Anti-thyroperoxida antibodies from patients with Hashimoto's
encephalopathy bind to cerebellar astrocytes
Stéphanie Blanchin a,⁎,Christine Coffin b ,Fausto Viader c ,Jean Ruf d ,Pierre Carayon d ,
Francette Potier a ,Estelle Portier a ,Elisabeth Comby a ,Stéphane Allouche e ,
Yann Ollivier f ,Yves Reznik b ,Jean Jacques Ballet a
a
Laboratoire d ’Immunologie et d ’Immunopathologie,UPRES-EA 2128,CHU Clémenceau,14033Caen cedex,France
b
Service d ’Endocrinologie et Maladies Métaboliques,CHU Côte de Nacre,14033Caen cedex,France
c
Service de Neurologie,CHU Côte de Nacre,14033Caen cedex,France
d
INSERM U555,Facultéde médecine Timone,Universitéde la Méditerranée,13385,Marille cedex 5,France
e
Laboratoire de Biochimie,CHU Côte de Nacre,14033Caen cedex,France f
Service de Médecine Interne,CHU Côte de Nacre,14033Caen cedex,France
Received 28June 2007;received in revid form 31July 2007;accepted 6August 2007
Abstract
A cohort of 10Hashimoto's encephalopathy (HE)patients,33patients with unrelated neurological symptoms,12Hashimoto's thyroiditis patients and 4healthy adult donors was studied to explore the neurological targets of anti-thyroperoxida (TPO)autoantibodies (aAb)in HE.High levels of anti-TPO aAb were only detected in HE group's cerebrospinal fluids.In immunofluorescence assays on monkey brain cerebellum ctions,both HE patients'ra and anti-TPO monoclonal antibodies (mAb)
were able to bind cerebellar cells expressing glial fibrillary acid protein.Normal human astrocytes from primary cultures also reacted with anti-TPO mAb.Specific astrocyte binding of anti-TPO aAb suggests a role of the aAb in the HE pathogenesis.©2007Elvier B.V .All rights rerved.
Keywords:Anti-thyroperoxida antibodies;Astrocytes;Cerebrospinal fluid;Hashimoto's encephalopathy;Hashimoto's thyroiditis
1.Introduction
Hashimoto's encephalopathy (HE)is a rare,often misdiag-nod and poorly understood corticosteroid-responsive neuro-logical syndrome occurring in patients with autoimmune thyroid dia (AITD).Since the first description in 1966by Brain et al.(Brain et al.,1966),ca reports described heterogeneous clinical manifestations such as psychiatric impairments and neurological disturbances (Kothbauer-Margreiter et al.,1996).The diagnosis was established on the basis of unspecific central nervous dys-function with the prence of rum anti-thyroid autoantibodies (aAb),regardless of the thyroid disorder (Fatourechi,2005).An autoimmune pathogenic link was suggested between HE and AITD bad on HE features such as (i)prence of rum anti-thyroid aAb indicating active thyroiditis (Brain et al.,1966),(ii)cerebrospinal fluid (CSF)biochemistry suggestin
g inflammatory process (Ferracci et al.,2003),(iii)association with other autoimmune dias such as systemic lupus erythematosus and rheumatoid arthritis (Mulhern et al.,1966),(iv)x ratio and age distribution similar to AITD (Chaudhuri and Behan,2003;Sawka et al.,2002),(v)the improvement of symptoms after corticoste-roid therapy (Chong et al.,2003),and (vi)prence of rum aAb against neuronal alpha-enola (Fujii et al.,2005;Ochi et al.,2002;Yoneda et al.,2007)and an unidentified 36kDa autoantigen (Oide et al.,2004).HE is also characterized by an increa in the CSF anti-thyroid,especially anti-thyroperoxida (TPO)aAb levels (Ferracci et al.,2003)which was propod as a valuable marker of diagnosis (Ferracci et al.,2004).
