Structural Insights into Autoreactive Determinants in Thyroid ...

1 downloads 0 Views 404KB Size Report
Sep 7, 2006 - ... Danuta Plochocka, Grazyna Hoser, E. Helen Kemp, Brian J. Sutton, ..... Dickinson, Franklin Lakes, NJ) with CellQuest software (Becton Dick-.
0013-7227/06/$15.00/0 Printed in U.S.A.

Endocrinology 147(12):5995– 6003 Copyright © 2006 by The Endocrine Society doi: 10.1210/en.2006-0912

Structural Insights into Autoreactive Determinants in Thyroid Peroxidase Composed of Discontinuous and Multiple Key Contact Amino Acid Residues Contributing to Epitopes Recognized by Patients’ Autoantibodies Marlena Dubska, J. Paul Banga, Danuta Plochocka, Grazyna Hoser, E. Helen Kemp, Brian J. Sutton, Andrzej Gardas, and Monika Gora Department of Biochemistry and Molecular Biology (M.D., A.G., M.G.), and Department of Clinical Cytology (G.H.), Medical Centre of Postgraduate Education, 01-813 Warsaw, Poland; Division of Gene and Cell-Based Therapy (J.P.B.), King’s College London School of Medicine, London SE5 9PJ, United Kingdom; Department of Bioinformatics (D.P.), and Department of Genetics (M.G.), Institute of Biochemistry and Biophysics PAS, 02-106 Warsaw, Poland; Division of Clinical Sciences (North) (E.H.K.), University of Sheffield, Northern General Hospital, Sheffield S5 7AU, United Kingdom; and The Randall Division of Cell and Molecular Biophysics (B.J.S.), King’s College London SE1 1UL, United Kingdom Thyroid peroxidase (TPO) is a major autoantigen of thyroid autoimmune disease, and the autoantibodies that are produced recognize two immunodominant regions (IDR) of the molecule, termed IDR-A and -B. Based upon our structural model of the TPO ectodomain, we recently identified R225 and K627 as key residues in IDR-A and -B, respectively. We report here on rational mutagenic investigations to identify additional residues surrounding R225 and K627 that affect the binding of recombinant human Fabs (rhFabs) specific for each IDR. Two residues R646 and D707 were identified from the model as promising surface-exposed amino acids adjacent to R225. Similarly, residues E604, D620, D624, and D630 were identified in the vicinity of K627. These residues were substituted in different combinations of single, double, and multiple mutations, and stably expressed in Chinese hamster ovary

A

UTOIMMUNE THYROID DISEASES (AITD) encompass destructive Hashimoto’s thyroiditis, which leads to hypothyroidism, and Graves’ disease, which leads to hyperthyroidism. A large number of AITD patients are positive for high-titer autoantibodies to thyroid proteins, in particular thyroid peroxidase (TPO) (reviewed in Ref. 1). Recently, many findings support the hypothesis that antibodies against TPO are responsible for the autoimmune destruction of thyrocytes, either by fixing complement or through cell mediated cytotoxicity (reviewed in Refs. 2 and 3). It has been also shown that the complement pathway may be directly activated by component C4 binding to TPO itself (4). Furthermore, induction of autoimmunity in mice immunized with TPO fragments and whole protein (5–7) adds a new

First Published Online September 7, 2006 Abbreviations: AITD, Autoimmune thyroid diseases; CCP, complement control protein; CHO, Chinese hamster ovary; EGF, epidermal growth factor; FACS, fluorescence-activated cell sorting; HRP, horseradish peroxidase; hTPO, human TPO; IDR, immunodominant region; mAb, monoclonal antibody; MPO, myeloperoxidase; rhFab, recombinant human Fab; TPO, thyroid peroxidase. Endocrinology is published monthly by The Endocrine Society (http:// www.endo-society.org), the foremost professional society serving the endocrine community.

cells. By fluorescence-activated cell sorting and capture ELISA, we found that R225A, R646A, and D707N specifically led to the loss of binding of IDR-A rhFabs, whereas E604A, D620R, K627G, and D630N specifically abrogated the binding of IDR-B rhFabs. Further supportive evidence of the importance of these residues for the IDR epitopes was obtained with patients’ sera. We conclude that R646 and D707 together with R225 constitute a functional epitope within IDR-A, and that residues E604, D620, and D630, together with K627, constitute a functional epitope within IDR-B. This identification of key residues within the autoreactive epitopes will help in understanding the structural basis for the breakdown of immune tolerance to TPO in thyroid autoimmune disease. (Endocrinology 147: 5995– 6003, 2006)

