follicular cells is partlyomposed of tight junctions (TJ) and adherens junctions (AJ), mak-ng two belt-like structures running along the luminal pole andimiting paracellular permeability. Tight junctions are characteris-ically located at the apicolateral borders of adjacent epithelial cells

aintaining the epithelium barrier function. They are composed ofeveral proteic components, such as the Claudin proteins, a keylement of TJ, Occludin and junctional adhesion molecule (JAM).ccludin has been localized to the TJ strand with Claudin. JAM haseen identified as a component of TJ necessary to allow transmi-ration of monocytes through paracellular space (Del Maschio et al.999). As concerns other epithelial linings, a tight barrier betweenxtracellular compartments, the lumen and extra-follicular space,s critical to normal thyroid function, because it promotes cell polar-ty and the establishment of trans-epithelial solute, e.g. iodine andG, gradients. Conversely, destruction of the paracellular barrierould challenge thyroid function and, in the context of autoim-unity, might facilitate the exposure to the immune system of

ormally secluded auto-antigens: Tg in the follicular lumen andPO at the apical plasma membrane.

Cytokines are known to modify epithelium integrity via alter-tion of junction proteins expression in various epithelia (Ewertt al. 2010; Gretzer et al. 2000; Shaw et al. 2001). During theevelopment of AITD, environmental factors and immune system-nd target tissue disorders are simultaneously involved (Weetman994). The infiltration of thyroid by immune T and B cells, a char-cteristic of AIDT, allows the local release of pro-inflammatoryytokines. In addition, cytokines such as the IL-1, IL-6 and IL-8, arelso produced by thyroid in AITD and directly affect thyroid func-ion, perpetuating inflammation (Gianoukakis et al. 2006). Indeed,

strong negative regulation of TJ by IL-1 has been described inultured thyroid cells by decreases in trans-epithelial resistance,n increase in paracellular flow and the rearrangement of junctionomplex (Nilsson et al. 1998). Interferon (IFN) and Tumor Necrosisactor (TNF), known to have an effect on intestinal and renal epithe-ium, are ineffective on paracellular permeability and functional

orphology of thyroid cells (Madara 1989; Madara and Stafford989; Mullin et al. 1992; Nilsson et al. 1998). IL-6, produced byhyrocytes, does not affect junctions between thyroid cells unliken cancerous breast cells where the cytokine is able to decreaseell–cell association (Tamm et al. 1989; Weetman et al. 1990).

Until now, the physiopathological role of cytokines in the devel-pment of AITD is not clearly established. It could involve a lossf thyroid epithelial barrier function, promoting the access formmune cells to intra-follicular antigens. To explore this issue, weompared the expression of junction proteins in the two majorubtypes of AITD, Hashimoto and Graves’ diseases. We also investi-ated the effect of recombinant IL-1�, on the expression of junctionroteins in cultured thyroid cells. Our data bring evidence that HTnd GD display marked differences in junction proteins expression,etermined by low JAM-A and ZO-1 levels in the former. Interest-

ngly, we could also show that the pro-inflammatory cytokine IL-1�

s able to decrease Claudin and ZO-1 expression and modify theirub-cellar distribution. These data could be of physiopathologicalelevance; they argue for a role of IL-1� in the pathogenesis of AITDia alterations in follicular tightness.

logy 218 (2013) 285– 291

Materials and methods

Cytokines and antibodies

Human recombinant cytokines, IL-1� (50 ng/mL) were pur-chased from BD Biosciences (Bedford, MA). Rabbit anti-humanzonula occudens (ZO)-1, anti-Claudin-1 and anti-JAM-A were fromZymed Laboratories (San Francisco, US). Rabbit anti-human IL-1,IL-6R, IL-1R1 and Goat anti-human IL-6 were from Santa CruzBiotechnology (Santa Cruz, CA, US). Peroxydase-congugated sec-ondary antibody (anti-rabbit and anti-mouse PO) were from Sigma,(Saint Louis, US), Texas red (TR) and FITC-conjugated anti-rabbitand anti-goat antibody from Vector laboratories (Peterborough,UK).