The main thyroid autoantigens,TPO and thyroglobulin (Tg),are the enzyme-substrate pair involved in thyroid hormone production (Ruf and Carayon,2006).TPO is a 933-amino-acid
Journal of Neuroimmunology 192(2007)13–
20
/locate/jneuroim
⁎Corresponding author.Tel.:+33231272551;fax:+33231272550.E-mail address:blanchin-s@chu-caen.fr (S.Blanchin).
0165-5728/$-e front matter ©2007Elvier B.V .All rights rerved.doi:10.1016/j.jneuroim.2007.08.012
long,type I integral membrane protein,which carries out the iodination and the intramolecular coupling of tyrosine residues of Tg to form thyroid hormones.Human(Chazenbalk et al., 1993)and murine(Ruf et al.,1989)monoclonal antibodies (mAb)have been ud for characterization of the TPO immuno-dominant region(IDR).
The aim of the prent study was to asss the reactivity of anti-TPO aAb from HE patients with central nervous system (CNS)tissues.In immunofluorescence assays,both HE patient rum anti-TPO aAb and murine anti-TPO mAb were bound to astrocytic cells.
2.Materials and methods
2.1.HE patients and control subjects
atthemomentSera and CSF from10HE patients admitted between January 1999and April2006to the neurological ward of Caen University Hospital were collected and kept at−80°C until ud.Diagnosis was bad on the neurological clinical manifestations and the prence of anti-Tg and-TPO aAb in the rum and CSF in the abnce of other central neurological conditions.Serum and CSF samples were also obtained from a panel of33control patients who were referred to the emergency ward with meningitis(15), intractable headache(8),peripheral neuropathy(4),multiple sclerosis(6).They were free of AITD except two who were diagnod with an Hashimoto's thyroiditis(HT)(rum anti-TPO aAb:80and100IU/mL,normal values:b25IU/mL).Sera from 12patients with a documented HTand rum anti-TPO aAb levels ranging from90to N200IU/mL were obtained from experienced endocrinologists.Normal ra were generous gifts from4healthy adult volunteers without AITD or CNS dia.Sera from HE patients and healthy control donors were tested for the prence of anti-nuclear aAb as described(Comby et al.,2006)and of anti-onconeuronal aAb(anti-Ri,-Hu,-Yo,-CV2,-amphiphysin,-Ma2 assays,Euroimmun GmbH,Gross Grönau,Germany,and Ravo Diagnostika GmbH,Freiburg,Germany)according to the manufacturers'instructions.
The procedures ud in this study were approved by the local institutional review committee,and info
rmed connt was ob-tained from all donors.
2.2.ELISA for the detection of rum and CSF anti-thyroid aAb
mieWells of microtiter plates(Dynatech Laboratories Inc., Chantilly,V A)were coated with phosphate-buffered saline (PBS),pH7.4,containing300ng human TPO(Biodesign Inter-national,Saco,ME)or1000ng human Tg(Valbiotech,Paris, France).After being incubated overnight at4°C,the wells were saturated with1%bovine rum albumin(BSA;Laboratoires Eurobio,Courtaboeuf,France)in PBS.They were then incubated with1/800rum dilution for anti-TPO aAb,1/400rum dilution for anti-Tg aAb or1/5dilution for CSF samples in PBS,1%BSA with0.05%Tween20for1h30min at37°C.A calibration curve was drawn up using World Health Organization standards (NIBSC Codes:66/387and65/93for anti-TPO and-Tg aAb,respectively)and the BP114as substandard plasma(The Binding Site,Saint Egreve,France).After a washing step,the anti-Tg and -TPO aAb levels were both determined using alkaline phospha-ta-coupled goat antibodies to human IgG(P.A.R.I.S,Com-piègne,France)and p-nitrophenyl phosphate(Interchim, Montluçon,France)as substrate.Optical density(OD)was read at405nm using a Bio-Tek Elx808microplate reader(Bio-Tek Instruments Inc,Winooski,VT).Results were expresd as means of duplicate measurements in IU/mL in the ca of ra and OD [ΔOD405nm=(OD405nm of coated wel
ls)minus(OD405nm of uncoated wells)]in that of CSF.For18ra from healthy adult volunteers,threshold values for both anti-TPO and-Tg aAb was 25IU/ml,determined as mean values plus3SD from the mean.