dimension to understanding thyroid cell destruction mediated by TPO. The majority of the autoimmune response to TPO is directed toward two discontinuous surface determinants on the molecule (reviewed in Refs. 2 and 3). Competition experiments with a panel of murine monoclonal antibodies (mAb) and patients’ autoantibodies defined immunodominant region-A (IDR-A) and immunodominant region-B (IDR-B) (8), and subsequent studies with recombinant human anti-TPO Fab fragments confirmed these findings (9, 10). It is confusing however, that the terminology has been inverted: IDR-A recognized by murine mAbs is identical with IDR-B recognized by human Fab fragments, and vice versa. The terminology used in this article refers to the original IDR definitions of Ruf et al. (8). Interestingly, there are no statistically significant differences between Graves’ and Hashimoto patients’ sera in the proportion of autoantibodies to the IDR-A and IDR-B determinants (11). TPO is the central enzyme in the biosynthesis of thyroid hormone by the thyroid gland, catalyzing the iodination and coupling of tyrosine residues in thyroglobulin that leads to the synthesis of T3 and T4 (12). Human TPO (hTPO) consists of 933 amino acid residues and encom-

5995

5996

Endocrinology, December 2006, 147(12):5995– 6003

passes the extracellular region (ectodomain) comprising 848 residues, the transmembrane domain and the short cytoplasmic region. Three regions in the TPO ectodomain exhibit a high degree of sequence homology to myeloperoxidase (MPO-like domain), complement control protein (CCP-like domain), and epidermal growth factor (EGFlike domain), respectively (13). Epitope mapping of autoantigens is valuable for understanding the mechanisms that trigger autoimmunity and to guide the design of possible therapeutic agents. Several attempts have been made to map the two immunodominant regions on TPO (reviewed in Refs. 2, 3, and 14). Assays of proteolytic peptides generated by enzymatic hydrolysis of TPO have provided support for the hypothesis that residues of the MPO-like, CCP-like, and EGF-like domains contribute to IDR-A and -B determinants in the native TPO quaternary structure (15, 16). Using peptide phage-displayed libraries, two regions in the MPO-like domain and one region in the CCPlike domain were identified as sites of autoantibody binding (17). However, other studies produced different results and showed that deletions of the CCP-like and EGF-like domains did not significantly influence patients’ autoantibody binding (18), nor change the reaction with human recombinant Fab fragments (19). Determination of the three-dimensional structure of TPO would clearly assist interpretation of these data, but reported crystals of TPO have diffracted only to low resolution (20, 21). We have constructed a molecular model of the TPO ectodomain based on the known structure of myeloperoxidase (22) because the two proteins have 46% identical and 71% similar amino acid residues. Examination of the model led us to select sequences exposed on the TPO surface, and we then raised antisera to these synthesized peptides. Studies with these antibodies identified a region encompassing both IDR-A and -B epitopes, defined by three peptides (22– 24). Subsequently, by site-directed mutagenesis, we identified the residues R225 and K627 as key components of IDR-A and IDR-B, respectively (25). At the same time, Bresson et al. (26) assigned the residues 713–717, recognized by murine mAb 47, as part of IDR-A. Using peptide spot technology the sequences 597– 604 and 611– 618 in TPO were identified as targets for IDR-〉-specific autoantibodies, either expressed as recombinant Fab or obtained from Graves’ disease patients (27). More recently, residues 353–363 of TPO have also been shown to be important for the binding of autoantibodies (28). In this study, we extend the mapping of the IDR-A and IDR-B determinants in TPO by molecular modeling and sitedirected mutagenesis. The identification of R225 and K627 allowed us to propose additional amino acid residues that might contribute to the IDRs, and we now report that R225, R646, and D707 are key residues for IDR-A-specific of recombinant human Fab (rhFab) binding, and that E604, D620, K627, and D630 are key residues for IDR-〉-specific rhFab binding. This provides the most detailed description yet of these immunodominant epitopes and identifies individual residues that are critical not only for rhFab binding, but also for recognition by autoantibodies in patients’ sera.