Tissue samples and clinical assessment

Specimens of thyroid tissues were obtained from patientsundergoing surgery. GD was diagnosed on the basis of clinicaland laboratory evidence of hyperthyroidism requiring treatment,palpable diffuse goiter, high thyroid hormonal rates, positiveanti-TSHR, anti-thyroid peroxidase, and/or anti-thyroglobulin anti-bodies. Before surgical ablation all GD patients included inthis work were treated with anti-thyroid drugs (carbimazol,benzylthio-uracil or propylthio-uracil) to decrease hormonal thy-roid level and to reach euthyroidism. HT was diagnosed bydocumented clinical and biochemical hypothyroidism requiringthyroid hormone replacement and presence of antibodies tothyroid peroxidase, with/or without antibodies to thyroglobu-lin.

Primary culture of thyroid cells

Tissues from patients suffering from Graves’ (GD) andHashimoto disease (HT) were obtained from Dr. B. Guerrier (Guide Chauliac Hospital, Montpellier, France). Primary cultures havebeen carried out from biopsies of Graves’ or Hashimoto’s thy-roids according to a previously established protocol (Rebuffat et al.2008). Each patient’s tissue was used to prepare a single pri-mary cell culture as previously described (Estienne et al. 2002).Briefly, the tissue was minced, digested with collagenase (Sigma,St Louis, MO) in HBSS medium (PAA Laboratories, Austria) for1 h at 37 ◦C. The digested tissue was filtered through a sterilegauze, washed with Coon modified Ham’s F-12 medium (Sigma)containing 2.6 g/L sodium carbonate (Sigma), 50 U/mL penicillin,50 �g/mL streptomycin, and 2 mmol/L glutamine (Life technol-ogy, France) and supplemented with 5% fetal calf serum (FCS)(PAA Laboratories). In order to get closer to a more “physiologi-cal” thyroid environment, thyrocytes were seeded in this culturemedium supplemented with a mixture of 5 or 6 hormones:Insulin (10 �g/mL), Transferrin (5 ng/mL), Glycyl-l-Histidyl-l-Lysinacetate (100 ng/mL), Somatostatin/SRIF (10 ng/mL), Hydrocorti-son (10 nM) to compose mixture 5H, and with TSH (1 mU/mL,(Sigma)) to complete mixture 6H. The cells were incubatedat 37 ◦C in a humidified atmosphere containing 5% CO2. After24 h, culture supernatant was removed by aspiration. Proteinextraction was performed either directly on these cells or afteran additional 24 and 48 h exposure to fresh medium supple-mented or not with IL-1� cytokine (50 ng/mL). Experiments wereperformed at least three times for each experimental condi-tion.

Immunofluorescence

For immunofluorescence studies on cultured cells, thyrocyteswere grown directly on a LabTech chamber slide system (Sigma)

nobio

ac3wsdoftcmpc

wsa

M

fwf4beof4eu

D

GpwTaCmaiwB

t5ttdnfa1fpmtcII

S.A. Rebuffat et al. / Immu

nd used at 75% confluence. Cells were washed with PBS (pH 7.4)ontaining CaCl2 and MgCl2, fixed with 3% paraformaldehyde for0 min and quenched with 50 mM NH4Cl for 10 min. After twoashings with PBS, slices were saturated with a 2% BSA-gelatine

olution and then incubated with either anti-human zonula occu-ens (ZO)-1-, anti-Claudin-1 or anti-JAM-antibodies (diluted 1:50)vernight at 4 ◦C. After three washings, cells were then incubatedor 1 h with FITC-conjugated anti-rabbit antibody (Vector labora-ories). After three additional washings, cells were covered withitifluor (Citifluor, UK) and observed with an upright fluorescenceicroscope using the facilities of RIO imagery platform (Mont-

ellier, France). The negative control was performed using theonjugated antibody alone.

