Induction of experimental autoimmune encephalomyelitis in transgenic mice expressing ovalbumin in...

Induction of experimental autoimmuneencephalomyelitis in transgenic mice expressingovalbumin in oligodendrocytes

Yi Cao*1, Catherine Toben1, Shin-Young Na1, Kirsten Stark1, Lars Nitschke*1,Alan Peterson2, Ralf Gold3, Anneliese Schimpl1 and Thomas H�nig1

1 Institute for Virology and Immunobiology, University of W�rzburg, W�rzburg,Germany

2 Laboratory of Developmental Biology, McGill University, Montreal, Canada3 Institute for MS Research, University of G�ttingen and Gemeinn�tzige Hertie-Stiftung, G�ttingen, Germany

We have used the 50 flanking sequence of the myelin basic protein gene known toinclude the core promoter and a strong oligodendrocyte (ODC)-specific enhancer totarget expression of the well-studied model antigen ovalbumin (OVA) to ODC intransgenic mice. OVA protein was detected in a tissue- and cell-specific manner in these"ODC-OVA" mice. Without immunization, CD4 T cells and B cells remained ignorant ofthe neo-self antigen expressed in the central nervous system (CNS), as indicated byunimpaired development and lack of activation of OVA/IAb-specific TCR transgenic Tcells in these mice, and the ability to mount normal OVA-specific recall and antibodyresponses. Upon immunization with OVA in complete Freund's adjuvant, about half ofthe transgenic mice developed neurological symptoms characteristic of experimentalautoimmune encephalomyelitis (EAE). Mononuclear infiltrates in the brain and spinalcord contained both macrophages and Tcells, similar to classical models of EAE inducedby immunization with CNS antigens in adjuvant. The wealth of immunological reagentsavailable to study and manipulate the OVA-specific response should make this newmodel useful for the investigation of components and mechanisms involved in CNS-specific autoimmunity.

Introduction

Transgenic mice expressing well-characterized modelantigens in a tissue-specific manner have provenextremely useful for the investigation of mechanismsinvolved in autoimmune disease. In particular, studies

on autoimmune diabetes in mice expressing the OVAantigen in pancreatic islets have generated a wealth ofinformation on the development of type 1 diabetes, dueto the availability of mouse lines expressing MHC class Iand class II restricted transgenic TCR that recognizedefined OVA-derived peptides and their variants [1].With regard to multiple sclerosis (MS), an autoimmunedisease of the central nervous system (CNS), no suchtransgenic mouse model is available so far. An MSmodelusing a well-defined antigen in which all components ofspecific antigen recognition are available would be ofparticular interest because all lymphocyte subsets and a

Clinical immunology

* The first two authors contributed equally to this work.

Correspondence: Dr. Thomas H�nig, Institute for Virology andImmunobiology, University of W�rzburg, Versbacher Str. 7,D-97078 W�rzburg, GermanyFax: +49-931-201-49243e-mail: [email protected]

Received 30/6/05Revised 20/9/05

Accepted 10/11/05

[DOI 10.1002/eji.200535211]

Key words:Autoimmunity � EAE

� Ovalbumin� Transgene

Abbreviations: CNS: central nervous system �HA: hemagglutinin � MBP: myelin basic protein � MS: multiplesclerosis � ODC: oligodendrocyte

Eur. J. Immunol. 2006. 36: 207–215 Clinical immunology 207

f 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji.de

* Present address: Department of Genetics, University ofErlangen, Erlangen, Germany

variety of effector mechanisms have been implicatedboth in the human disease itself and in the variousrodent models of EAE [2, 3].

So far, one mouse strain has been described thatharbors a transgene directing the expression of a modelantigen under the control of a CNS-specific promoter:The "GFAP-HA" mouse, which expresses viral hemag-glutinin (HA) under the control of the glial fibrillaryacidic protein promotor in astrocytes of the CNS andglial cells of the enteric nervous system [4]. AlthoughHA-specific TCR transgenic CD4 cells developing inthese mice remain ignorant of, and reactive to the HA-antigen, immunization with the relevant peptide in CFAfails to induce autoimmunity [5]. On the other hand,introduction of a class I-restricted transgenic TCR in thismodel led to the early death of the double transgenicpups due to an acute bowel inflammation caused byautoimmune destruction of HA-expressing enteric glialcells [6].

