JOURNAL OF IMMUNOLOGICAL METHODS

ELSEVIER Journal of Immunological Methods 195 (1996) 81-92

Autoantibody-mediated capture and presentation of autoantigen to T cells via the Fcs receptor by a recombinant human

autoantibody Fab converted to IgE

Jin Guo a, Sonia Quaratino b, Juan Carlos Jaume a, Giuseppe Costante ay1, Marco Londei b, Sandra M. McLachlan a, Basil Rapoport a3 *

’ Thyroid Molecular Biology Unit (1 I ITi, Veterans’ Administration Medical Center and the Uniwrsity of California. San Francisco, 4150 Clement Street, San Francisco. CA 94121, USA

b The Mathilda and Terence Kennedy Institute of Rheumatology. Sunley Division, London W6 8LW, UK

Received 30 January 1996; revised 26 March 1996: accepted 19 April 1996

Abstract

FCC receptor (CD23)-mediated capture of IgE-antigen complexes by B cells provides a powerful antigen presenting system. Our goal was to develop a system using high affinity, human, organ-specific monoclonal autoantibodies for antigen capture by B cells. For this purpose, we converted a recombinant human autoantibody to TPO from a Fab (SP1.41 to an IgE molecule. Sera from all patients with autoimmune thyroid disease contain autoantibodies with the same epitope as SP1.4. The SP1.4 H and L chain V region genes were spliced by overlap PCR to a mammalian, non-immunoglobulin signal peptide and transferred to expression vectors for human IgGl and K, respectively. After inserting the IgE constant region genes into the H chain vector, the K and IgE H chain vectors were expressed in SP2/0 cells. SPl.CIgE retains its high affinity (K,) for TPO (-2X lo-” M), recognizes the same epitope as Fab SP1.4 and, importantly, binds to a different epitope than does Fab TR1.9. Binding of preformed complexes of SPl.4-IgE and biotinylated TPO to EB virus transformed B cells (EBVL) was weakly detectable by flow cytometry and was displaced by unlabeled TPO. SP1.4-IgE/ ‘251-TP0 complex binding to EBVL was much more clearly evident, was also inhibited by the addition of unlabeled TPO, and was greatly reduced by preincubation of the EBVL with antiCD23. Further, autologous EBVL preincubated with SPl.CIgE/TPO complexes stimulated proliferation of TPO-specific T cells. IgE autoantibody-mediated antigen focusing to B cells is unlikely to operate in vivo but is, instead, a powerful investigative tool.

In conclusion, SP1.4-IgE is the first monoclonal human autoantibody to be developed for IgE-mediated antigen presentation to T cells by EBVL. Recombinant human autoantibodies converted to IgE, possibly in combinations if their epitopes permit simultaneous binding to the same molecule, provide a unique system to generate human T cell lines and clones specific for peptides naturally processed from internalized high affinity autoantibody/autoantigen complexes.

Abbreviations: EBVL, Epstein-Barr virus transformed B cell line; H, heavy; L, light; pAH4604, pAG4447 and pAN4621, vectors for

expression of IgGl H chain, IgE H chain and K L chains respectively; TPO, thyroid peroxidase; SP2/0, mouse myeloma line; V, variable. * Corresponding author. Tel.: (415) 476-5984; Fax: (415) 752-6745. ’ Present address: University of Reggio Calabria, Catanzaro,

Italy.

0022.1759/96/$15.00 Copyright 0 1996 Elsevier Science B.V. All rights reserved

PII SOO22-1759(96)0009 1-9

82 J. Guo et al. / Jonrnal qflnlmlrnologiccrl Methods 195 f 1996) 81-92

Keywords: Autoantigen capture; Autoantigen presentation: CD23: Fcs receptor: Fab, human; IgE; T cell: Thyroid peroxidase; Thyroid peroxidase autoantibody

1. Introduction

Antigen focusing by specific membrane-bound antibodies on B lymphocytes is well recognized and may dominate presentation for antigen-specific T cells (reviewed in Lanzavecchia, 1990). In addition, secreted IgG class antibody complexed with antigen can be internalized and processed for T cell presenta- tion via binding to Fc y receptors (Fc~RI) on macrophages. However, the B lymphocyte isoform of the Fc y receptor (Fc y RII), produced by altema- tive mRNA splicing, contains a 42 amino acid inser- tion in the cytoplasmic tail that prevents FcrRII- mediated endocytosis and subsequent antigen pre- sentation (Amigorena et al., 1992).

Nevertheless, B lymphocytes have a second pow- erful antigen presenting system involving Fcs recep- tor (CD23)mediated capture, internalization and processing of IgE-antigen complexes (reviewed in Mudde et al., 1995). For example, hapten-specific IgE enhances N lOO-fold B cell presentation of hap- tenized carrier proteins to murine T cells (Kehry and Yamashita, 1989). Similarly, hapten-specific chimeric murine Fab/human IgE focuses haptenized antigen (tetanus toxoid or allergens) to Epstein-Barr virus transformed B lymphocytes (EBVL) for pre- sentation to human T cell clones (Santamaria et al., 1993; Pirron et al., 1990). Finally, serum IgE from atopic individuals facilitates allergen presentation by EBVL (Van der Heijden et al., 1993).

Recombinant human autoantibodies specific for thyroid peroxidase (TPO) have been generated by the immunoglobulin gene combinatorial library ap- proach using mRNA from intrathyroidal B cells (re- viewed in Rapoport et al., 1995). These autoantibod- ies, expressed as Fdb, closely resemble serum TPO autoantibodies in terms of IgG class and subclass, light chain type and high affinity ( - lo- “’ M K,) for TPO (Chazenbalk et al., 1993b; Portolano et al., 1992: and reviewed in McLachlan and Rapoport, 1995). The Fab define overlapping, conformational epitopes in two domains (A and B) on TPO (Chazenbalk et al., 1993a,b: Portolano et al., 1992). These epitopes constitute an immunodominant re-

gion because they are recognized by TPO autoanti- bodies in all patients and by > 80% TPO autoanti- bodies in an individual serum (Jaume et al., 1995a,b; Nishikawa et al., 1994). Consequently, these Fab are potentially valuable in studies of TPO presentation to human T cells.