Dubska et al. • Autoreactive Residues on TPO

Materials and Methods Single and multiple mutant TPO cDNA and their expression in Chinese hamster ovary (CHO) cells Single mutants R225A, K627G, and D624S used in this study have been described in our earlier work (24). Figure 1 shows the positions of mutations and the restriction sites used in the construction of the TPO mutants. Mutants in the vicinity of R225 (R225A/R646A, R225A/ D707N, R225A/R646A/D707N) and mutants in the vicinity of K627 (E604A, K627G/E604A, K627G/E604A/D620R/D624S, K627G/E604A/ D620R/D624S/D630N) were generated using Altered Sites in Vitro Mutagenesis System (Promega, Madison, WI). Mutagenic oligonucleotides were designed by introducing or removing appropriate restriction sites (Table 1). The hTPO cDNA carrying R225A or K627G substitutions, subcloned into pALTER1 (24), served as a template in the following reactions: R225A template plus a646 oligo to give the mutant R225A/ R646A; R225A template plus a707 oligo (R225A/D707N); R225A template plus a646 and a707 oligos (R225A/R646A/D707N); K627G template plus a604 oligo (K627G/E604A); K627G template plus a604 and a620/624 oligos (K627G/E604A/D620R/D624S). The resulting hTPO cDNA carrying K627G/E604A/D620R/D624S substitutions in pALTER1 served as a template in the subsequent mutagenesis reaction with a630 oligo to obtain the multiple mutant K627G/E604A/D620R/D624S/ D630. The single E604A substitution was generated on the pALTER1 vector carrying the wild-type hTPO cDNA (24) in the reaction with a604 oligonucleotide. Mutations in the vicinity of K627 were introduced into the pcDNA5/FRT vector carrying the wild-type hTPO cDNA by exchanging the BamHI-ClaI fragment. The pCIneo vector (Promega) in which the internal SacI restriction site has been deleted, was used to introduce mutations in the vicinity of R225. Mutations R225A/R646A, R225A/D707N, and R225A/R646A/D707N were introduced into the full-length hTPO cDNA in this vector by exchanging the Eco47III-SacI fragment. The full-length hTPO cDNA carrying the introduced mutations was then cloned into pcDNA5/FRT (Invitrogen, Carlsbad, CA) between NheI and NotI sites. Single mutants R646A, D707N and the double mutant R646A/D707N were constructed by exchanging the ClaINotI fragment from the wild-type hTPO cDNA in pcDNA5/FRT with the corresponding fragments from previously generated R225A/R646A, R225A/D707N and R225A/R646A/D707N mutants. Single mutants D620R, D630N, and the double mutant D620R/D624S were generated by an overlap extension PCR (29), with pairs of internal complementary oligonucleotides for the desired mutation (Table 1) and two external oligonucleotides (Table 1, ext1 and ext2). The hTPO cDNA in pcDNA3.1 (5) was used as a template in the first round of PCR. The resulting PCR products carrying D620R and D620R/D624S mutations were cut with BamHI and ClaI, and were introduced into the hTPO cDNA in pcDNA5/ FRT by swapping the BamHI-ClaI fragment. This approach could not be used in the case of the D630N substitution as the ClaI restriction site had been removed during mutagenesis. Therefore, the pGEMzf(⫹) vector (Promega), in which the SacI restriction site has been removed, was used to subclone the BamHI-NotI fragment of the wild-type hTPO cDNA, followed by introduction of D630N by exchanging the BamHI-SacI fragment. Finally, the BamHI-NotI fragment carrying the D630N substitution was exchanged with the corresponding fragment from the hTPO cDNA

FIG. 1. Schematic representation of the TPO cDNA. The diagram represents the coding sequence of TPO (blank box) and the positions of mutations constructed in this study (shown in bold). The noncoding sequence is shown as the horizontal line. The restriction sites used in the construction of the TPO mutants are indicated. Details of these constructs are described in Materials and Methods. aa, Amino acids.

Dubska et al. • Autoreactive Residues on TPO

Endocrinology, December 2006, 147(12):5995– 6003

5997

TABLE 1. Oligonucleotides used in construction of TPO mutants Name

a604 a620/624 a630 a646 a707 f620 r620 f620/624 r620/624 f630 r630 ext1 ext2

Restriction sitea

Sequence 5⬘–3⬘ b

GTCAGCGGGGGTAGCTAGCCGAGGCAGGCCG GGTGACCGTACAAGGACAAGATCTTGCGGGCCACGCTCCTGC GACATCGATGTTGTTAGGATGACCGTACAAG GCCCTGTCCGAGCTGCAGGGAGGAAGTTTTC GACTTGGAAGGCGTTCATGGGCACCCTG CTTGTACAAGTCCAAGATCTTGCGGGCCACGCTCCT AGGAGCGTGGCCCGCAAGATCTTGGACTTGTACAAG ATGCTTGTACAAGGACAAGATCTTGCGGGCCACGCTCCTG CAGGAGCGTGGCCCGCAAGATCTTGTCCTTGTACAAGCAT CAGCCAGACATCAATATTGTTAGGATGCTTGTAC GTACAAGCATCCTAACAATATTGATGTCTGGCTG TCAATGCGCACTGGAGCG CCGCCTGTCTCCGAGATG

NheI introduced BglII introduced BstEII removed PstI introduced NcoI removed BglII introduced BglII introduced BglII introduced BglII introduced SspI introduced, ClaI removed SspI introduced, ClaI removed

f, Forward oligonucleotide; r, reverse oligonucleotide; ext1 and 2, two external nucleotides. Introduced sites are underlined in the oligonucleotide sequence. b Mutated nucleotides are in boldface type. a

in pcDNA5/FRT. All mutants were sequenced to verify the presence of the desired mutations and to ensure that no other mutations were introduced. Transfection and expression of TPO mutants in CHO cells using the Flp-In expression system was previously described (25). Cell lines resistant to 600 ␮g/ml Hygromycin B were propagated for stable expression. Mutants R225A, D707N, and R646A/D707N were further subcloned by isolation of single cells because preliminary studies on mutant TPO expression showed that these three mutants were mixed populations of expressing and nonexpressing cells.