Oligodendrocytes (ODC) express most of the majorself antigens relevant for the classical EAE models andprobably also for MS [2]. We therefore sought to targetOVA-specific CD4 and CD8 T cells to ODC. Transgenicmice were generated expressing OVA under the controlof the myelin basic protein (MBP) promoter lacking theupstream enhancer element required for its activity inSchwann cells, and previously shown to confer selectiveexpression in ODC [7, 8]. Here, we show that ODC-specific expression of OVA is indeed observed in theseanimals, and that CD4 Tcells and B cells of the ODC-OVAmice appear ignorant of this artificial self antigen as theymount normal recall and antibody responses againstOVA. Importantly, active immunization induces mildEAE in about 50% of ODC-OVA mice, indicating that theOVA expressed in ODC indeed provides a target forautoimmune attack.

Results

ODC-specific expression of OVA in ODC-OVAmice

In an attempt to express the well-studied model antigenOVA as a tissue-specific autoantigen in the brain,transgenic mice were generated with an OVA cDNAunder the control of an oligodendrocyte-specific MBPpromotor construct, containing the 6.5-kb proximal partof the MBP promoter [7, 8] (Fig. 1A). For future studieson the requirement for antigen persistence in chronicinflammation of the CNS, the transgene was flankedwith loxP sites, thereby allowing its deletion uponactivation of an inducible Cre transgene. Three trans-genic founder lines were obtained and confirmed byPCR. Of these, two (lines 2 and 7) expressed the OVAprotein at levels readily detectable by Western blotting

both in brain and in spinal cord (Fig.1B, C). Importantly,OVA expression was maintained in adult mice, indicat-ing that the most 50 module contained in the promoter/enhancer construct, which is required for MBP expres-sion in the post-weaning period, is functional. Noprotein was detected by Western blotting or immuno-histochemistry in sciatic nerve (Fig. 1C), spleen, thymus,kidney, heart, lung, stomach, colon, liver, testicle, uterusand ovary (not shown).

Immunohistochemistry performed on brain sectionsrevealed distinct OVA-positive cells in young adulttransgenic but not in WT mice (Fig. 2A, B). Doublelabeling with anti-OVA and anti-20, 30cyclic nucleotide30phosphodiesterase identified the OVA-expressing cellsas ODC (Fig. 2C–E). For further experiments, line 2 wasselected. In reference to the ODC-specific expression ofOVA, these mice will be referred to as ODC-OVA mice.

Figure 1. Expression of OVA protein in the CNS of mice with anOVA transgene under the control of the proximal 6.5-kb MBPpromoter. (A) Construct and location of PCR primers used fortyping. (B) Western blots of protein extracts from brain (lane 1:WT, lane 2: ODC-OVA 4 wk old, lane 3: 12 wk old). (C) Westernblots of protein extracts from spinal cord (lane 1: WT, lane 2:ODC-OVA 4 wk old, lane 3: ODC-OVA 12 wk. old) and sciaticnerve (lane 4: WT, lane 5: ODC-OVA 4 wk old). The weak bandobserved in the extract from WT brain (panel B, lane 1) is anunrelated band of slightly higher molecular weight consis-tently observed in non-transgenic brain extracts.

Yi Cao et al. Eur. J. Immunol. 2006. 36: 207–215208


TCR transgenic CD4 cells are ignorant of OVA inODC-OVA mice

Clonal thymic deletion and peripheral T cell activationare very sensitive means of detecting even minuteamounts of 'self' antigens in vivo. Accordingly, OT-II miceexpressing a transgenic TCR on CD4 Tcells reactive withOVA peptide 323–339 presented by IAb [9] were crossedwith ODC-OVA mice, and cellularity as well as expres-sion of the OT-II TCR and of activation markers wascompared between these double transgenic and OT-IIsingle transgenic mice. As shown in Fig. 3A, the featuresdistinguishing OT-II from C57BL/6 WT mice, i.e. amarked reduction in peripheral CD8 T cells as well asthymic CD8 single-positive cells, and a mild reduction inCD4, CD8 double-positive thymocytes were maintainedin the double transgenic mice. Moreover, levels of thetransgenic TCR (Fig. 3B) and of the CD4 coreceptor (notshown) on cells expressing the OT-II receptor wereindistinguishable in OT-II single and OT-II/ODC-OVAdouble transgenic mice, indicating that neither imma-ture thymocytes nor mature CD4 T cells were recogniz-

ing their cognate antigen. This notion was furthersupported by identical levels of CD44, CD62L, CD69 andCD25 on CD4 T cells from OT-II single and OT-II/ODC-OVA double transgenic mice (data not shown).