In the present study, we describe the conversion of one of these TPO-specific Fab, SP1.4 (Portolano et al., 1992) to a full length IgE molecule secreted by mammalian cells in vectors previously used to express murine hybridoma V regions with human IgGl and K constant regions (Coloma et al., 1992). We show that this human TPO autoantibody con- verted to IgE class retains its high affinity for TPO as well as its original epitopic specificity. Most important. it mediates TPO Fee receptor (CD23) and specific T cell clone.

binding to EBVL via the presentation to a TPO-

2. Materials and methods

2.1. K light (L) chain expression vector containing the TPO Fab SP1.4 K chain variable (V) region

Recombinant TPO-specific Fab such as SP1.4 (Chazenbalk et al., 1993b) were generated by the immunoglobulin gene library approach with a prokaryotic signal peptide. In the absence of specific information for SP1.4, we used overlap PCR to splice the signal peptide from the human thyrotropin (TSH) receptor, a cDNA previously cloned and ex- pressed in our laboratory (Nagayama et al., 1989) to the first amino acid of the SPl.4 V K region (Fig. 1). The upstream primer introduced an EcoRV site at the 5’ end of the signal peptide. In order to maintain the intron between VL and CK, which is necessary to obtain L chain expression (Coloma et al., 1992) a Sal I site was introduced after the splice donor sequence downstream of SP 1.4 J K. After restriction, the DNA was ligated into the EcoRV and SalI sites in pAN462 1 (Fig. 1), an expression vector for human K L chains (Coloma et al., 1992), kindly provided by Dr. S. Morrison. The fidelity of the PCR products

J. Guo et al. /Journal of Immunological Merhods 195 (1996) 81-92 83

MR P ADLLQLVL

GATATCCACCATGAGGCCGGCGGAC‘TTGCTGCAGCTGGTGCTG

EcoRV L LDLPRDLGG

CTGCTCGACCTGCCAAGGGACCTGGGCGGA

TSHR VK JK

BamHl

PWI BamHl

Fig. 1. Construction of K light chain expression vector pAN462 I- SP-K containing the SP1.4 Fab V K and J K regions (Section 2).

The signal peptide (boxed) from the human thyrotrophin receptor

(TSHR) (Nagayama et al., 1989) was spliced to VK by overlap

PCR and inserted into the expression vector pAN4621 (Coloma et

al., 1992).

and ligation sites were confirmed by nucleotide se- quencing.

2.2. IgE heavy (H) chain expression vector contain- ing the TPO Fab SP1.4 H chain V region

As described for the K chain construct, overlap PCR was used to splice the human TSH receptor signal peptide on to the SP1.4 VH region. The oligonucleotide primers included EcoRV and NheI restriction sites at the 5’ and 3’ ends, respectively. After digestion with EcoRV and NheI, the DNA was ligated into the same sites in pAH4604, an expression vector for the human IgGl H chain (Col-

oma et al., 1992), also from Dr. S. Morrison. Expres- sion of this construct (pAH4604-SP-Gl) (Fig. 2) was confirmed by electroporation into J558L myeloma cells which secretes murine A L chains and screen- ing for expression of IgG, as previously described (Coloma et al., 1992). Subsequently, the BamHI site upstream of the EcoRV site in PAH4604-SP1.4 was removed by limited digestion, blunting of the ends with Klenow DNA polymerase and religation.

Plasmid DNA (2 pg pAH-SP-E and 20 p,g pAN4621-SP-K) were linearized with PuuI and co- transfected into 10’ SP2/0 cells by electroporation using an IBI gene zapper (950 pF, 200 V). Subse- quently, 100 ~1 aliquots of 2 X lo4 cells were plated in 96 well flat bottom culture dishes in Iscoves’ modified Dulbecco’s medium containing 10% fetal bovine serum (Gibco BRL, Grand Island, NY), 2.5 pg/ml fungizone, 50 p,g/ml gentamicin and 100 U/ml penicillin. Selection was initiated 48 h later by adding 100 pi/well medium containing G4 18 (Gibco BRL, 0.6 mg/ml). 2 days later, half the supematant was removed from each well and replaced with medium containing the same concentration of G418. The procedure was repeated after a further 48 h to achieve a final concentration of N 0.6 mg/ml G418. Colonies appeared after N 12- 16 days, at which time 100 ~1 aliquots of supematant were screened for IgE class TPO antibody by ELISA (see below). Cells from positive wells were expanded in 24 well plates and cloned by limiting dilution (0.5 cells/well).

2.4. ELISA for TPO antibody of IgE class

The IgE constant region was derived from the The technique was adapted from an ELISA previ- pAG4447 vector (also kindly provided by Dr. S. ously developed for IgG class TPO antibodies Morrison). The lack of suitable restriction sites pre- (Portolano et al., 1992). In brief, TPO in conditioned

cluded direct transfer of the constant region into PAH4604-SP-Gl. Instead, this aim was achieved in two stages. First, PCR was used to amplify CE I, the first 13 amino acids of CE II and the intervening intron. The 5’ oligonucleotide primer included an NheI site. The 3’ primer contained a BamHI site downstream of the natural BglII site. This N 0.6 kb DNA product was cloned into the NheI-BamHI sites of pAH4604-SP-Gl (Fig. 2). This intermediate con- struct was opened with BglII and BarnHI and the _ 2.6 kb BglII-BglII fragment of pAG4447 (con- taining the remainder of CE II, CE III, IV and intervening introns) was inserted (Fig. 2). Fidelity of the PCR amplified fragment and the correct orienta- tion of the insert were confirmed by nucleotide sequencing.

2.3. Expression ofSP1.4 H and L chain DNA encod- ing IgE class TPO autoantibody

84 J. Guo et al. / Jounml of Intnntnological Methods 195 (19961 81-92

culture medium from CHO cells (Kaufman et al., 19911 was used to coat ELISA plates (Immulon 4, Dynatech Laboratories, Chantilly, VA). Binding of IgE TPO antibody was detected using murine mono- clonal antibody to human IgE (Clone GE-l, Sigma Chemical Co., St. Louis, MO). The signal was devel- oped with affinity-purified anti-mouse IgG conju- gated to horseradish peroxidase (Sigma) and o-phen- ylene diamine + H,O, as substrate and optical den- sities (OD) were read at 492 nm. As control antibod- ies, we used murine monoclonals to human IgGl and IgG4 (NL16 and RJ4 respectively, Recognition Sci- ences, Birmingham, UK). Serum from a patient with autoimmune thyroid disease and a normal donor

were used to provide positive and negative signals respectively for binding by TPO autoantibodies of subclasses IgGl and IgG4.