Antibodies Recombinant human Fabs (rhFabs) specific for IDR-A (126TP1 and 126TO10) and specific for IDR-B (126TP5 and 126TP14) were obtained and characterized as described (9, 30). Analysis of epitope specificity was carried out on rhFabs prepared in culture medium. Sera from patients with thyroid autoimmune disease were obtained from the Regional Hospital Siedlce (Siedlce, Poland). Pooled sera from 20 patients with thyroid autoimmune disease, positive for TPO antibodies, were used as a positive control. Autoantibodies to TPO were measured by ELISA, standardized to the World Health Organization/Medical Research Council international standard 66/387 (22). Murine mAb A4 directed to a linear epitope on TPO (31) was used in capture ELISA and Western blotting.

Immunodetection of TPO The production of recombinant TPO proteins was confirmed by Western blot analysis. Membrane proteins were extracted from stably transfected CHO cells as described (25). Samples were subjected to SDSPAGE under reducing conditions and electrophoretically transferred to nitrocellulose membranes. The membranes were probed with anti-TPO mAb A4 followed by HRP-conjugated rabbit antimouse IgG (Dako, Carpinteria, CA), and developed using chemiluminescence SuperSignal West Pico Substrate System (Pierce, Rockford, IL). The amount of protein expressed by different CHO transfectants was standardized by densitometry.

Flow cytometry analysis Cell surface expression of mutant TPO was analyzed by flow cytometry with anti-TPO-specific rhFabs (1:50 dilution) and a pool of 20 sera from patients with thyroid autoimmune disease (1:30 dilution). Seventy to 80% confluent monolayer cultures of CHO cells expressing wild-type or mutant TPO were detached from culture plates with Cell Dissociation Solution (Sigma, St. Louis, MO) and washed with PBS containing 0.1% BSA. Cells (2 ⫻ 105) were incubated for 30 min on ice with specific primary antibodies, washed, and incubated for 30 min on ice with antihuman ␬ R-phycoerythrin conjugate (Caltag, Burlingame, CA) or antihuman IgG fluorescein isothiocyanate conjugate (Sigma). After washing, cells were fixed in 2% formaldehyde prepared in PBS and

analyzed (10,000 events) on a FACSCalibur flow cytometer (Becton Dickinson, Franklin Lakes, NJ) with CellQuest software (Becton Dickinson). The binding of all antibodies was analyzed at least three times.

Capture ELISA The binding of rhFabs and individual patients’ sera to mutant TPO was tested by capture ELISA as previously described (25). Briefly, the wells of microtitre plates (Costar, Corning, NY) were coated with mAb A4 at 20 ␮g/ml in carbonate-bicarbonate buffer/0.1% sodium azide and were incubated overnight at 4 C. After each incubation, wells were washed three times with TBS containing 0.1% Tween 20 (TBST). After blocking by incubation with TBS containing 2 mg/ml BSA, membrane proteins (10 ␮g/ml) extracted from CHO cells expressing TPO were captured onto mAb A4-coated wells in TBST containing 2 mg/ml BSA (TBST-BSA). Anti-TPO-specific rhFabs or individual patients’ sera were diluted in TBST-BSA to give an OD at 450 nm of between 1.0 and 2.0 and incubated for 1 h at room temperature. The bound Fab fragments or patients’ antibodies were detected after incubation for 1 h at room temperature with horseradish peroxidase (HRP)-conjugated goat antihuman IgG Fab-specific (Sigma) or HRP-conjugated rabbit antihuman IgG (Sigma), respectively. The plates were developed with solution of tetramethylbenzidine (0.1 mg/ml) in 0.1 m citric-phosphate buffer (pH 4.0), the reaction was stopped by adding 1 m sulfuric acid and optical density was measured at 450 nm. All experiments were repeated at least three times, each in triplicate. Background binding to membrane proteins prepared from untransfected CHO cells was subtracted from the antibody binding to proteins extracted from CHO cells expressing wildtype and mutant TPO.