2.3.Depletion of HE patients’ra in anti-TPO aAb
Anti-TPO aAb in ra of HE patients were depleted by performing affinity chromatography(Ruf et al.,1992).Briefly, 10mg of purified human TPO was coupled to25mL Affi-Gel15 (Bio-Rad,Marnes la Coquette,France).After extensive PBS washing,patients’ra were incubated overnight with the coupled gel at4°C under shaking.Unbound material was then checked by ELISA to ensure that it contained no anti-thyroid aAb.
cereal是什么意思
2.4.Primary normal human astrocyte cell cultures
Normal primary human astrocytes derived from the whole brain of one donor,18weeks old fetus,(Cambrex Bio Science, Verviers,Belgium)were grown on coverslips in Astrocyte Basal Medium supplemented with the Astrocyte Growth Medium SingleQuots(Cambrex Bio Science).
2.5.Immunofluorescence assays
Staining was performed on cerebrum,cerebellum and thy-roid tissue ctions from Macaccus rhesus monkeys(The Binding Site)and normal human astrocyte coverslips.Human astrocytes were fixed and permeabilized in100%acetone.After BSA saturation,tissues and astrocytes were incubated for1h with HE patients’ra(1/15dilution),HE patients’ra devoid of anti-TPO aAb(1/15dilution),HT patients’ra and normal ra(1/15dilution),4murine anti-TPO mAb directed towards TPO IDR:mAb47,60,15and9(Ruf et al.,1989)and1murine anti-Tg mAb:mAbJ7C9.3(Ruf et al.,1983)(1/300dilution) and with anti-glial fibrillary acid protein(GFAP)mAb(1/1000 dilution;Sigma-Aldrich,Saint Quentin Fallavier,France)in PBS,3%BSA.Dilutions of ra and antibodies yielding optimal signal/background fluorescence were experimentally determined by rial tests.Staining patterns were revealed with rhodamine or FITC coupled goat antibodies against mou IgG (Jackson ImmunoRearch Laboratories,West Grove,PA)or FITC coupled goat antibodies against human IgG(Euroimmun GmbH)for1h.Sections were then mounted with Vectashield-DAPI(Vector Laboratories,Peterborough,UK)and viewed using an Axioskop2Plus fluorescence microscope(Carl Zeiss Vision GmbH,Munchen,Germany)with a40×oil immersion lens.Pictures were acquired with an AxioCam HRC digital
14S.Blanchin et al./Journal of Neuroimmunology192(2007)13–20
七年级数学补习
camera equipped with an AxioVision 3.2software program.The Photoshop software program (version 6;Adobe System Inc.,San Jo,CA)was ud to superimpo the pictures for colocalization viewing.
2.6.SDS-PAGE and western blot analysis
Two μg of purified human TPO were incubated in Laemmli sample buffer and loaded onto 7.5%acrylamide gels.Proteins were either stained with Coomassie brilliant blue or electro-transferred onto a nitrocellulo sheet (Amersham Biosciences,Buckinghamshire,UK)for western blotting.After BSA satura-tion,the blotted membranes were incubated overnight at 4°C with HE patients'ra (1/100dilution)or mAb47(1/300dilu-tion)in PBS,0.05%Tween 20.Membranes were then incubated with alkaline phosphata-labeled goat antibodies against human IgG or rabbit antibodies against mou IgG (Rockland Immunochemicals for rearch,Gilbertsville,PA)for 2h.Finally,a revelation step was performed using 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium as the appro-priate substrate (Sigma-Aldrich).2.7.Statistical analysis
Results are expresd as means ±1SD.Significance was determined using unpaired t -tests or one-way ANOVA with Dunnett's post test.Linear correlations were evaluated using R values.Significance levels were t at p b 0.01.