2.5. “‘I-TPO binding by SPl.4.IgE anti-TPO

Recombinant TPO (Kaufman et al., 1991; Foti et al., 1990) was labeled with lz51 to a specific activity of - 50 pCi/ug by the iodogen method (Salacinski et al., 198 I ). Duplicate aliquots of IgE class anti- TPO, serially diluted in assay buffer (0.15 M NaCl containing 10 mM Tris-HCl pH 7.5 and 0.5% bovine serum albumin). were incubated with “51-TP0 (- 20000 cpm) and mouse monoclonal antibodies to

TSHR VU D JH

i Bglii

C&l CE2

\ BamHl PCR /

C&l C&2 (partId)

ml ~hel f3glll-EarnHI

Fig. 2. Construction of IgGl and 1gE heavy chain expression vectors (pAH4604-SP-G 1 and pAH-SP-E, respectively) containing the SPl.4 Fab VH, D and J regions. The TSH receptor signal peptide was spliced to SPl.4 VDJ by overlap PCR and inserted into the IgG I expression

vector pAH4604 (Coloma et al., 1992). Subsequently. C E 1 . 2, 3 and 4 and the intervening introns from pAG4447 were transferred in two stages to form pAH4604-SP-E (Section 2). pAG4447 was kindly provided by Dr. S. Morrison.

J. Guo et al. /Journal of Immunological Methods I95 (1996181-92 85

human immunoglobulins in a total volume of 200 ~1. The following monoclonals were used: anti-IgE (clone GE-I, Sigma); anti+ (QEI 1). anti-IgG1 (NL 16) and anti-IgG4 (RJ4) (Recognition Sciences). After 1 h at room temperature, 100 l.~l donkey anti-mouse Sac-ccl (IDS, Boldon, Tyne and Wear, UK) was added, and the incubation was continued for 30 min. Subsequently, 1 ml assay buffer was added, the tubes were vortexed and centrifuged for 30 min at 1000 X g to sediment the immune com- plexes which were then counted to determine the % radiolabeled TPO bound. Background binding ( * 5% total counts), determined using normal human serum diluted l/100, was subtracted from the values for SPl.4-IgE in calculating % ‘251-TP0 bound. This immunoprecipitation protocol was used to measure the affinity of SP1.4-IgE by Scatchard analysis (Scatchard, 1949) as previously described for TPO Fab SP1.4 (Portolano et al., 1992).

2.6. TPO epitope recognized by SPI.4-IgE

Because TPO Fab are not precipitated using anti- IgE, the immunoprecipitation assay described above was used to determine Fab inhibition of ‘251-TP0 binding by IgE anti-TPO. The assays were per- formed by incubating IgE anti-TPO, diluted to give N 10% ‘251-TP0 binding, in the absence or presence

of four different TPO-specific Fab (SP1.4, TR1.8, TR1.9 and WR1.7) (Chazenbalk et al., 1993b) at a final concentration of lo-’ M. TPO binding by serum from a patient, precipitated with protein A as previously described (Chazenbalk et al., 1993b), was used to provide a positive control.

2.7. Production and characterization of EBVL

Lymphocytes were isolated from lymph nodes draining the thyroid gland of a patient undergoing partial thyroidectomy for Graves’ hyperthyroidism. The lymphocytes were incubated for 2 h at 37°C with supernatant from the B95-8 marmoset line and subsequently cultured (N 3 X IO6 cells/2 ml) with phytohemagglutinin (15 p+l/ml, Gibco BRL) in RPM1 1640 containing 10% fetal calf serum, 2 mM L-glutamine (all from Gibco BRL), 250 p,g/ml fun- gizone, 50 pg/ml gentamicin and 100 U/ml peni- cillin. Expression of Fcs receptors on EBVL was

examined (10000 events) by flow cytometry using a FACScan (Becton Dickinson, San Jose, CA) and FITC-conjugated murine monoclonal anti-human CD23 (MHCD2301, Caltag Laboratories, San Fran- cisco CA). Controls included unstained cells and FITC-conjugated isotype control (IgG3, Pharmingen).

2.8. Binding of biotinylated TPO to EBVL

Recombinant human TPO (50 pg) was biotinyl- ated using the ECL protein biotinylation module (RPN 2202, Amersham International, Little Chal- font, Buckinghamshire, UK). To confirm that the protein retained antigenic activity, ELISA wells coated with monoclonal TPO antibody (SP1.4 IgGl) were exposed to biotinylated TPO. Binding was detected using streptavidin-horseradish peroxidase (Amersham International) followed by color devel- opment with substrate (o-phenylene diamine + H,O,). As a control, bovine serum albumin (BSA, 5 kg, Sigma) was biotinylated in the same way.

Complexes were produced by overnight incuba- tion at 4°C of culture supematant containing IgE anti-TPO (200 ~1 aliquots) with _ 2 +g biotinylated TPO, alone or together with unlabeled TPO (5 pg). Supematants containing IgE anti-TPO were incu- bated with biotinylated BSA (N 2 p,g) with or with- out cold BSA (5 kg) as controls. Subsequently, each anti-TPO/TPO mix (or anti-TPO/BSA mix) was added to 1.5 X lo6 EBVL for 1 h at 4°C. After washing three times with 2 ml buffer B (phosphate- buffered saline, 2% BSA), the EBVL were incubated (30 min, 4°C) with 1.5 p,g FITC-conjugated strepta- vidin (Caltag Laboratories). Following washing (3 ml, two times buffer B) and resuspension in 0.5 ml buffer B, the cells were analyzed by flow cytometry as described above.