Modeling of the TPO structure The three-dimensional model of TPO was constructed as described previously (22) and was regenerated for this study. hTPO has almost 50% sequence identity with human MPO, the structure of which has been determined by x-ray crystallography (32, 33) (PDB entries: 1MHL and 1CXP). The monomer of MPO consists of two polypeptide chains resulting from posttranslational excision of the A107-G112 fragment (numbering according to Ref. 32). Atomic coordinates of the TPO threedimensional models were obtained from the SWISS-MODEL server (34) for fragments T145-T248 and A257-T735, which were aligned with C1A104 (light chain) and V113-A578 (heavy chain), respectively. Missing loops were inserted using the HOMOLOGY module in INSIGHT II (version 2005; Accelrys, San Diego, CA). To check the effect of the introduced mutations on the protein structure, model structures of the wild-type and mutated proteins were subjected to energy minimization using consistent valence force field as implemented in DISCOVER version 2004.1 (Accelrys). Minimizations were performed in vacuo. Surface areas of selected patches were calculated using WHAT IF program (35).

5998

Endocrinology, December 2006, 147(12):5995– 6003

Results Construction and expression of TPO mutants

Our previous studies on TPO mutants identified two amino acid residues, R225 and K627, as key components of the dominant autoreactive IDR-A and IDR-B epitopes, respectively (25). To discover additional amino acid residues that contribute to IDR-A, we selected on our structural model two surface-exposed residues within the vicinity of R225. Single or double mutations were performed at positions 646 and 707: R646A, D707N, and R646A/D707N. Furthermore, R646A and D707N substitutions were combined with the previously reported R225A mutation in the following variants: R225A/R646A, R225A/D707N, R225A/R646A/D707N (RRD). The same approach was applied to characterize the extent of the IDR-B region. Three single mutations were generated at positions 604, 620, and 630 (E604A, D620R, D630N). They were also combined with previously reported K627G and D624S substitutions in double and multiple mutants: K627G/E604A, D620R/D624S, K627G/E604A/ D620R/D624S (KDDE) and K627G/E604A/D620R/D624S/ D630N (KDDED). To avoid major changes in the tertiary structure of the protein, all the selected amino acids were substituted by the homologous residues in the MPO sequence or by alanine. All mutants were stably expressed in CHO cells as surface membrane proteins (25). Membrane extracts from recombinant CHO cells were analyzed for their TPO content by immunodetection with anti-TPO murine mAb A4 (Fig. 2). Densitometry analysis of immunoblots revealed that all the mutant proteins were produced essentially at the wild-type TPO level (100 ⫾ 19%). Mutations that disrupt the native folding of the protein usually cause its degradation by proteases in vivo, and thus we assume that the substitutions introduced in TPO did not interfere with protein folding.

FIG. 2. Western blot analysis of mutant TPO proteins expressed in CHO cells. Cell lysates were prepared by Triton X-100 solubilization and 5 ␮g proteins/lane were subjected to 8% SDS-PAGE, transferred to nitrocellulose membrane and probed with anti-TPO murine monoclonal antibody A4. A, Mutants in the IDR-A epitope, lane 1, control sample expressing wild-type TPO, lane 2, R225A mutant, lane 3, R646A mutant, lane 4, D707N mutant, lane 5, R225A/R646A mutant, lane 6, R225A/D707N, lane 7, R646A/D707N, lane 8, R225A/R646A/ D707N (RRD). B, Mutants in the IDR-B epitope, lane 1, control sample expressing wild-type TPO, lane 2, E604A mutant, lane 3, D620R mutant, lane 4, D624S mutant, lane 5, K627G mutant, lane 6, D630N, lane 7, K627G/E604A, lane 8, D620R/D624S, lane 9, K627G/E604A/ D620R/D624S (KDDE) mutant, lane 10, K627G/E604A/D620R/ D624S/D630N (KDDED) mutant.

Dubska et al. • Autoreactive Residues on TPO

Fluorescence-activated cell sorting (FACS) analysis of the binding of rhFabs to TPO mutants expressed on the cell surface