The expression level of junction proteins (ZO-1, Claudin, JAM-A)as also investigated on thyroid slices (5 �m thickness) using the

ame protocol, as well as the expression of cytokines (IL-1�, IL-6)nd their receptors.

embrane protein extraction

Thyroid cells were grown in T75 tissue culture flasks (Corning)or membrane protein extraction. Cells were washed three timesith D-PBS, scraped at 4 ◦C and membrane protein extraction per-

ormed as follows: after centrifugation at 1000 rpm for 5 min at◦C, membrane proteins were solubilized by adding 500 �L of lysisuffer (50 mM Tris/HCl, pH 7.3, 150 mM NaCl, 5 mM EDTA, 10% glyc-rol, 1% Triton X-100 and a protease inhibitor mixture tablet/10 mLf lysis buffer (Roche Molecular Biochemicals) and subjected to tworeeze–thaw cycles. After centrifugation at 13,000 rpm for 30 min at◦C, the supernatants containing membrane proteins were recov-red and stored at −80 ◦C. Protein concentrations were evaluatedsing the BCA protein assay reagent (Pierce).

ot blot and Western blot analysis

For Dot blot assays, 1 and 3 �g of thyroid protein extract fromD or HT were adsorbed onto a nitrocellulose membrane (Milli-ore) by gravity flow. The membrane was then washed three timesith PBS, saturated with 5% non-fat powdered milk in PBS-0.1%

ween 20 (PBS-T; blocking buffer) for 1 h at room temperaturend then incubated with polyclonal anti-JAM-A, anti-ZO-1 or anti-laudin antibodies. After three washings with PBS-T for 10 min,embranes were probed with peroxydase-congugated secondary

ntibody (anti-rabbit IgG whole molecule) diluted 1:3000 (Sigma)n blocking buffer for 1 h at room temperature. Then, the signal

as detected by chemiluminescence ECL (Amersham Pharmaciaiotech) on a sensitive film.

For Western blotting, approximately 20 �g of membrane pro-eins were mixed with protein buffer (0.2 mM Tris/HCl, pH 6.8,0% glycerol, 1% SDS and 0.1% bromophenol blue). Each native pro-ein sample was loaded into individual wells and electrophoresedhrough 10% SDS-PAGE. Proteins were transferred to polyvinyli-ene fluoride membranes which were then saturated with 5%on-fat powdered milk in D-PBS-Tween 0.1% (blocking buffer)

or 60 min at room temperature. The first antibody: anti-Claudin,nti-ZO-1 or anti-JAM-A Ab was incubated at a concentration of

�g/mL overnight at 4 ◦C. After three washings with D-PBS-Tweenor 10 min at room temperature, membranes were probed witheroxidase-conjugated secondary antibody (anti-rabbit IgG wholeolecule) diluted 1:3000 in blocking buffer for 60 min at room

emperature. After three washings, the signal was detected byhemiluminescence ECL (Amersham BioScience) on a sensitive film.ntensity values were obtained with ImageJ software (Nationalnstitut of Health, Bethesda, MD).

logy 218 (2013) 285– 291 287

Results

Expression of Claudin, ZO-1 and JAM-A in the thyroid of patientswith AITD

Tight junction (TJ) are composed of multi-protein complexesincluding the transmembrane proteins Occludin and Claudin thatexert fence and gate functions, and a set of cytoplasmic proteins(ZO-1, ZO-2, and ZO-3) that link the junction to microfilamentsand regulate its assembly (Tsukita et al. 2008) (Fig. 1A). JAM-A,a small transmembrane immunoglobulin-like adhesion molecule,co-distributes with TJ components. To define if there is a relationbetween AITD and the integrity of thyroid epithelia, we exploredjunction proteins involved in tight junction in thyroid tissue frompatients suffering of AITD. Expression of Claudin, JAM-A and ZO-1 proteins were studied in thyroid tissues from patients sufferingfrom HT and GD. Although Claudin is strongly expressed in thyroidtissues from HT patients, JAM-A and ZO-1 appear to be more weaklyexpressed when compared to tissues from GD patients (Fig. 1B andC).