To test for a possible induction of functional anergy,LN cells from WT, single transgenic and doubletransgenic mice were stimulated with OVA 323–339 invitro (Fig. 3C). Anti-CD3 stimulation served as a control.Cells from OT-II and OT-II/ODC-OVA double transgenicmice responded equally well to the antigenic peptide,whereas as expected, neither C57BL/6 WT nor singletransgenic ODC-OVA mice mounted a detectableresponse.

Taken together, these experiments show that OVA-derived peptides presented by IAb are not recognized bydeveloping or mature T cells in these mice. Accordingly,it can be assumed that OVA expressed byODC in the CNSdoes not reach the thymus or the periphery of ODC-OVAmice in amounts sufficient to generate stimulatory IAb/peptide complexes on APC via processing and presenta-tion.

Induction of EAE by immunization with OVA

WT and ODC-OVA mice were immunized subcuta-neously with OVA in CFA, and monitored for symptomsof EAE and for weight loss (Table 1 and Fig. 4). Abouthalf (14/26 in line 2 and 2/5 in line 7) of the transgenic,but none of the control mice developed neurologicalsymptoms characteristic of classical EAE between 8 and20 days post-infection, which, however, only rarelyprogressed beyond grade 1 (limp tail). Of 16 mice withEAE, 2 developed complete paraplegia before they weresacrificed for histological analysis. Development of EAEwas accompanied by typical weight loss, which wasabsent in immunized WT animals and animals notdisplaying disease (Fig. 4).

Histopathology of active OVA-induced EAE inODC-OVA mice

As illustrated in Fig. 5 and listed in Table 1, cellularinfiltrates were observed in the white matter of the brainand the spinal cord of ODC-OVA mice undergoing EAE,but were absent inWTmice immunized with OVA. In theCNS, the majority of the infiltrating cells were presentoutside of the perivascular cuff. In one case of severeEAE, infiltrates were also observed in the meninges ofthe brain (Table 1). Immunohistochemistry revealedboth T cells and macrophages at the sites of inflamma-tion. Furthermore, in both of the severely affectedanimals studied, focal lesions were observed in the whitematter that included demyelination, myelin degenera-tion, and the presence of degenerative and necrotic cells(not shown).

Figure 2.Detection and cellular localization of OVA in the brainof transgenic mice of line 2. Binding of polyclonal OVA-specificantibodies detected by alkaline peroxidase conjugated sec-ondary reagent identifies OVA in the brain of the transgenic (B),but not of the wild-type mouse (A). (C–E) Double fluorescentlabeling for OVA (C) and 2,3–cyclic nucleotide 3-phosphohy-drolase (CNPase, (D) in the brain of the transgenic mouselocalizes OVA to oligodendrocytes as shown by overlay (E).



Antibody production and recall responses in OVA-immunized ODC-OVA mice

In other model systems of EAE, immunization with CNSantigens in CFA primarily activates autoreactive CD4cells and antibodies, both of which can contribute toCNS inflammation, depending on the model studied[10]. To investigate if an anti-OVA response can beinduced in both of these components of the adaptiveimmune system in ODC-OVA mice, their proliferativerecall response to OVA protein as well as their serumantibody titers were compared to those induced by thesame immunization protocol in WT C57BL/6 mice.Whereas unimmunized WT or transgenic mice did notexhibit detectable proliferative responses (Fig. 6A) or

serum antibody titers (Fig. 6B) to OVA, both groups ofmice responded equally when immunized with OVA inCFA. Thus, as already suggested by the phenotype of OT-II/ODC-OVA double transgenic mice, the CD4 T cellcompartment is fully reactive to OVA in mice expressingOVA in ODC. Furthermore, there is no tolerization of theB cell compartment towards OVA in these animals,making this model useful for the study of T and B cellcontributions to CNS-specific autoimmunity.