Flow cytometry was used to determine cell surface exposure and reactivity of the mutant TPO proteins (Fig. 3). The mean fluorescence intensity is indicated for each mutant. The positive reaction with human anti-TPO serum pool showed that all mutant proteins were expressed on the cell surface and the level of expression is comparable to wild-type TPO. Moreover, binding of rhFab TP14 specific for the IDR-B was not substantially affected in the case of mutants in the IDR-A domain (Fig. 3A, middle column). On the contrary, binding of rhFab TO10 specific for the IDR-A epitope was not influenced when mutants in the IDR-B epitope have been tested (Fig. 3B, first column). These internal controls confirmed that introduced mutations did not influence the conformational integrity of the mutants for the binding of rhFabs for the alternative IDR determinant. Single mutations R646A and D707N did not cause a substantial effect on binding rhFab TO10 specific for the IDR-A epitope (Fig. 3A, first column). However, introducing these two mutations together in the mutant R646A/D707N reduced approximately 3-fold the mean fluorescence intensity. Surprisingly, the substitution of R225, previously identified as important for IDR-A, led only to slightly (⬃2-fold) decreased binding of TO10. But double mutants R225A/R646A and R225A/ D707N caused approximately 4- to 6-fold reduction of the mean fluorescence intensity relative to the wild-type TPO value. The most dramatic effect was observed in binding of rhFab TO10 to the triple mutant R225A/R646A/D707N: the mean fluorescence intensity was approximately 55-fold lower than the normal value, whereas the recognition by rhFab TP14 was not affected (Fig. 3A; RRD mutation). All these results indicated that residues R646 and D707 contribute to the IDR-A epitope together with the residue R225. In the case of candidate residues within IDR-B epitope (Fig. 3B), none of the single substitutions influenced binding of rhFab TP14 specific for the IDR-B region, even K627G previously identified as a component of IDR-B (25). However, the double replacement K627G/E604A caused approximately 2.5-fold reduction of rhFab TP14 binding, indicating an additive effect of these two mutations. The double D620R/D624S substitution in TPO did not diminish binding of rhFab TP14, but a combination of four mutations K627G/E604A/D620R/D624S (KDDE) led to approximately 6-fold reduction of rhFab TP14 binding, without an effect on rhFab TO10 binding. Importantly, introduction of the fifth mutation D630N to this mutant (K627G/E604A/ D620R/D624S/D630N; KDDED) completely abolished binding of rhFab TP14, whereas rhFab TO10 reacted with this multiple mutant to the same extent as the wild-type TPO protein. These data identify E604, D620, D624, and D630 in TPO as residues involved in the IDR-B region, together with the previously identified K627. Characterization of autoantigenic epitopes on TPO mutants by capture ELISA

Although flow cytometry analysis allows study of the native protein expressed on the cell surface, the method appeared to have too low sensitivity to detect the effects of

Dubska et al. • Autoreactive Residues on TPO

Endocrinology, December 2006, 147(12):5995– 6003

5999

FIG. 3. Flow cytometry evaluation of IDR-A and IDR-B epitopes on TPO mutants expressed on the surface of CHO cells. Stably transfected CHO cells expressing wild-type and mutant TPO were incubated with rhFab TO10 (specific for IDR-A), rhFab TP14 (specific for IDR-B) and a pool of 20 sera from patients with AITD. Fluorescence was developed as described in Materials and Methods. Untransfected CHO cells were used as a negative control. The binding of antibodies was analyzed at least three times. Data of representative experiments are presented. Mean values of fluorescence are shown in boxes. A, Mutants in the IDR-A epitope, B, mutants in the IDR-B epitope.

single amino acid substitutions in TPO on autoantibody recognition. No differences in the binding of rhFabs between CHO cells expressing wild-type and point mutated TPO by flow cytometry were observed by Bresson et al. either (26), who eventually used BIACORE to analyze their TPO mutants. In this study, we have applied the capture ELISA with mAb A4 to characterize IDR-A and -B epitopes on TPO mutants more quantitatively (Fig. 4). We showed in our earlier work (25) that this provides an accurate method to study the effects of point mutations on the autoimmune epitopes of TPO. Different sensitivities of the FACS and ELISA methods are not unexpected in view of the different physical presentation of the TPO molecules in these two assays. The results of binding of rhFabs TP1 and TO10 specific for IDR-A confirmed the important effect of the mutation R225A (25) and demonstrated an equally dramatic reduction in binding to D707N, and a 50% reduction for R646A (Fig. 4A, left panel). The double and triple mutants, as expected, also showed dramatically reduced binding, but importantly, none of the single or multiple mutations in IDR-A affected binding of IDRB-specific rhFabs (Fig. 4A, right panel). For the IDR-B candidate residues, the results confirmed our earlier identification of K627 (24), and demonstrated

similar effects for E604A and D630N (Fig. 4B, right panel). The D620R mutation caused about 40% reduction of binding, whereas the D624S mutation had no effect on the reactivity of IDR-B-specific rhFabs. However, introducing D620R and D624S mutations together in the mutant D620R/D624S caused about 80% reduction of binding of IDR-B-specific rhFabs, and this additive effect confirms that D620 and D624 participate in the structure of the IDR-B region. The other double and triple mutations reduced binding of the IDR-Bspecific Fab to almost undetectable levels. Again, none of these single or multiple mutations affected binding of IDRA-specific Fab (Fig. 4B, left panel). The reactions of all mutants with sera from 12 patients with Hashimoto’s thyroiditis were also tested by capture ELISA (Fig. 5). Patients’ sera were positive for anti-TPO antibodies, with autoantibody levels higher than 500 IU/ml. The greatest reduction of autoantibody binding relative to the wild-type TPO protein was apparent with the multiple mutants: R225A/R646A (mean, 30%), R646A/D707N (mean, 39%), R225A/R646A/D707N (mean, 38%) and K627G/E604A/ D620R/D624S/D630N (mean, 35%). These findings provide compelling evidence that the functionally important autoantibody IDRs on TPO—as judged by patients’ autoantibody