We then evaluated the integrity of the epithelia by perform-ing immunofluorescence stainings for ZO-1, JAM-A and Claudin onthyroid slices obtained from GD patients; a high expression of thethree junction proteins, ZO-1, JAM-A and Claudin was observed,confirming Western blot analysis data. ZO-1 staining pattern dis-closed a remarkable continuous distribution, mainly at the surfaceof thyrocytes in contact with the lumen. JAM-A and Claudin arelargely expressed on follicular cells and make up a continuous ringaround each cell (Fig. 2A).

Effect of TSH treatment on junction proteins

GD is characterized by glandular hyper-function whereas HTresults into glandular hypo-function. In the former, production ofthyroid-stimulating antibodies (stimulating anti-TSHR Ab) leads tothe production of T3 and T4 hormones exerting a negative feedbackon TSH secretion; in the second case TSH levels increase. Thereforethe differences in the junction proteins expression we observed inthyroid tissues from GD and HD patients, prompted us to evaluatethe possible modulation of junction proteins expression by TSH.Thyroid cells were cultured in the presence or absence of TSH, andClaudin and JAM-A expressions were analyzed by Western blot. Pri-mary cultures of thyrocytes were obtained from four thyroid tissuesissued from GD patients. No significant variation in the expressionof junction proteins, Claudin and JAM-A, could be observed in pri-mary cultures of thyroid cells in the presence of TSH (6H versus5H) (Fig. 3A and B) suggesting that TSH has no effect on the expres-sion level of these proteins. TSH treated thyrocytes appeared as amonolayer of polarized cells connected by typical tight junctions.Our immunofluorescence data show that in the absence of IL-1�,TSH did not change ZO-1 and Claudin localization (Fig. 4a and b), butprovoked a sub-cellular redistribution of JAM-A, characterized by amore punctuated staining at the polygonal cells periphery (Fig. 4c).

Effect of IL-1 ̌ treatment on junction proteins

During AITD, locally released pro-inflammatory cytokines pro-duced by lymphocytes and/or thyrocytes are known to be criticallyinvolved in the pathology. Because cytokines play an importantrole in the development of AITD, possibly through changes in theexpression profile and the organization of junction proteins in thy-roid epithelial cells. We therefore investigated the expression of

IL-1� and IL-6 as well as that of their receptors on thyroid slicesfrom GD patients. A punctuated staining is observed on thyroidfollicles for IL-1� and IL-6. Moreover, their corresponding recep-tors, IL-1R1 and IL-6R, are also strongly expressed on thyroid cells

288 S.A. Rebuffat et al. / Immunobiology 218 (2013) 285– 291

F tein oZ yroidia in exp

(rp

pwt

Flr

ig. 1. Proteins involved in tight junctions. (A) Schematic representation of the proO-1) in thyroid tissues of patients with Graves’s diseases (GD) and Hashimoto’s thnd densitometric analysis of three independent experiments are shown. The prote

Fig. 2B), suggesting a possible control by the cytokines of thy-oid epithelium function and thereby their contribution to AITDathogenesis.

We then analyzed the effects of IL-1� (50 ng/mL) on thyrocytes

rimary cultures in the presence or not of TSH. IL-1� concentrationas chosen to mimic the physiological levels detected in thyroid

issue. After a 24 h culture of thyrocytes in 5H and 6H culture

ig. 2. Immunofluorescence staining for junction proteins, cytokines and their receptorsocalization of junction proteins, (a) ZO-1, (b) JAM-A and (c) Claudin. (B) Expression of cyepresentative of a minimum of 3 slides of 3 patients.

rganization in tight junctions. Expression of junction proteins (Claudin, JAM-A andtis (HT) by (B) Dot blot and (C) Western blot analysis. Representative Western blotression was normalized by the constitutive expression of �-actin.