Discussion

We present the initial characterization of a newtransgenic mouse model for human MS, in which EAE

Figure 3. OVA expressed in ODC-OVA mice is ignored by thymocytes and CD4 T cells expressing an OVA+IAb specific transgenicTCR. (A) CD4/CD8 profiles of thymocytes and LN from ODC-OVA and OT-II single and double transgenic mice. Numbers givepercentages of events in quadrants. (B) Level of transgenic TCR expression in OT-II single and OT-II/ODC-OVA double transgenicmice shown as staining intensity with Vbeta5-specific antibody on gated CD4+Valpha2+ LN cells. (C) Proliferative in vitro responseof whole LN cells from OT-II single and OT-II/ODC-OVA double transgenic mice.



can be induced by immunization with the model antigenOVA. Key features of this model are (1) selectiveexpression of the transgene in ODC, (2) absence oftolerance with regard to CD4 T cells and B cells, and (3)mononuclear infiltration and development of neurolo-gical disease in response to OVA immunization in CFA.

In two mouse lines derived from two independenttransgenic founders that expressed OVA in the CNS, thisexpression was restricted to ODC, in agreement with thereported cell-type specificity inferred on the MBPpromoter by deletion of the upstream enhancer [7].The absence of detectable OVA protein in the peripheralnervous system (Fig. 1C) and of infiltrates or lesions inthe sciatic nerve (not shown) indicates that thepromoter/enhancer construct used successfully targetedthe transgene to the CNS alone, making ODC-OVA micea useful model for CNS-specific autoimmunity (MS)without simultaneous features of autoimmunity in theperipheral nervous system (Guillain-Barr� syndrome).Importantly, the ectopically expressed model antigenOVA appears to be sufficiently sequestered in ODC toobtain a state of immunological ignorance in CD4 Tcellsand B cells of unimmunized mice. Interestingly, thiseven holds true for double transgenic ODC-OVA/OT-IImice, in which the majority of all Tcells are OVA-specificCD4 T cells. A similar observation has been made in

double transgenic mice expressing HA in astrocytestogether with a class II-restricted HA-specific TCR [5]. Inthat case, however, even active immunization did notlead to an inflammation of the CNS or to neurologicalsymptoms, whereas both could be induced in our ODC-OVA mice.

It is noteworthy that in both transgenic systems, CD8T cells apparently have an easier access to the neo-selfantigen. Thus, in the HA-GFAP system, double trans-genic mice expressing a class I-restricted TCR die at anearly age from acute enteritis due to destruction of theenteric nervous system [6]. Similarly, ODC-OVA/OT-Idouble transgenic mice develop an acute CD8-T cellmediated EAE, a phenomenon currently under investi-gation. These findings are in line with the constitutivepresentation of these model antigens in the class Ipathway, whereas cross presentation by class II-positivecells is required for recognition by CD4 T cells.

Classical EAE models rely on immunization withbrain antigens in adjuvant, a protocol preferentiallyinducing Th1 cells and antibodies. Indeed, either Th1cells by themselves or both Th1 cells and antibodies havebeen shown to contribute to EAE development,dependent on the rodent strain and the antigenemployed [2, 11]. As a possible alternative to CFA,which allows induction of mild EAE in about half of

Table 1. Summary of clinical and histological observations in ODC-OVA mice with active EAE

Onset (days Histological analysis

Animal no. Strain Sex Clinical grade post-immunization) Duration Inflammation/Days after onset

Distribution

1 2 Female 1 14 5 +/25 Med, SPb)

2 2 Female 1 14 6 +/26 Med, SP

3 2 Female 1 14 5 +/25 Med, SP

4 2 Female 5 8 >2a) +++/2 Med, SP, Mng

5 2 Female 1 12 5 +/11 Med, SP

6 2 Female 1 16 >1a) +/1 Med, SP

7 2 Male 1 14 >1a) +/1 Med, SP

8 2 Male 1 14 >1a) +/1 Med, SP

9 2 Female 4 8 >3a) +++/3 Med, SP

10 2 Male 1 14 >1a) +/1 Med, SP

11 2 Male 1 20 6 +/7 Med, SP

12 2 Female 1 14 8 +/16 Med, SP

13 2 Female 1 13 5 +/18 Med, SP

14 2 Female 1 12 10 ++/12 Med, SP

15 7 Female 1 14 >3a) +/3 Med, SP

16 7 Female 1 16 6 +/21 Med, SP

a) Sacrificed on day after onset given.b) Med: Medulla; SP: Spinal cord; Mng: Meninge.