6000

Endocrinology, December 2006, 147(12):5995– 6003

Dubska et al. • Autoreactive Residues on TPO

FIG. 4. Relative reactivity of rhFabs specific for IDR-A and IDR-B epitopes to mutant TPO proteins determined by capture ELISA. Membrane proteins extracted from CHO cells expressing wildtype and mutant TPO were captured onto microtitre plates coated with antiTPO mAb A4, followed by incubation with rhFabs TP1 and TO10 (specific for IDR-A) or TP5 and TP14 (specific for IDR-B). Data are expressed as percentage values of wild-type TPO reactivity. A, Mutants in the IDR-A epitope, B, mutants in the IDR-B epitope.

responses—were affected by the substitutions investigated here. Three-dimensional localization of residues contributing to IDR-A and IDR-B on TPO

All mutated residues are solvent accessible according to the TPO monomer model (Fig. 6). Residues R225, R646, and D707, clearly belong to one cluster on the protein surface and residues E604, D620, D624, K627, and D630 to another. The centers of the two patches are about 40Å apart. Energy minimization calculations for each of the mutations showed that none affected the course of the polypeptide backbone, and only differences in the side-chain are predicted to occur. Residues K713, P715, E716, and D717, which have been reported as components of IDR-A (26), are located at the protein surface in the neighborhood of R225, R646, and D707, and together form an elliptical patch (major axis 20Å, minor axis 18Å, surface area 940Å2). This is entirely compatible with an antigen antibody binding area (36). Similarly, residues T605, P606, A607, D608, and R616, identified previously as part of IDR-B (27), lie immediately adjacent to E604, D620, D624, K627, and D630, together forming a patch of longest dimension about 18Å and area 1000 Å2, again compatible with an antibody binding area. Discussion

Our earlier studies on conformational epitopes of hTPO demonstrated that combined use of molecular modeling and site-directed mutagenesis is a very useful tool for epitope mapping. We provided evidence supporting the role of R225

as a key residue contributing to the IDR-A epitope, and K627 as a key residue for IDR-B (25). These two residues located the autoreactive regions on the TPO structural model but clearly did not define their full extent. In the present work, we extended the mapping to identify several key residues in each of the IDRs that, together with other data, provide a detailed description of their nature and extent. Mutations were incorporated in different combinations into TPO, and mutant proteins were expressed in CHO cells at levels comparable to wild type. Flow cytometry analysis showed that all combinations of double substitutions at positions R225, R646, and D707 substantially reduced the reaction of mutant TPO with IDR-A-specific rhFab TO10, with almost complete inhibition for the triple mutant R225A/ R646A/D707N. A similar effect was observed for the IDR-B candidate residues: the double K627G/E604A and multiple K627G/E604A/D620R/D624S mutations substantially weakened the binding of the IDR-B-specific rhFab TP14, and substitution of all five residues K627G/E604A/D620R/ D624S/D630N totally abolished binding. The results clearly demonstrate that these residues contribute to the formation of these two IDRs. Studies of antigen-antibody complexes have revealed that between 15- and 22-amino acid residues of the antigen form a typical structural epitope. However, only three to five residues may contribute significantly to the binding energy and constitute the functional epitope (reviewed in Ref. 36). Replacement of such critical residues within an epitope usually causes a reduction or complete loss of antibody binding, whereas mutations of other residues elsewhere within the

Dubska et al. • Autoreactive Residues on TPO

FIG. 5. Reduction of patients’ autoantibodies binding to TPO mutants determined by capture ELISA. Sera from 12 patients positive for anti-TPO autoantibodies were tested individually with all the mutants. Data are expressed as percentage values after subtraction of the mutant’s reactivity from the reactivity of wild-type TPO, considered as 100%. Each point represents a mean value of the individual serum binding. SDs for all results were negligible. Horizontal bar in each data series denotes the mean value of reduction of all sera binding.