medium, we could detect variations in the expression, localizationand organization of junction proteins. In the absence of TSH, IL-1�did not affect the expression profile of JAM-A (Fig. 4c); in contrast,we observed a decreased intracellular and cell surface immunos-

tainning as well as a reorganization of Claudin and ZO-1 proteinsafter the 24 h exposure of thyrocytes to IL-1� (Fig. 4a and b). The cel-lular ring appeared irregular in shape with alterations in junctions’

on thyroid slices from patients suffering from Graves’s disease. (A) Expression andtokines and their receptors, (a) IL-1�, (b) IL-1R1, (c) IL-6 and (d) IL-6R. Images are

S.A. Rebuffat et al. / Immunobiology 218 (2013) 285– 291 289

Fig. 3. Western blot analysis of the expression of junction proteins (Claudin and JAM-A) on membrane extracts of thyroid cells from 4 patients suffering from Graves’sd in prE y valut ans ±

tdotoT

FTto

isease. (A) The expression of junction proteins (Claudine and JAM-A) is analyzedxperiments were all duplicated with highly reproducible data. (B) Relative intensithe constitutive expression of �-actin. Results of protein level quantification are me

hickness. Furthermore, in TSH-stimulated cells, claudin and ZO-1isplayed a pattern of disorganization qualitatively similar to that

bserved in the absence of TSH but more pronounced for both pro-eins after 24 and 48 h exposures to IL-1�. No further alterationf JAM-A expression was observed in the presence of IL-1� plusSH.

ig. 4. IL-1� and TSH effects on Claudin, ZO-1 and JAM-A distribution and expression.SH, stimulated or not with IL-1� (50 ng/mL) for 24 and 48 h and then stained for junctiohyrocytes primary culture exposed to IL-1� for 24 and 48 h. Expression and localization of

f cultures of thyrocytes from 3 patients suffering from Graves’ diseases.

imary cultures of thyroid cells in the absence or presence of TSH (5H versus 6H).es were obtained with ImageJ software. The protein expression was normalized by

SD (n = 4 patients).

Discussion

Alterations of tight junctions between thyroid follicular cellsare probably major actors in the development and perpetuationof AITD. In the present study, we show that in the thyroid tissue ofpatients suffering from AITD (HT and GD), expression, localization

Thyroid cells from Graves’ patients were cultured in the presence or absence ofn proteins. Immunofluorescence staining for (a) Claudin, (b) ZO-1 and (c) JAM-A injunction proteins were detected using specific antibodies. Images are representative

2 nobio

aeaT

cbceodA(t

iiuMseAivTaaltrtc

iGaipaiTcehf

bslNamrbccarsaahAcot

90 S.A. Rebuffat et al. / Immu

nd organization of junction proteins (Claudin, ZO-1 and JAM-A),nsuring the epithelial barrier function are modified and that theselterations are dependent on the pro-inflammatory cytokine IL-1�;SH plays only a minor role.

In HT, JAM-A and ZO-1 are less expressed than in GD. Thisould account for the loss of epithelium integrity and tightnessetween the cells designing the thyroid follicle in this disease andould participate to and explain (i) antigen presentation of proteinsxpressed at the apical pole of thyrocytes by enabling the transitf immunocompetent cells through the inter-cellular space, (ii) theestruction of thyroid gland by cytotoxicity mechanisms such asDCC and CDC, resulting from the recognition by auto-antibodies

aAbs) of proteins (TPO and Tg) expressed at the apical pole ofhyroid cells.

In both AITD, lymphocytic infiltration of the thyroid glands a histopathological hallmark. The importance of tissuenfiltrating immune cells for the development and perpet-ation of autoimmune thyroid disease is now indisputable.acrophages/monocytes are involved in the processing and pre-

entation of thyroid auto-antigens and also mediate cytotoxicffects. We previously demonstrated that TPO aAbs are, via bothDCC and CDC, able to damage human thyroid cells upon bind-

ng to TPO expressed on the cell surface and that monocytes are,ia their Fc�RI, important effector cells in ADCC mediated by anti-PO aAbs (Rebuffat et al. 2008). However, a major question is howntigen-presenting cells get access to thyroid auto-antigens (suchs TPO and Tg) located at the apical membrane or in the follicleumen? This is particularly true for GD in which the integrity ofhe epithelium seems not to be altered. In HT, the cellular immuneesponse leads to epithelium destruction and consequently allowshe accessibility of TPO and Tg present at the apical pole or in theolloid of thyroid follicle to competent immune cells.