ODC-OVA mice but only rarely supports its progressionto a fulminant stage, we also tried to induce disease withOVA conjugated to CpG oligonucleotides, previouslyshown to induce massive Th1 and CTL responses [12,13]. This conjugate was, however, unable to induce EAEin ODC-OVA mice (data not shown). Furthermore, wehave attempted to increase the incidence and severity ofEAE by depletion of CD25+ regulatory T cells in eithersingle ODC-OVA or double transgenic OT-II/ODC-OVAmice. However, in spite of effective depletion, we did notobserve an increased susceptibility. This is in contrast toEAE in MBP-specific class II-restricted TCR transgenicmice where CD4+CD25+ regulatory T cells effectivelyprevented disease [14]. It will be of interest toinvestigate whether other types of regulatory T cellssuppress more fulminant disease development in oursystem.

The reproducible induction of EAE in about half ofODC-OVA animals provides a solid basis for suchmechanistic studies, as well as for experimentaltherapies. The well-characterized reagents available in

Figure 4.Weight changes inmice immunizedwith OVA in CFA.(A) ODC-OVAmicewithout EAE symptoms (n = 7), (B) ODC-OVAmice with EAE symptoms (n = 7) and (C) WT mice (n = 9).

Figure 5.Histopathology of EAE induced by OVA-immunizationin ODC-OVA mice. ODC-OVA and WT mice were immunizedwith OVA/CFA. Brain (A) and spinal cord (B) of a transgenicmouse with grade 4 EAE display infiltration by inflammatorycells (arrows, A) and nervous tissue damage (arrow, B);Immunohistochemical staining shows CD3+ T cells (C) andmacrophages (arrows, D) in the CNS of the transgenic mouse.

Figure 6. Proliferative recall response and antibody productionin CFA/OVA immunized transgenic and WT mice. (A) Lymphnode cells fromOVA transgenic mice andWTmice immunizedwith OVA in CFA 15 days earlier were stimulated for 3 dayswithOVAbefore pulsingwith [3H]thymidine. (B) Serum levels ofOVA-specific antibody in OVA-immunized OVA transgenicmice and WT mice were measured by ELISA at day 10 p.i.



the OVA system, which beyond the TCR transgenic OT-Iand OT-II lines include mAb leading to demyelination,peptides suitable for stimulation and tetramer staining,and peptide variants allowing immunomodulation willfacilitate these experiments and will make this modeluseful for further dissection of the mechanisms by whichimmune responses to CNS antigens lead to inflammationand permanent damage. One feature of particularinterest for future studies is the potential to delete the"neo-autoantigen" by Cre-mediated deletion of thefloxed transgene. This will allow distinguishing betweenchronic antigen-specific responses and "antigen-spread".

Finally, both bacterial and viral vectors expressingOVA have been extensively characterized, making theODC-OVAmodel suitable for the study of the importanceof anti-microbial immune responses, in particular toantigens cross-reactive with brain antigens, in theinduction of EAE and MS.

Materials and methods

Generation of ODC-OVA transgenic mice

To generate the pMBPOVA construct, the standard cloningvector pSP72 2.5 kb (Promega, GeneBank accession numberX65332) was modified by inserting an additional NotIrestriction site into the Bglll site with a 22 nucleotide longlinker. Two directly repeated loxP sites were included to allowdeletion of the transgene by Cre recombinase. Before insertingthe second loxP site, the 1.8-kb OVA cDNA was released frompAc-NEO-OVA [15] and cloned into the EcoRI site of the vector.Prior to insertion of the 6.5-kb MBP promoter, an oligonucleo-tide linker containing KpnI and SalI restriction sites was clonedinto the respective sites in pSP72LX/HOVA. Before removingthe 6.5-kb MBP promoter from the pMBP clone #8 plasmid[7], a ClaI restriction site was destroyed by blunting (Klenow;MBI Fermentas, Heidelberg, Germany) before re-ligating withT4 ligase (Gibco-Invitrogen, Karlsruhe, Germany). Themodified 6.5-kb MBP promoter was subsequently releasedwith KpnI and SalI and ligated 50 upstream of the OVA cDNAinto the respective sites in pSPOVAHKS, generating pMBPO-VAHKS. The second loxP site contained within an oligonucleo-tide and supporting an AseI overhangwas cloned together withthe 0.68-kb polyA tail (ClaI/AseI) into ClaI/NotI sites inpMBPOVAHKS, resulting in pMBPOVA. TheMBPOVA constructwas released with NotI/XhoI from prokaryote sequences andeluted from a 0.8% TAE agarose gel using the QiagenQuiaquick kit and a TE-based (pH7.5) injection buffer. B.Kanzler (Freiburg, Germany) kindly carried out microinjec-tions of the linearized pMBPOVA construct into C57BL/6oocytes.