2.9. IgE-mediated binding of ‘251-TP0 to EBVL

Culture supematant (400 ~1) containing SP1.4-IgE anti-TPO was incubated overnight at 4°C with radio- labled TPO (200000 cpm, N 5 ng) in the absence or presence of excess unlabeled TPO (1 Fg). To pro- vide controls, 400 ~1 aliquots of culture medium without IgE anti-TPO were incubated with ‘251-TP0 (200,000 cpm) alone or with unlabeled TPO. The following day, duplicate or triplicate aliquots of

86 .I. Guo et al. / Journal of Imnumological Methods 195 (1996) 81-92

EBVL (5 X IO6 cells) were resuspended in the anti- body/antigen mixtures, incubated for 2 h at 4°C washed twice with 3 ml buffer B (see above) and ‘251-TP0 bound to the pellets was determined by gamma counting. In some experiments, aliquots of EBVL were pre-incubated (1 h, 4°C) with anti-CD23 (MHCD2300, Caltag; 50 p.1/5 X lo6 cells), washed twice in 3 ml buffer B before incubation with IgE anti-TPO/radiolabeled TPO complexes.

2. IO. TPO-specific T cell clone and proliferation assay

T cell clone 36 from a Graves’ disease thyroid infiltrate was established as reported previously in detail (Londei et al., 1985). In brief, activated T cells were expanded for 1 week with recombinant inter- leukin-2. After a second week of expansion with

irradiated autologous peripheral blood mononuclear cells, OKT3 anti-CD3 and interleukin-2, cells were cloned by limiting dilution (Dayan et al., 1991; Londei et al., 1985). Further expansion of the T cell clone was achieved by restimulation every 2 weeks with irradiated allogeneic peripheral blood mononu- clear cells and phytohemagglutinin, or with OKT3 coated to plastic, with the addition of interleukin-2 at 4 day intervals. Culture medium was RPM1 1640 containing 10% pooled human serum, 100 U/ml penicillin and 50 kg/ml streptomycin.

The response of T cell clone 36 to TPO was tested in two ways at least 2 weeks after the last stimulation with phytohemagglutinin or 0KT3 and 5 days after the last addition of interleukin-2. First, 10” T cells were co-cultured with 3 X 10’ glutaralde- hyde-fixed autologous EBVL, stably expressing TPO, or (as controls) expressing chloramphenicol acetyl transferase (CAT) (Mullins et al., 1994). In the second approach, presenting cells were untransfected autologous EBVL pre-incubated in IgE-TPO com- plexes or appropriate controls. EBVL (5 X 10’) were incubated (3 h at 37°C) with SPl.4-IgE (2.5 ml) complexed with 25 p,g purified recombinant TPO, followed by washing and glutaraldehyde treatment (Dayan et al., 1991). In both types of proliferation assay, T cells and antigen-presenting cells were cul- tured in triplicate for 72 h in round bottom 96-well microtiter plates and then pulsed for 8 h with [ 3H]thymidine (1 l&i per well; Amersham, UK).

3. Results

3.1. IgE anti-TPO production by transfected SP2 / 0 rnyeloma cells

SP2/0 myeloma cells were co-tranfected with pAN462 I-SP-K and pAH-SP-E (Figs. I and 2). TPO autoantibodies were detectable by ELISA using murine monoclonal anti-human IgE in 15/88 wells. Two high level secretors were cloned and one was characterized. IgE specificity of anti-TPO autoanti- body in the culture supematant was confirmed by lack of recognition by antihuman IgGl or IgG4 (Fig. 3A). In contrast, serum TPO autoantibodies in a patient with autoimmune thyroid disease were de- tectable using anti-IgGl and anti-IgG4, but not anti- IgE (Fig. 3B).

3.2. Afinity and epitopic profile of SP1.4-IgE anti- TPO

Using either anti-k or anti-IgE, culture super- natants containing SPl .CIgE anti-TPO were capable of precipitating N 40% of added tracer “‘1-TPO (Table 1). This activity was reduced by serial dilu- tions of the culture supematant. In contrast, anti-IgGl and IgG4 precipitated only negligible amounts of tracer TPO. Unlabeled TPO could compete for SPl.4-IgE binding to ‘251-TP0 (Fig. 4). Scatchard analysis of these data indicated a TPO binding affin- ity (K,) of 2.20 + 0.15 X 10plo M, consistent with the value obtained for Fab SP1.4 (Portolano et al., 1992).

I~~~~!

2 J 8 (5 32 64 128 100 200 400 800 1500 3200 6400 DILUTION (Reaprocal) DILUTION (Reciprocal)

CULTURE SUPERNATANT SERUM

Fig, 3. Human autoantibody binding to TPO by ELBA. Autoanti-

bodies were detected using anti-IgE, anti-IgC I or anti-IgG4 (Set- tion 2). A: culture medium containing SP1.4-IgE serially diluted from l/2 to 1/ 128; B: serum (diluted l/ 100 to I /6400) from a

patient with autoimmune thyroid disease.

J. Guo et al./ Journal of Immunological Methods 195 (1996) 81-92 87

Table 1 Table 2

“‘I-TPO binding by serial dilutions of culture supematant con-

taining SPl .CIgE anti-TPO

SPl.4.IgE % “sI-TPO bound by SP1.4.IgE using

Epitopic identity of IgE SP1.4 and Fab (IgGl) SPl.4

Antibody Inhibition (%) of ‘*‘I-TPO

binding by Fab

Dilution Anti-K Anti-IgE Anti-IgG1/4

Neat 36.5 40.5 4.1 l/IO 26.9 29.3 _

l/30 23.3 14.9 _

l/l00 6.9 8.2 l/300 2.7 2.9 _

The epitopic profile of SP1.4-IgE anti-TPO was determined in competition studies using four recom- binant human IgG class Fab (Chazenbalk et al., 1993b) that define the immunodominant region on TPO for serum autoantibodies (Table 2). SPl.CIgE binding of 1251-TP0 was completely inhibited by Fab SP1.4 and WR1.7. Inhibition was partial with Fab TR1.8 but TR1.9 had no effect. These data are in complete agreement with the epitopic profile of the SP1.4 IgG class Fab (Chazenbalk et al., 1993b) which recognizes domain A on TPO (Chazenbalk et al., 1993b). Serum autoantibodies from some pa- tients (such as Hashimoto patient X, Table 2) also interact predominantly with domain A (Nishikawa et al., 1994). The functional activities of Fab TR 1.8 and TRl.9 in these assays were confirmed by their ability to completely inhibit tracer TPO binding by serum from a different patient (Y) with autoantibodies pri- marily to the TPO B domain (Table 2).