Our data now show a decrease of JAM-1 and ZO-1 expressionn thyrocytes of patients suffering from HT when compared toD patients. In GD, the presence of stimulating anti-TSH-receptorsntibodies leads to the hyperfunctionning of thyroid gland with anncrease of T3 and T4 and a decrease of TSH. In contrast, in HT theresence of TSH-stimulation blocking antibodies, which block thection of the TSH, leads to the atrophy of thyroid gland and to anncrease TSH secretion. We first explored the possible role of theSH on epithelium integrity in primary culture of human thyro-ytes. In agreement with the study by Nilsson et al., we found novidence for a major effect of TSH on epithelial barrier function inuman cultured thyroid epithelial cells, which suggests that other

actors could be involved (Nilsson et al. 1996).Pro-inflammatory cytokines are well-known to alter epithelial

arrier function. The effects of IL-1� and IFN� have been exten-ively studied. Nilsson et al. have shown that IL-1� provokes aoss of thyroid cells integrity and function (Gretzer et al. 2000;ilsson et al. 1998). Locally produced by thyroid epithelial cellsnd monocytes, IL-1� is likely to be involved in the develop-ent of autoimmune thyroid disease. On the other hand, IFN�

elease by infiltrating T-helper 1 cells (Tedelind et al. 2003) haseen demonstrated to severely impair tight junctions in primaryultured human thyrocytes obtained from GD thyroid tissue andonsequently, may provide a possible mechanism for the releasend exposure to the immune system of otherwise hidden thy-oid auto-antigens. Interestingly, more recently, IL-1� has beenuggested to play a role in AITD. In a recent study, we found anssociation between IL1B + 3954 C/T polymorphism and AITDs in

large Tunisian family (Kammoun-Krichen et al. 2007). So, IL-1�as to be considered as a valuable candidate that could operate in

ITD pathogenesis. In the present study, we found that IL-1� affectslaudin and ZO-1 expression and localization in primary culturesf human thyrocytes. The claudin family of transmembrane pro-eins plays a critical role in maintaining TJ integrity and could be

logy 218 (2013) 285– 291

responsible for the selective passage of molecules through the para-cellular space (Tsukita et al. 2001). In our study we show thatclaudin is down regulated in thyroid of patients with HT, andaffected by exposure of primary thyroid cell culture to IL-1� sug-gesting that this protein is involved in the regulation of the TJmediated by cytokines in the thyroid epithelium. We also foundthat ZO-1 is down regulated in thyroid of HT patients and redis-tributed after exposure of cultured thyroid cells to IL-1�. Theseresults further argue for a possible potential role of IL-1� in AITDspathogenesis by altering thyroid follicle epithelium integrity.

Concerning JAM-A a low expression level has been reported inthe thyroid of HT patients, whereas the protein has been foundhighly expressed in the thyroid of patients with GD. This high JAM-1protein expression in thyroid epithelium of patients with GD agreeswith the role of JAM in monocyte transmigration across the paracel-lular pathway (Ozaki et al. 1999). This allows presentation of apicalantigens to the immune system and consequently the productionof aAbs against TPO and Tg, but also the possibility for monocytesto play a role in ADCC mediated by anti-TPO aAbs (Rebuffat et al.2008). Therefore transmigration of cells across thyroid epitheliumis considered to be a cytokine-dependent process; in this respectand interestingly, we could not detect any effect of IL-1� on theexpression or localizationtion of JAM-A in primary culture of thy-rocytes. This suggests that, unlike for Claudin and ZO-1, IL-1� alonedoes not contribute to changes of JAM-A expression and organiza-tion and consequently to the related cellular transmigration.