Genotyping of ODC-OVA transgenic mice

Ova was detected by PCR in DNA from tails of ODC-OVAtransgenic mice using the primers (50 TGAGGAFATGCCAGA-CAGAT) and (50 TTCCAGGATTCGGAGACAGT). PCR reagents

were purchased from PAN Biotech GmbH (Aidenbach,Germany). After 35 cycles (30 s 94�C, 30 s 57�C, 90 s72�C), the amplified fragment of 733 bp was detected on a2% agarose gel containing ethidiumbromide. As positivecontrols, primers specific for mouse A1 were used.

SDS-PAGE and Western blot for OVA expression

About 500 mg brain, spinal cord or sciatic nerve tissues wereremoved and put in 1 mL of 1% Triton X-100 (10 mM sodiumphosphate pH 7.5, 150 mM NaCl) containing a mixture ofprotease inhibitors (PMSF, phenanthroline and N-ethylmalei-mide, 2 mM each) (Sigma), minced finely and homogenizedwith oscillation at 4�C for 24 h. After centrifugation for 10 minat 15 000 � g in an Eppendorff centrifuge, the supernatantswere taken and stored frozen (–20�C) until use.

Total protein extracts were boiled in SDS-PAGE reducingbuffer for 5 min and subsequently separated by SDS-PAGE andtransferred to a polyvinylidene difluoride transfer membrane.The membrane was incubated in 0.1% Tween 20 in PBScontaining 5% nonfat dry milk overnight at 4�C, incubatedwith rabbit polyclonal anti-OVA (Sigma) for 1 h at roomtemperature, washed with 0.1% Tween 20 in PBS, andincubated with peroxidase-conjugated anti-rabbit immuno-globulin (Sigma) for 1 hour at room temperature. Afterwashing, membranes were incubated with the Western BlotChemiluminescence Reagent (NEN Life Science Products,Boston, MA) for 1 min at room temperature. After drainingexcess detection reagent, the blots were exposed to X-ray filmfrom 5 s to 5 min and then developed.

Induction and evaluation of EAE

Mice were injected s.c. at flanks and base of backs with 200 lgOVA whole protein (Sigma, Deisenhofen, Germany) in PBSemulsified in an equal volume of CFA containing Mycobacter-ium tuberculosis (Sigma) at a final concentration of 1 mg/mL.Two i.p. injections of 400 ng pertussis toxin (Sigma) weregiven 24 and 72 h later. Mice were weighed and scored fordisease on a daily basis. Disease severity was assessed using ascale ranging from 0 to 5 [16, 17]: 1, limp tail; 2, partial hindleg paralysis; 3, total hind leg paralysis; 4, hind and front limbparalysis; 5, moribund or dead.

All experiments were performed according to the Bavarianstate regulations for animal experimentation and approved bythe responsible authorities. Animals were kept at standardconditions with free access to food and water.

Histology and immunohistochemistry

Mice were anesthetized with pentobarbital and transcardiallyperfused with saline followed by 4% of paraformaldehyde.Spinal cord, brain, optic nerve, sciatic nerve, and other organswere removed and processed for routine paraffin embeddingand histological staining.

Immunohistochemistry was performed with 5-lm paraffinsections as described [18]. If necessary, antigen unmaskingwas achieved by heat pretreatment of sections for 30 min in10 mM citric acid buffer (Mac-3) or 1 mM EDTA (CD3) in amicrowave oven (850 W). After blocking of unspecific binding



sites with 10% BSA, sections were incubated overnight at 4�Cwith the appropriate primary antibody in 1% BSA. Secondaryantibodies were used as indicated below. After blocking ofendogenous peroxidase with H2O2, the peroxidase-based ABCdetection system (DAKO, Hamburg, Germany) was employedwith DAB as the chromogenic substrate. Specificity of stainingwas confirmed by omitting the primary antibody as a negativecontrol. T cells were labeled by rat anti-CD3 (Serotec,Wiesbaden, Germany; 1:300) and macrophages by rat anti-mouse Mac-3 (PharMingen; 1:200), each with a rabbit anti-ratsecondary antibody (1:100, Vector via Linavis, Wertheim,Germany). Staining for OVA was done using rabbit anti-OVAantiserum (Sigma). Pancreas sections of transgenic RIP-mOVAmice expressing OVA in islet b cells [1] were used as positivecontrol for OVA staining, and normal mouse spleen sections forCD3 and Mac-3.