Fig. 4. Inhibition of SPl.CIgE binding of ‘*sI-TPO by increasing

concentrations of unlabeled TPO. Data shown are the mean * SE

of three separate experiments. In the absence of unlabeled TPO, radiolabeled TPO binding in the three experiments was 9.9%,

9.4% and 7.2%.

SPl.4 wR1.7 TRl.8 TR1.9

IgE SPl.4 93 99 47 -1

Fab SP1.4 a 80 94 36 2

Hashimoto serum X 87 79 53 7

Hashimoto serum Y 20 36 93 89

‘?sI-TPO binding by IgE SP1.4, immobilized Fab or Hashimoto

serum was determined in the absence and the presence of the

indicated TPO-specific free Fab (lo-* Ml. IgE SPl.4 was precipi-

tated using anti-IgE and Sac-eel and TPO autoantibodies in two

Hashimoto sera were precipitated with protein A. The ability of

the individual free Fab to inhibit TPO binding by Fab SP was

determined in a two-step assay in which Fab SP was first immobi-

lized using anti-a and Sac-ccl, as previously described (Chazen-

balk et al., 1993b). In the absence of free Fab, IgE SPl.4, Fab SP

and Hashimoto sera X and Y were diluted to provide 13%, 9%,

18% and 20% binding of the tracer TPO, respectively.

aData shown are as previously published for Fab SPl.5 (Portolano

et al., 19921, which has the identical epitope as SPl.4 (both have

the same heavy chain and very similar light chains).

3.3. Flow cytometric analysis of CD23 expression and @E-mediated TPO binding to EBVL

Expression of CD23 on EBVL was clearly evi- dent on flow cytometry (Fig. 5A). Binding of pre- formed complexes of SPl.CIgE and biotinylated TPO to EBVL was detectable, but only weakly and when gated on single cells (Fig. 5B). The specificity of this weak signal was established in two ways. First, by displacement of the peak with unlabeled TPO (Fig. 5B). Second, no shift in fluorescence was observed when the cells were incubated with bio- tinylated BSA, in the absence or presence of unla- beled BSA (Fig. 5C).

3.4. IgE-mediated binding of “‘I-TPO to EBVL uia CD23

In preliminary studies we observed that SP1.4 IgE recognition of 1251-TP0 was more sensitive than that of the biotinylated TPO used in the flow cytometry experiments (see above). We, therefore, re-examined capture by EBVL of SPl.4-IgE complexed to lZ51- TPO. In contrast to the flow cytometry studies, binding of these complexes to EBVL was much

J. Guo et al./Journcd oflnmunolo~icul Methods I95 (1996) 81-92

FLUORESCENCE INTENSITY

Fig. 5. A: FCE receptor (CD23) expression on EBVL analyzed by

flow cytometry. Black line, anti-CD23; gray line, IgG3 isotype

control. B: binding of preformed complexes of SP1.4-IgE and

biotinylated TPO to EBVL in the absence (gray line) and presence

(black line) of unlabeled TPO. C: binding of preformed com-

plexes of SPl.4.IgE and biotinylated BSA to EBVL in the ab-

sence (gray line) and presence (black line) of unlabeled BSA. In

B and C gating was restricted to single cells.

more clearly evident (Fig. 6). Again, the binding of complexes of SPl.4-IgE and labeled TPO was inhib- ited by the addition of unlabeled TPO. Further, this binding was greatly reduced by preincubation of the EBVL with antiXD23.

3.5. IgE-mediated TPO presentution to T cells

T cell clone 36 was first shown to be TPO-specific. Thus, T cell proliferation was observed in response to autologous EBVL stably expressing TPO antigen, but not to the same EBVL stably transfected with a plasmid expressing CAT (Fig. 7A). In subsequent

SP-IgE aCD23 TPO 012345678

+

+ - -

+ - +

+ _

+ + -

1251-TP0 (cpm x 1 O-3)

Fig. 6. FCC-mediated capture of complexes of SPI .4-IgE and

radiolabeled TPO by EBVL. EBVL were incubated with com-

plexes of “‘I-TPO and IgE SP1.4 (Section 2). Where indicated.

the complexes were preformed in the presence of 10-s M unla-

beled TPO. As controls, cells were exposed to “‘I-TPO, with or

without unlabeled TPO, in medium lacking SP1.4-IgE. Some

aliquots of EBVL were preincubated with antiKD23. The data

represent the mean and SE of triplicate determinations. Similar

data were obtained in a repeated experiment.

EEVL-TPO

ESVL-CAT + T CELLS

EBVL-TPO + T CELLS

EBVL + SPI 4-IgE

EBVL + SPI 4-l~G4

B

EEVL + WI 4-IQE + TPO + T CELLS

BH-THYMIDINE INCORPORATION (c.p m )

Fig. 7. Response of TPO-specific T cell clone 36 (Londei et al.,

1985) to TPO presented in two different ways. A: T cells were

co-cultured with autologous EBVL stably expressing TPO or

chloramphenicol acetyl transferase (CAT). B: T cells were co-cul- tured with untrunsfected autologous EBVL pre-incubated in either

SPI .4-IgE/TPO or SPl.4.IgG4. In both experiments, appropriate EBVL controls were cultured in the absence of T cells. Results are

expressed as the mean+SD of triplicate determinations of [‘Hlthymidine incorporation.

J. Guo et al. / Journal of Immunological Methods 195 (1996) 81-92 89

experiments, we demonstrated the ability of non- transfected autologous EBVL to present TPO to clone 36 following incubation in SPI.CIgE/TPO complexes (Fig. 7B). The IgE specificity of this response was evident by the inability of SP1.4- IgG4/TPO complexes to induce clone 36 prolifera- tion. SPl .CIgE and SPl.4-IgG4 (the construction and expression of the latter to be reported elsewhere) differ only in their heavy chain constant regions.