3.Results
3.1.Clinical and biological characteristics of HE patients The HE cohort consists of 1man and 9women.Their ages ranged from 25to 85(mean:53years)at the ont of the dia.The autoimmune thyroid disorders were HT for 8patients and Graves'dia (GD)with anti-TPO aAb for 2patients.Clinical data are prented in Table 1according to major criteria from Chaudhuri and Behan (2003)and Ferracci et al.(2004).The clinical manifestations included impaired consciousness,tremor,cephalalgia,coma,izure,paraparesis,hemiparesis,confusion,ataxia,memory loss,myoclonus,apraxia and aphasia.Neither dementia nor behavioural disturbance was obrved.In 4patients,non-migraine-type cephalalgia was initially recorded as a prominent symptom,and resolved within 2months in one while persisting for over 6months in the three others.Abnormal electroencephalography (EEG)tracing changes were obrved in 3/10patients with diffu intermittent runs of synchronous sharp waves.Brain magnetic resonance imaging (MRI)was performed in 7/10patients with normal imaging in all subjects except one with unspecific hyperinten spots in the white matter.One day to 4months after initial manifestations,9/10patients underwent a corticosteroid trial (prednisone equivalent:1mg/kg/day)during a minimum period of 6months.One patient (no.2)remained in spontaneous remission for at least 2years and did not receiv
家务的英文e corticosteroid therapy.Two months after initiation of therapy,symptoms were prent in 5/9treated patients and abnt in 4/9.
Table 1
Clinical features of HE patients Patient no.12345678910Sex F F F F F F F F F M Age
51413545858457552553Thyroid disorder
HT HT HT HT HT HT GD HT GD HT Symptoms
Impaired consciousness +−−++++−++Tremor ++++++++++Cephalalgia
−−++−−−++−Migraine-type headache −−−−−−−−−−Coma −−−−−−−−+−Seizure −−+−−−−−+−Paraparesis −−+−−−−−−−Hemiparesis +−+−−−−−−−Confusion −−−++++−++Ataxia
+++−−−−++−Memory loss +−−−−−−+−+Myoclonus −−−+−−−−++Apraxia −+−−−−−−−−Aphasia
−+−−−−−+++Focal neurological deficit −
−
−
−
party girl−
michael page−
−
−
−
−
EEG Normal Diffu abn Normal Normal Normal Normal Normal Normal Diffu abn Diffu abn MRI
NA Normal Normal Hypersignal NA Normal Normal Normal NA Normal Corticosteroid therapy +−++++++++Evolution at 6months
CR CR CD CR Death CD CD CR CD CR Persisting symptoms at 6months/12months
发型设计学校−/−
−/−
withoutu+/−
+/−terrell owens
+/NA
−/NA
+/NA
+/−
+/+
−/−
F:female,M:male,HT:Hashimoto's thyroiditis,GD:Graves ’dia,+:prence,−:abnce,abn:abnormalities,NA:not available,CR:complete remission,CD:corticodependence.
15
S.Blanchin et al./Journal of Neuroimmunology 192(2007)13–20
At6months,steroid therapy was discontinued in8/9patients. From6months to12months,corticodependence was obrved in 4/9patients(no.3,6,7,9).All9steroid-treated patients were clinically improved.At12months,1(no.9)was symptomatic under therapy,5(no.1,3,4,8,10)were asymptomatic after cessation of therapy and2(no.6,7)were under therapy with a follow-up limited to7months.One patient(no.5)had died of an unrelated condition.From12months,therapy was continued in only1/10patient(no.9).All10HE patients exhibited high levels (N50IU/mL)of anti-TPO or-Tg aAb in their ra,or both (Table2).Nine HE patients were evaluated for CSF protein concentration and IgG index which were within normal or near normal ranges(Table2).No rum antibodies against nuclear and onconeuronal antigens were found in HE patients nor healthy donors(data not shown).