To study the localization of OVA in the CNS, cryostatsections of brains and spinal cords were examined for doublelabeling with rabbit anti-OVA (Sigma) and mouse anti-20,30cyclic nucleotide 30phosphodiesterase (CNPase, SternbergerMonclonals via Szabo Scandic, Vienna, Austria). Brain andspinal cord were removed and frozen in Tissue-Tek OCT. Thecryostat sections were fixed with –20�C acetone for 20 min.The sections were treated with normal goat serum for 30 minand treated with Vector M.O. M. Immunodetection Kit (VectorLaboratories, Burlingame, USA). The thoroughly washed slideswere sequentially treated with rabbit anti-OVA, goat-anti-rabbit-Cy2, mouse anti-CNPase, biotinylated anti-mouse IgG,and streptavidin-Cy3 for 1 h at room temperature.

ELISA

The 96-well ELISA plates were coated with 10 lg/mL OVA incarbonate coating buffer pH 9.6 for 12 h at 4�C. Plates wereblocked for 2 h at 37�C with 0.2% gelatin and washed twice inPBST (PBS with 0.1% Tween 20). After washing, 100 lL ofdiluted sera (1:200, 1:400, 1:800, 1:1600) in PBS were addedand incubated for 2 h at room temperature. Sera from naiveWTmice were used as negative controls while mouse anti-OVAmAb was used as a positive control. After washing in PBST, theplates were incubated for 1 h at room temperature withalkaline phosphatase-conjugated goat anti-mouse IgG (c-chain) and goat anti-mouse IgM (l-chain). The reactionproduct was visualized using p-nitrophenylphosphat indiethanolamin buffer and the absorbance read at 405 nm.

Flow cytometry

Single-cell suspensions of freshly isolated thymocytes andpooled LN cells were resuspended in PBS (0.1% BSA, 0.02%NaN3) at a concentration of 1 � 106cells/mL followed byincubation at 4�C for 20 min with 50 lL of properly dilutedmAb. Four-color stainings were carried out with mAb eitherdirectly labeled with FITC, R-PE, allophycocyanin, or werebiotinylated. For the latter, avidin-Cy-chrome (PE-Cy5) wasused as a secondary reagent for detection. After staining, cellswere washed twice with PBS/BSA/NaN3. Relative fluorescenceintensities were then measured with FACScalibur (BectonDickinson, Heidelberg, Germany) using CellQuestTM software(Becton Dickinson).

T cell proliferation

A total of 2 � 105 LN cells/0.2 mL supplemented RPMI 1640/5% FCS were incubated at 37�C in triplicate in round-bottom96-well tissue culture plates with various concentrations ofindicated OVA (grade V, Sigma) or synthetic OVA peptides,323–339 (ISQAVHAAHAEINEAGR) or 257–264 (SIINFEKL) for72 h. In parallel, soluble anti-CD3emAb (5 lg/mL) was addedas a polyclonal stimulator to the cell culture. During the last16 h of the culture [3H] thymidine (1 lCi/well) was addedafter which its incorporation into DNA of proliferatinglymphocytes was measured using a Betaplate scintillationcounter.

Acknowledgements: We thank Drs. B. Kanzler and T.Boehm for microinjection of our constructs into C57BL/6 oocytes, and Dr. C. Kurts for providing tissue samplesand reagents for OVA-specific immune histology. Thisproject was supported by DFG through SFB 581 and bythe IZKF W�rzburg.

References

1 Kurts, C., Heath, W. R., Carbone, F. R., Allison, J., Miller, J. F. andKosaka, H., Constitutive class I-restricted exogenous presentation of selfantigens in vivo. J. Exp. Med. 1996. 184: 923–930.

2 Gold, R., Hartung, H. P. and Toyka, K. V., Animal models for autoimmunedemyelinating disorders of the nervous system. Mol. Med. Today 2000. 6:88–91.

3 Lassmann, H. and Ransohoff, R. M., The CD4-Th1 model for multiplesclerosis: a critical [correction of crucial] re-appraisal. Trends Immunol.2004. 25: 132–137.

4 Cabarrocas, J., Bauer, J., Piaggio, E., Liblau, R. and Lassmann, H.,Effective and selective immune surveillance of the brain by MHC class I-restricted cytotoxic T lymphocytes. Eur. J. Immunol 2003. 33: 1174–1182.