4. Discussion

The goal of this study was to develop a system using high affinity, human, organ-specific mono- clonal autoantibodies for autoantigen capture by B cells and presentation to T cells. If successful, this approach has the important potential for studying antibody-mediated antigen presentation in human au- toimmune disease. In principle, this goal could be attained by immortalizing IgG class autoantibody- specific B cells using EB virus. Indeed, this was the approach successfully used to demonstrate the im- portance of antibody-mediated capture and modula- tion of tetanus toxoid presentation to T cells (Lanzavecchia, 1985); and reviewed in (Lanzavec- chia, 1990). To date, however, very few stable EB transformed B cell lines have been produced that secrete organ specific autoantibodies (reviewed in Rapoport et al., 1995).

We report herein the conversion of a TPO-specific recombinant human Fab (SPl.4) to a full length IgE molecule expressed by mammalian cells. SPl.4-IgE is capable of capturing a human autoantigen, TPO, via the Fcs receptor on EBVL. Further, this capture can lead to antigen presentation to TPO-specific T cells. In previous studies, antigen focusing via CD23 was mediated by hapten-specific murine antibodies (Kehry and Yamashita, 1989), chimeric (human Fe/mouse Fab) anti-hapten antibodies (Santamaria et al., 1993; Pirron et al., 1990), or sera from atopic individuals (Van der Heijden et al., 1993). SPl .CIgE

is the first monoclonal human autoantibody to be developed for CD23 facilitated antigen capture and presentation by EBVL.

The immunoglobulin gene combinatorial ap- proach, which we used to generate TPO-specific Fab like SP1.4 (Chazenbalk et al., 1993b), involves

cloning in a vector which contributes a prokaryotic signal peptide. In a previous conversion of a recom- binant human Fab to a full length IgGl, leader sequences were selected from the VH and V K gene families from which the H and L chain genes were derived (Bender et al., 1993). However, H and L chain signal peptides vary considerably for different

germline genes (Kabat et al., 1991). Therefore, we used the signal peptide from a mammalian protein previously cloned and expressed in our laboratory, the human thyrotropin receptor (Nagayama et al., 1989). This approach was successful, demonstrating that immunoglobulin secretion does not require H or L chain specific signal peptides.

In order for SP1.4-IgE to be effective in antigen capture and processing, it is necessary that this molecule retain its high affinity. This, indeed, was the case. Thus, the affinity of SPl .CIgE was - 2 X lo- lo M (K,), in agreement with the value observed for the recombinant SP1.4 Fab (Portolano et al., 1992). Affinity is important both for the initial anti- body-mediated antigen capture, as well as for degra- dation following internalization and degradation. Only antibodies which bind with high affinity to the antigen are likely to remain complexed in the acidic endosomal compartment (reviewed in Lanzavecchia, 1990).

Antibodies can modulate antigen processing and thereby enhance or suppress presentation of different T cell determinants (e.g. Simitsek et al., 1995; Watts and Lanzavecchia, 1993; Davidson and Watts, 1989; Manta et al., 1988). Some human autoantibodies interact with a restricted region of the autoantigen, for example the immunodominant region on TPO (Chazenbalk et al., 1993b; Ruf et al., 1989b) and the major immunodominant region on the acetylcholine receptor 01 subunit (Heidenreich et al., 1988; Tzartos et al., 1982). It is possible that processing of high affinity autoantibody/autoantigen complexes, per- haps by exposing ‘cryptic epitopes’ (Lanzavecchia, 1995), is responsible for this restricted epitopic recognition. It was, therefore, important to confirm conservation of the epitopic specificity of SP 1.4-IgE. Indeed, different constant regions have been ob- served to alter epitopic recognition, at least for multi- valent antigens or different epitopic densities (Horgan et al., 1993; Cooper et al., 1993). Fortunately, SP1.4- IgE recognizes the same epitope as the IgGl class

90 J. Guo et al./ Journal of Immunological Methods I95 (19%) 81-92

Fab (Chazenbalk et al., 1993b). Thus, SP1.4-IgE binds to an epitope in domain A. Further, SPl.4-IgE binds to a different epitope than does Fab TRl.9. Because both molecules can bind simultaneously to TPO, TR1.9 could possibly be captured in ‘piggy back’ fashion (Simitsek et al., 1995) by SPl.4-IgE complexes bound to EBVL. Further experiments will be necessary to test this hypothesis.

Efficient IgE specific antibody-mediated allergen presentation to T cells has been suggested as a mechanism for lowering the threshold of atopic indi- viduals to mount allergen-specific T cell responses (Maurer et al., 1995; and reviewed in Mudde et al., 1995). Observations made with crude thyroid micro- somes, the principal component of which is TPO (reviewed in McLachlan and Rapoport, 19921, sug- gest that IgE class TPO autoantibodies are probably absent from patients with autoimmune thyroid dis- ease (Matsui et al., 1978). Therefore, IgE-mediated focusing of TPO to B cells must be viewed as a powerful investigative tool rather than as an in vivo mechanism. Nevertheless, B cells expressing mem- brane-bound IgG specific for thyroid autoantigens like TPO probably play an important role in antigen-uptake and presentation in vivo.

Several investigators have generated T cell lines or clones specific for TPO peptides (Kawakami et al., 1992; Tandon et al., 1991; Dayan et al., 1991). A recent comparison of T cell clones generated using full-length acetylcholine receptor CY subunit versus panels of peptides emphasizes the importance of ‘naturally’ processed peptides (Matsuo et al.. 1995). EBVL cell lines expressing TPO provide one ap- proach for generating naturally processed peptides to cell lines or clones (Mullins et al., 1994; Martin et al., 1993; McIntosh et al., 1993). However. anti- body-mediated uptake and presentation of thyroid autoantigens may not only influence the peptides presented but also permits B cells to capture and present minute amounts of antigen (reviewed in Lan- zavecchia. 1990).

In summary, recombinant human autoantibodies converted to IgE, possibly in combinations if their epitopes permit simultaneous binding to the same molecule, provide a unique system to generate hu- man T cell lines and clones specific for peptides naturally processed from internalized high affinity autoantibody/autoantigen complexes.