3.2.High frequency of anti-TPO aAb in CSF from HE patients
Thirty three CSF control samples corresponding to diver CNS dias with or without HT(2patients)status were quantified for their respective anti-TPO and-Tg aAb level.All the CSF exhibited lowest levels of anti-TPO and-Tg aAb which were ud to determine threshold values.The values expresd inΔOD405nm(meanΔOD405nm+2SD),were t at0.228for anti-TPO aAb and0.115for anti-Tg aAb.High levels of anti-TPO (ΔOD N0.228)and-Tg(ΔOD N0.115)aAb were detected in all the HE patients'CSF(p b0.01;Fig.1).No correlation was found between rum and CSF levels of anti-TPO and-Tg aAb or between the respective anti-Tg and-TPO aAb levels.
3.3.Serum anti-TPO autoantibody titer follow-up after initiation of corticosteroid therapy
In6HE patients(no.1,2,3,4,9,10),rum anti-TPO aAb levels were evaluated2and12months after initiation of corticosteroid therapy(Fig.2).Four out of6exhibited decread anti-TPO aAb levels after2months(p b0.01),and in3,antibody titers fell below the threshold anti-TPO aAb value (b25IU/mL)after12months.Interestingly,no alteration of anti-TPO aAb(N200IU/mL)was noted in one HE patient(no.
9)after a5-year therapy(data not shown).
3.4.Anti-TPO aAb and mAb recognize the same TPO isoforms
To determine the specificity of anti-TPO aAb from HE patients’ra,a western blotting assay was performed on affinity-purified human TPO.TPO at100kDa was detected by performing Coomassie brilliant blue staining(Fig.3,A2).In western blotting assays,anti-TPO mAb47,reacted with the2 main isoforms forming the characteristic TPO doublet around 100kDa(Fig.3,B2).Both TPO isoforms were recognized by ra from the10HE patients as shown in Fig.3,C2which depicts the typical reactivity of the HE patient's no.9rum.
3.5.Anti-TPO aAb bind to primate cerebellar tissues
To investigate more cloly the role of anti-TPO aAb in CNS abnormalities,indirect immunofluorescence assays with10HE patients,12HT patients and4healthy donors’ra were performed on CNS and thyroid tissues from M.rhesus.Fig.4A depicts the typical reactivity of one of HE patients'ra(no.9), the same rum depleted from anti-TPO aAb,one HT patient rum(anti-TPO aAb N200IU/mL)and normal rum on cerebellum and thyroid tissues.The anti-TPO aAb reactivity of all ra were first evaluated by specific staining of the thyroid
Table2
Biological features of HE patients
Patient no.12345678910
CSF proteins(NR:0.15–0.45g/L)NA0.420.360.28NA0.35NA0.56NA0.69 IgG index(NR b0.7)NA0.58NA0.580.45NA NA0.47NA0.66 Serum anti-Tg aAb(NR b25IU/mL)17844N300225N30065150N30020041 Serum anti-TPO aAb(NR b25IU/mL)25150N20060N200N20080N200N200170 CSF anti-Tg aAb(ΔOD405nm b0.115)NA0.20 1.500.50 1.500.250.31 2.25 1.470.14 CSF anti-TPO aAb(ΔOD405nm b0.228)NA 3.49 2.320.280.78 2.670.75 4.03 3.91 1.89
CSF:cerebrospinal fluid,NA:not available,NR:normal range,aAb:autoantibodies,Tg:thyroglobulin,TPO:thyroperoxida,ΔOD405nm:optical densities at 405
nm.
Fig.1.Anti-TPO and-Tg aAb in CSF from HE patients.CSF from9HE patients
was analyd by ELISA to determine the anti-TPO(A,white bars)and-Tg(B,black
bars)aAb levels.Results were expresd as the meanΔOD405nm value±1SD
(error bars).Dotted lines give threshold values for anti-TPO and-Tg aAb.