For a better understanding of AITD, our study deservesfurtherinvestigations on cultured thyroid cells exposed to a panel of pro-inflammatory cytokines known to be induced and occur in humanthyrocytes.

Acknowledgments

This work was supported by a DGRST-CNRS grant (06/R09-06)and a LIA135-CNRS grant. The authors thank the MRI-RIO imageryplatform (Montpellier, France) for it technical assistance.

References

Chiovato, L., Bassi, P., Santini, F., Mammoli, C., Lapi, P., Carayon, P., Pinchera, A., 1993.Antibodies producing complement-mediated thyroid cytotoxicity in patientswith atrophic or goitrous autoimmune thyroiditis. J. Clin. Endocrinol. Metab. 77,1700–1705.

Del Maschio, A., De Luigi, A., Martin-Padura, I., Brockhaus, M., Bartfai, T., Fruscella, P.,Adorini, L., Martino, G., Furlan, R., De Simoni, M.G., Dejana, E., 1999. Leukocyterecruitment in the cerebrospinal fluid of mice with experimental meningitis isinhibited by an antibody to junctional adhesion molecule (JAM). J. Exp. Med.190, 1351–1356.

Estienne, V., Duthoit, C., Reichert, M., Praetor, A., Carayon, P., Hunziker, W., Ruf, J.,2002. Androgen-dependent expression of FcgammaRIIB2 by thyrocytes frompatients with autoimmune Graves’ disease: a possible molecular clue for sexdependence of autoimmune disease. FASEB J. 16, 1087–1092.

Ewert, P., Aguilera, S., Alliende, C., Kwon, Y.J., Albornoz, A., Molina, C., Urzúa, U.,Quest, A.F., Olea, N., Pérez, P., Castro, I., Barrera, M.J., Romo, R., Hermoso, M.,Leyton, C., González, M.J., 2010. Disruption of tight junction structure in salivaryglands from Sjögren’s syndrome patients is linked to proinflammatory cytokineexposure. Arthritis Rheum. 62, 1280–1289.

Gianoukakis, A., Douglas, R.S., King, C.S., Cruikshank, W.W., Smith, T.J., 2006.Immunoglobulin G from patients with Graves’ disease induces interleukin-16and RANTES expression in cultured human thyrocytes: a putative mechanismfor T-cell infiltration of the thyroid in autoimmune disease. Endocrinology 147,1941–1949.

Gretzer, C., Thomsen, P., Jansson, S., Nilsson, M., 2000. Co-culture of humanmonocytes and thyrocytes in bicameral chamber: monocyte-derived IL-1alphaimpairs the thyroid epithelial barrier. Cytokine 12, 32–40.

Guo, J., Jaume, J.C., Rapoport, B., McLachlan, S.M., 1997. Recombinant thyroidperoxidase-specific Fab converted to immunoglobulin G (IgG) molecules: evi-dence for thyroid cell damage by IgG1, but not IgG4, autoantibodies. J. Clin.

Endocrinol. Metab. 82, 925–931.

Kammoun-Krichen, M., Bougacha-Elleuch, N., Makni, K., Rebai, M., Peraldi-Roux,S., Rebai, A., Mnif, M., Abid, M., Jouida, J., Ayadi, H., 2007. Association analy-sis of interleukin gene polymorphisms in autoimmune thyroid diseases in theTunisian population. Eur. Cytokine Netw. 18, 196–200.

nobio

M

M

M

M

N

N

O

P

R

6930–6938.Weetman, A.P., 1994. The potential immunological role of the thyroid cell in autoim-

S.A. Rebuffat et al. / Immu

adara, J.L., 1989. Loosening tight junctions. Lessons from the intestine. J. Clin.Invest. 83, 1089–1094.

adara, J.L., Stafford, J., 1989. Interferon-gamma directly affects barrier function ofcultured intestinal epithelial monolayers. J. Clin. Invest. 83, 724–727.