Acknowledgements

We thank Dr. Sheri Morrison for providing us with the vectors and Dr. Morrison and her colleagues for their advice on plasmid construction and expres- sion. This work was supported by NIH grants DK36182 (B.R.), DK48216 (S.M.M.), the Arthritis and Rheumatism Council (S.Q. and M.L.) and AISM (Associazione Italiana Sclerosi Multipla) (S.Q. and M.L.). J.C.J. is the recipient of a UCSF Molecular Medicine Training Program Fellowship Award.

References

Amigorena, S., Bonnerot. C., Drake, J.R.. Choquet, D., Hunziker,

W.. Guillet, J.-G.. Webster, P., Sautes, C., Mellman. I. and

Fridman. W.H. (1992) Cytoplasmic domain heterogeneity and

functions of IgG Fc receptors in B lymphocytes. Science 256,

1808.

Bender, E., Woof, J.M.. Atkin. J.D., Barker, M.D.. Bebbington.

C.R. and Burton, D.R. (1993) Recombinant human antibodies:

linkage of an Fab fragment from a combinatorial library to an

Fc fragment for expression in mammalian cell culture. Hum.

Antibod. Hybridomas 4, 74.

Chazenbalk. G.D., Costante. G.. Portolano, S.. McLachlan, S.M.

and Rapoport, B. fl993a) The immunodominant region on human thyroid peroxidase recognized by autoantibodies does

not contain the monoclonal antibody 47/c?] linear epitope. J.

Clin. Endocrinol. Metab. 77, 1715.

Chazenbalk. G.D., Portolano. S.. Russo, D., Hutchison. J.S.,

Rapoport. B. and McLachlan, SM. (1993bl Human organ-

specific autoimmune disease: molecular cloning and expres-

sion of an autoantibody gene repertoire for a major autoanti-

gen reveals an antigenic dominant region and restricted im-

munoglobulin gene usage in the target organ. J. Clin. Invest.

92. 62.

Coloma, M.J., Hastings, A., Wims, L.A. and Morrison, S.L.

(1992) Novel vectors for the expression of antibody molecules

using variable regions generated by polymerase chain reaction.

J. Immunol. Methods 152, 89.

Cooper. L.J.N., Shikhman. A.R.. Glass. D.D.. Kangisser, D..

Cunningham, M.W. and Greenspan, N.S. (1993) Role of

heavy chain constant domains in antibody-antigen interaction,

J. Immunol. 150, 2231. Davidson. H.W. and Watts, C. (1989) Epitope-directed processing

of specific antigen by B lymphocytes. J. Cell Biol. 109. 85.

Dayan, C.M., Londei. M.. Corcoran, A.E.. Grubeck-Loebenstein,

B., James, R.F.L.. Rapoport. B. and Feldmann. M. (1991) Autoantigen recognition by thyroid-infiltrating T cells in Graves disease. Proc. Nat]. Acad. Sci. USA 88. 7415.

Foti, D., Kaufman. K.D., Chazenbalk, G. and Rapoport, B. (1990) Generation of a biologically-active. secreted form of human thyroid peroxidase by site-directed mutagenesis. Mol. En-

docrinol. 4, 786.

.I. Guo et al. /Journal of Immunological Methods 195 (1996) 81-92 91

Heidenreich, F., Vincent, a.. Roberts, A. and Newsom-Davis, J. L., Vincent, A., Newsom-Davis, J. and Willcox, N. (19951

(1988) Epitopes on human acetylcholine receptor defined by Peptide-selected T cell lines from myasthenia gravis patients

monoclonal antibodies and myasthenia gravis sera. Autoimmu- and controls recognize epitopes that are not processed from

nity 1, 285. whole acetylcholine receptor. J.Immunol. 155, 3683.

Horgan, C., Brown, K. and Pincus, S.H. (1993) Studies on antigen

binding by intact and hinge-deleted chimeric antibodies. J.

Immunol. 150, 5400.

Jaume, J.C., Costante, G., Nishikawa, T., Phillips, D.I.W.,

Rapoport, B. and McLachlan, S.M. (1995a) Thyroid peroxi-

dase autoantibody fingerprints in hypothyroid and euthyroid

individuals. I. Cross-sectional study in elderly women. J. Clin.

Endocrinol. Metab. 80, 994.

Maurer, D., Ebner, C., Reininger, B., Fiebiger, E., Kraft, D.,

Kinet. J.-P. and Stingl, G. (1995) The high affinity IgE

receptor (FcERI) mediates IgE-dependent allergen presenta-

tion. J. Immunol. 154, 6285.

McIntosh, R.S.. Mulcahy, A.F., Hales, J.M. and Diamond, A.G.

(1993) Use of Epstein-Barr virus-based vectors for expression

of thyroid auto-antigens in human B-lymphoblastoid cell lines.

J. Autoimmun. 6, 353.

Jaume, J.C., Parkes, A.B., Lazarus. J.H., Hall. R., Costante, Cl.,

McLachlan, S.M. and Rapoport, B. (1995b) Thyroid peroxi-

dase autoantibody fingerprints. II. A longitudinal study in

postpartum thyroiditis. J. Clin. Endocrinol. Metab. 80, 1000.

Kabat, E.A.. Wu, T.T.. Perry, H.M., Gottesman, K.S. and Foeller,

C. (1991) Sequences of proteins of immunological interest.

U.S. Department of Health and Human Services. HIH Publica-

tion No. 91-3242.

McLachlan, S.M. and Rapoport, B. (1992) The molecular biology

of thyroid peroxidase: cloning. expression and role as autoanti-

gen in autoimmune thyroid disease. Endocrinol. Rev. 13, 192.

McLachlan, S.M. and Rapoport, B. (1995) Genetic and epitopic

analysis of thyroid peroxidase (TPO) autoantibodies: markers

of the human thyroid autoimmune response. Clin. Exp. Im-

munol. 101, 200.

Kaufman. K.D., Foti. D., Seto, P. and Rapoport, B. (19911

Overexpression of an immunologically-intact, secreted form of

human thyroid peroxidase in eukaryotic cells. Mol. Cell. En-

docrinol. 78. 107.

Kawakami. Y., Fisfalen, M.-E. and DeGroot, L.J. (1992) Prolifer-

ative responses of peripheral blood mononuclear cells from

patients with autoimmune thyroid disease to synthetic peptide

epitopes of human thyroid peroxidase. Autoimmunity 13, 17.