16S.Blanchin et al./Journal of Neuroimmunology192(2007)13–20
cell membrane (Fig.4,A6).The 10ra from HE patients (Fig.4,A1)but not the 12from HT patients (Fig.4,A3)or the 4from healthy subjects (Fig.4,A4)bound to structures in primate cerebellar tissues.After the depletion procedure which removed more than 90%of anti-TPO aAb but not anti-Tg aAb (Fig.4D and data not shown),the rum did not bind significantly to either thyroid cells or cerebellar cells (Fig.4,A2and A7).As the HE patients'ra contained anti-TPO and -Tg aAb,immuno-fluorescence assays were also performed with mAb directed against human TPO or Tg (Fig.4,B and C)to confirm the specificity of cerebellar staining.Four mAb characteristic of the TPO IDR (mAb47,60,15,9)and one directed toward human Tg (J7C9.3)were lected.The 4anti-TPO mAb all s
howed a similar reactivity pattern as en with HE patients'ra on primate cerebellar cells (Fig.4B).Stained cells were mostly located in the subcortical cerebellar white matter,near the granular cell layer.No staining of cerebellar neuronal cells,including granular and Purkinje cells,was obrved.The anti-Tg mAbJ7C9.3did not bind to cerebellar tissues (Fig.4,C1).None of the above ra and mAb bound to primate cerebrum tissues (data not shown).The results further suggest that the sole anti-TPO aAb electively recognize primate cerebellum structures.3.6.Reactivity of anti-TPO antibodies to GF AP-expressing astrocytic cells
To identify further the cerebellar cell type(s)reacting with anti-TPO aAb,a double immunofluorescence assay with HE patients'ra and anti-GFAP mAb was performed on cerebellar tissues (Fig.5).Anti-TPO aAb from 2HE patients'ra (no.2and no.3)bound to GFAP-positive astrocytic cells.Similarly,the anti-TPO mAb47stained strongly at the membrane (Fig.5,D)of normal human astrocyte primary cell cultures expressing GFAP (Fig.5,E).The data therefore indicate that anti-TPO aAb from HE patients and anti-TPO mAb both bind to astrocytes.4.Discussion
Prent data suggest that anti-TPO and/or -Tg aAb are prent in the CSF of HE patients and that their rum anti-TPO aAb bind cells of astrocyte lineage.
A nsitive ELISA was performed to detect CSF anti-thyroid aAb in two 9HE patients and 33control adults with other neurological conditions.In accordance with the HE cas already reported in the literature (Chong et al.,2003),2GD patients with subclinical or overt hyperthyroidism were included.In the HE patients,anti-TPO and/or anti-Tg aAb were detected at variable levels in ra and significant levels in CSF,consistent with previous reports (Castillo et al.,2006;Ferracci et al.,2003).Their abnce in the 33control patients including 2patients with HT suggests that this finding was restricted to HE and thus may provide a clue for early HE diagnosisS inflammation was not documented since normal or near normal range CSF protein concentrations and IgG indices,and normal or unspecific brain MRI and EEG patterns were found.There is no evidence to conclude between either local antibody synthesis or blood brain barrier passage which is suggested in patients no.4and no.7with lowest rum and CSF aAb levels.
Anti-TPO aAb in the CNS of HE patients might contribute to CNS pathology by interacting with CNS tissues,although not excluding a role for some other pathogenic antibodies.A role of aAb is further suggested by the marked clinical
improvement
Fig.2.Time cour study of rum anti-TPO aAb level in HE patients.Serum anti-TPO aAb levels (IU/ml)were quentially determined in 6HE patients (no.1,2,3,4,9and 10)for 2to 12months after the ont of corticosteroid therapy.Results were expresd as mean values of triplicated experiments.Standard deviations from all means were lower than 10%of the mean.Threshold value for normal ra (25IU/mL)as described in the Materials and methods
ction.
Fig.3.Reactivity of anti-TPO aAb from HE patients and anti-TPO mAb with purified human TPO.Purified human TPO samples were run on SDS-PAGE gel and proteins were either directly stained with Coomassie brillant blue (A,2)or electrotransferred onto a nitrocellulo membrane and r
evealed in western blotting experiments using either anti-TPO mAb47(B,2)or HE patient's no.9rum (C,2).The molecular weights of the standards (column 1)are indicated in kDa on the left.Data shown are of a reprentative experiment.
17
S.Blanchin et al./Journal of Neuroimmunology 192(2007)13–20