etcalfe, R.A., Oh, Y.S., Stroud, C., Arnold, K., Weetman, A.P., 1997. Analysis ofantibody-dependent cell-mediated cytotoxicity in autoimmune thyroid disease.Autoimmunity 25, 65–72.

ullin, J.M., Kofeldt, L.M., Russo, L.M., Hagee, M.M., Dantzig, A.H., 1992. Basolateral3-O-methylglucose transport by cultured kidney (LLC-PK1) epithelial cells. Am.J. Physiol. 262, F480–F487.

ilsson, M., Husmark, J., Bjorkman, U., Ericson, L.E., 1998. Cytokines and thyroidepithelial integrity: interleukin-1alpha induces dissociation of the junctionalcomplex and paracellular leakage in filter-cultured human thyrocytes. J. Clin.Endocrinol. Metab. 83, 945–952.

ilsson, M., Husmark, J., Nilsson, B., Tisell, L.E., Ericson, L.E., 1996. Primary cultureof human thyrocytes in Transwell bicameral chamber: thyrotropin promotespolarization and epithelial barrier function. Eur. J. Endocrinol. 135, 469–480.

zaki, H., Ishii, K., Horiuchi, H., Arai, H., Kawamoto, T., Okawa, K., Iwamatsu, A., Kita,T., 1999. Cutting edge: combined treatment of TNF-alpha and IFN-gamma causesredistribution of junctional adhesion molecule in human endothelial cells. J.Immunol. 163, 553–557.

arkes, A.B., Othman, S., Hall, R., John, R., Richards, C.J., Lazarus, J.H., 1994. The role of

complement in the pathogenesis of postpartum thyroiditis. J. Clin. Endocrinol.Metab. 79, 395–400.

ebuffat, S.A., Nguyen, B., Robert, B., Castex, F., Peraldi-Roux, S., 2008. Antithyroper-oxidase antibody-dependent cytotoxicity in autoimmune thyroid disease. J. Clin.Endocrinol. Metab. 93, 929–934.

logy 218 (2013) 285– 291 291

Rodien, P., Madec, A.M., Ruf, J., Rajas, F., Bornet, H., Carayon, P., Orgiazzi, J., 1996.Antibody-dependent cell-mediated cytotoxicity in autoimmune thyroid dis-ease: relationship to antithyroperoxidase antibodies. J. Clin. Endocrinol. Metab.81, 2595–2600.

Shaw, S., Perkins, B.N., Lim, Y.C., Liu, Y., Nusrat, A., Schnell, F.J., Parkos, C.A.,Luscinskas, F.W., 2001. Reduced expression of junctional adhesion moleculeand platelet/endothelial cell adhesion molecule-1 (CD31) at human vascularendothelial junctions by cytokines tumor necrosis factor-alpha plus interferon-gamma does not reduce leukocyte transmigration under flow. Am. J. Pathol. 159,2281–2291.

Tamm, I., Cardinale, I., Krueger, J., Murphy, J.S., May, L.T., Sehgal, P.B., 1989. Inter-leukin 6 decreases cell–cell association and increases motility of ductal breastcarcinoma cells. J. Exp. Med. 170, 1649–1669.

Tedelind, S., Ericson, L.E., Karlsson, J.O., Nilsson, M., 2003. Interferon-gamma down-regulates claudin-1 and impairs the epithelial barrier function in primarycultured human thyrocytes. Eur. J. Endocrinol. 149, 215–221.

Tsukita, S., Furuse, M., Itoh, M., 2001. Multifunctional strands in tight junctions. Nat.Rev. Mol. Cell Biol. 2, 285–293.

Tsukita, S., Yamazaki, Y., Katsuno, T., Tamura, A., Tsukita, S., 2008. Tight junction-based epithelial microenvironment and cell proliferation. Oncogene 27,

mune thyroid disease. Thyroid 4, 493–499.Weetman, A.P., Bright-Thomas, R., Freeman, M., 1990. Regulation of interleukin-6

release by human thyrocytes. J. Endocrinol. 127, 357–361.