Kehry, M.R. and Yamashita, L.C. (1989) Low-affinity IgE recep-

tor (CD23) function on mouse B cells: Role in IgE-dependent

antigen focusing. Proc. Natl. Acad. Sci. USA 86, 7556.

Lanzavecchia. A. (1985) Antigen-specific interaction between T

and B cells. Nature 314, 537.

Mudde, G.C., Bheekha, R. and Bruijnzeel-Koomen, C.A.F.M.

(19951 Consequences of IgE/CD23_mediated antigen pre-

sentation in allergy. Immunol. Today 16, 380.

Mullins. R.J., Chernajovsky, Y., Dayan, C., Londei, M. and

Feldmann. M. (1994) Transfection of thyroid autoantigens into

EBV-transformed B cell lines. J. Immunol. 152, 5572.

Nagayama, Y.. Kaufman, K.D., Seto, P. and Rapoport. B. (1989)

Molecular cloning, sequence and functional expression of the

cDNA for the human thyrotropin receptor. Biochem. Biophys.

Res. Commun. 165, 1184.

Lanzavecchia. A. (1990) Receptor-mediated antigen uptake and its

effect on antigen presentation to class II-restricted T lympho-

cytes. Ann. Rev. Immunol. 8, 773.

Lanzavecchia. A. (19951 How can cryptic epitopes trigger autoim-

munity? J. Exp. Med. 181. 1945.

Nishikawa. T., Costante. G., Prummel, M.F., McLachlan, S.M.

and Rapoport, B. (1994) Recombinant thyroid peroxidase au-

toantibodies can be used for epitopic ‘fingerprinting’ of thy-

roid peroxidase autoantibodies in the sera of individual pa-

tients, J. Clin. Endocrinol. Metab. 78. 944.

Pirron. U.. Schlunck, T., Prim, J.C. and Rieber. E.P. (1990)

IgE-dependent antigen focusing by human B lymphocytes is

mediated by low-affinity receptor for IgE. Eur. J. Immunol.

20, 1547.

Londei. M., Bottazzo. G.F. and Feldmann, M. (1985) Human T

cell clones from autoimmune thyroid glands: Specific recogni-

tion of autologous thyroid cells. Science 228. 85.

Manta. F., Fenoglio, D.. Kunkl. A., Cambiaggi, C., Sasso, M. and

Celada. F. (19881 Differential activation of T cell clones

stimulated by macrophages exposed to antigen complexed

with monoclonal antibodies. A possible influence of paratope

specificity on the mode of antigen processing. J. Immunol.

140, 2893.

Portolano. S., Chazenbalk, G.D., Seto, P., Hutchison, J.S.,

Rapoport, B. and McLachlan, S.M. (1992) Recognition by

recombinant autoimmune thyroid disease-derived Fab frag-

ments of a dominant conformational epitope on human thyroid

peroxidase. J. Clin. Invest. 90. 720.

Martin. A., Magnusson, R.P., Kendler, D.L., Conception, E.,

Ben-Nun, A. and Davies, T.F. (1993) Endogenous antigen

presentation by autoantigen-transfected Epstein-Barr virus-

lymphoblastoid cells. 1. Generation of human thyroid peroxi-

dase-reactive T cells and their T cell receptor repertoire. J.

Clin. Invest. 91, 1567. Matsui, Y., Heiner, D.C. and Beall, G.N. (1978) IgE and IgE

autoantibodies in patients with autoimmune thyroid disorders

and their relatives. Proc. Sot. Exp. Biol. Med. 158. 73.

Matsuo, H., Batocchi, A.-P.. Hawke, S., Nicolle, M., Jacobson,

Rapoport, B., Portolano. S. and McLachlan, S.M. (1995) Combi-

natorial immunoglobulin gene libraries: new insights into hu- man organ-specific autoantibodies. Immunol. Today 16, 43.

Ruf. J.. Toubert. M.. Czamocka, B., Durand-Gorde. J., Ferrand,

M. and Carayon. P. (1989) Relationship between immunologi-

cal structure and biochemical properties of human thyroid

peroxidase. Endocrinology 125, 1211.

Salacinski, P.R.P., McLean. C., Sykes, J.E.C., Clement-Jones,

V.V. and Lowry, P.J. (1981) Iodination of proteins, glyco-

proteins. and peptides using a solid-phase oxidizing agent, 1,3,4,6-tetrachloro-3 alpha.6 alpha-diphenyl glycoluril (Iodo- gen). Anal. Biochem. 117, 136.

Santamaria, L.F., Bheekha, R., van Reijsen, F.C., Perez Soler. M.T., Suter. M., Bruijnzeel-Koomen, C.A.F.M. and Mudde,

92 J. Guo et al./ Journal of Immunological Methods I95 (1996) 81-92

G.C. (1993) Antigen focusing by specific monomeric Taartos. S.J., Seybold. M.E. and Lindstrom, J.M. (19821 Specifici-

immunoglobulin E bound to CD23 on Epstein-Barr virus- ties of antibodies to acetylcholine receptors in sera from

transformed B cells. Hum. Immunol. 37. 23. myasthenia gravis patients measured by monoclonal antibod-

Scatchard. G. (1949) The attractions of proteins for small ies. Proc. Natl. Acad. Sci. USA 79, 188. molecules and ions. Ann. NY Acad. Sci. 51. 660. Van der Heijden, F.L.. Joost van Neerven, R.J.. van Katwijk. M.,

Simitsek, P.D., Cmpbell, D.G., Lanzavecchia, A., Fairweather, N. Bos, J.D. and Kapsenberg, M.L. (19931 Serum-IgE-facilitated

and Watts, C. (19951 Modulation of antigen processing by allergen presentation in atopic disease. J. Immunol. 150. 3643. bound antibodies can boost or suppress class II major histo- Watts, C. and Lanaavecchia, A. (19931 Suppressive effect of compatibility complex presentation of different T cell determi- antibody on processing of T cell epitopes. J. Exp. Med. 178,

nants. J. Exp. Med. 181, 1957. 1459.

Tandon, N., Freeman, M. and Weetman, A.P. (1991) T cell

responses to synthetic thyroid peroxidase peptides in autoim-

mune thyroid disease. Clin. Exp. Immunol. 86, 56.