5 Cabarrocas, J., Piaggio, E., Zappulla, J. P., Desbois, S., Mars, L. T.,Lassmann, H. and Liblau, R. S., A transgenic mouse model for T cellignorance of a glial autoantigen. J. Autoimmun. 2004. 22: 179–189.

6 Cornet, A., Savidge, T. C., Cabarrocas, J., Deng, W. L., Colombel, J. F.,Lassmann, H., Desreumaux, P. and Liblau, R. S., Enterocolitis induced byautoimmune targeting of enteric glial cells: a possible mechanism in Crohn'sdisease? Proc. Natl. Acad. Sci. USA 2001. 98: 13306–13311.

7 Forghani, R., Garofalo, L., Foran, D. R., Farhadi, H. F., Lepage, P.,Hudson, T. J., Tretjakoff, I. et al., A distal upstream enhancer from themyelin basic protein gene regulates expression in myelin-forming schwanncells. J. Neurosci. 2001. 21: 3780–3787.

8 Farhadi, H. F., Lepage, P., Forghani, R., Friedman, H. C., Orfali, W.,Jasmin, L., Miller, W. et al., A combinatorial network of evolutionarilyconserved myelin basic protein regulatory sequences confers distinct glial-specific phenotypes. J. Neurosci. 2003. 23: 10214–10223.

9 Barnden, M. J., Allison, J., Heath,W. R. and Carbone, F. R.,Defective TCRexpression in transgenic mice constructed using cDNA-based alpha- andbeta-chain genes under the control of heterologous regulatory elements.Immunol. Cell Biol. 1998. 76: 34–40.

10 Lyons, J. A., Ramsbottom, M. J. and Cross, A. H., Critical role of antigen-specific antibody in experimental autoimmune encephalomyelitis inducedby recombinant myelin oligodendrocyte glycoprotein. Eur. J. Immunol.2002. 32: 1905–1913.

11 Tuohy, V. K., Sobel, R. A. and Lees, M. B., Myelin proteolipid protein-induced experimental allergic encephalomyelitis. Variations of diseaseexpression in different strains of mice. J. Immunol. 1988. 140: 1868–1873.

12 Vabulas, R. M., Pircher, H., Lipford, G. B., Hacker, H. and Wagner, H.,CpG-DNA activates in vivoT cell epitope presenting dendritic cells to trigger



protective antiviral cytotoxic T cell responses. J. Immunol. 2000. 164:2372–2378.

13 Maurer, T., Heit, A., Hochrein, H., Ampenberger, F., O'Keeffe, M., Bauer,S., Lipford, G. B. et al., CpG-DNA aided cross-presentation of solubleantigens by dendritic cells. Eur. J. Immunol. 2002. 32: 2356–2364.

14 Furtado, G. C., De Lafaille, M. A., Kutchukhidze, N. and Lafaille, J. J.,Interleukin 2 signaling is required for CD4(+) regulatory T cell function. J.Exp. Med. 2002. 196: 851–857.

15 Moore, M. W., Carbone, F. R. and Bevan, M. J., Introduction of solubleprotein into the class I pathway of antigen processing and presentation. Cell1988. 54: 777–785.

16 Baron, J. L., Madri, J. A., Ruddle, N. H., Hashim, G. and Janeway, C. A.,Jr., Surface expression of alpha 4 integrin by CD4 Tcells is required for theirentry into brain parenchyma. J. Exp. Med. 1993. 177: 57–68.

17 Gold, R., Giegerich, G., Hartung, H.-P. and Toyka, K. V., T cell receptorusage in Lewis rat experimental autoimmune encephalomyelitis: TCR bchain variable region Vb8.2 positive T cells are not essential for inductionand course of disease. Proc. Natl. Acad. Sci. USA 1995. 92: 5850–5854.

18 Linker, R. A., Maurer, M., Gaupp, S., Martini, R., Holtmann, B., Giess, R.,Rieckmann, P., et al., CNTF is a major protective factor in demyenilatingCNS disease: a neurotrophic cytokine as modulator in neuroinflammation.Nat. Med. 2002. 8: 620–624.



Induction of experimental autoimmune encephalomyelitis in transgenic mice expressing ovalbumin in...

Documents

Transcript of Induction of experimental autoimmune encephalomyelitis in transgenic mice expressing ovalbumin in...