structural epitope may lead to much smaller, if any, changes in affinity. Thus, we have assumed that a substantial loss of the reactivity of IDR-specific rhFabs by a point mutation indicates that this residue is a critical contributor to the binding energy in the antibody complex. We thus assigned the contribution of individual amino acid residues to the IDRs by capture ELISA. R225, R646, and D707 were identified as critical for IDR-A-specific rhFab binding, whereas E604, D620, K627, and D630 appeared to be essential for IDR-B-specific rhFab binding. D624 did not contribute individually to IDR-B-specific rhFab binding; however, an additive effect of D620R and D624S together indicated that D624 contributes to the structural epitope within IDR-B. It has to be underlined that in both FACS and ELISA analyses, mutations led specifically to the loss of binding of IDR-A- or -B-specific rhFabs, without affecting the binding of autoantibodies to the other determinant. This reciprocal relationship provides strong evidence that none of the mutations caused destabilization and/or misfolding of the expressed

Endocrinology, December 2006, 147(12):5995– 6003

6001

TPO, and also supports the idea that there are indeed two distinct IDRs. Antibody-antigen interfaces are often roughly circular (37) with an average surface area of 700 –900 Å and radius of 8 ⫾ 1.7 Å (38, 39). Most of the contacts are made through sidechain atoms of the residues (40). We identify on our structural TPO model seven amino acid residues comprising a patch that covers 940 Å2 corresponding to IDR-A. As demonstrated in this study R225, R646, and D707 are clearly key residues of a functional epitope within IDR-A, and in addition, K713, P715, E716, and D717 may contribute to the structural epitope. K713 was previously shown to be a contact residue for rhFab TR1.9 recognizing IDR-A (41). Subsequently, the sequence K713-D717 was identified by Bresson et al. (26) as being involved in IDR-A recognition. Substitutions of K713 and E716 did not influence IDR-A-specific rhFabs binding in our earlier work (25); however, our interpretation is that different rhFabs used by different groups, although binding to the same overall region on TPO, will not necessarily have identical key contact residues. Concerning the IDR-B determinant, we identify on the structural TPO model a group of ten residues that constitute a contiguous area of approximately 1000 Å2. We demonstrated in this study that E604, D620, K627, and D630 are key residues that form a functional epitope within IDR-B. Using point-mutated peptides, Bresson et al. (27) recently assigned E604 as an important residue recognized by antipeptide (P14) antibodies and human autoantibodies, and in the present work we confirmed the critical role of E604 in IDR-B by introducing the E604A mutation into the full-length TPO protein. T605, P606, A607, and D608 have been previously reported as critical for the binding of antipeptide (P14) antibodies to TPO (25, 27) and R616 has been identified as a component of the sequence recognized by the rhFab T1.8 (27). However, alanine replacements of D608 and R616 did not affect binding of the IDR-B-specific Fabs (25), and so these two are not key functional residues, but rather contribute to the wider structural epitope, as does D624 (this study), which only affects Fab binding when combined with mutation of D620. In summary, we have defined several key residues, i.e. the functional epitopes, through which autoantibodies recognize IDR-A and -B domains on the autoantigen TPO. These key residues represent the important structural information required to understand the molecular mimicry that might be responsible for the breakdown of tolerance, and to design possible therapeutic agents that may block the binding of patients’ autoantibodies. Acknowledgments We are grateful to Grzegorz Wieczorek for calculations of peptide patches area. Received July 10, 2006. Accepted August 29, 2006. Address all correspondence and requests for reprints to: Monika Gora, Department of Biochemistry and Molecular Biology, Medical Centre of Postgraduate Education, Marymoncka 99, 01-813 Warsaw, Poland. E-mail: [email protected]. This work was supported by grant from Ministry of Science and Higher Education, Poland, No. PBZ-MIN-015/PO5/2004.

6002

Endocrinology, December 2006, 147(12):5995– 6003

Dubska et al. • Autoreactive Residues on TPO

FIG. 6. Structural model of the MPO-like domain of hTPO. Localization of amino acid residues mutated in this study is shown on the whole MPO-like domain of TPO (left) and on the enlarged fragments of IDR-A and IDR-B regions (insets). The side chains of investigated residues are shown as sticks and labeled with their residue numbers.

References 1. Rapoport B, McLachlan SM 2001 Thyroid Autoimmunity. J Clin Invest 108: 1253–1259 2. Banga JP 1998 Thyroid peroxidase: developments in our understanding of its structure and the relevance of these findings to autoimmunity. Curr Opinion Endocrinol Diabetes 5:275–281 3. Ruf J, Carayon P 2006 Structural and functional aspects of thyroid peroxidase. Arch Biochem Biophys 445:269 –277 4. Blanchin S, Estienne V, Durand-Gorde JM, Carayon P, Ruf J 2003 Complement activation by direct C4 binding to thyroperoxidase in Hashimoto’s thyroiditis. Endocrinology 144:5422–5429 5. Flynn JC, Gardas A, Wan Q, Gora M, Alsharabi G, Wei WZ, Giraldo AA, David CS, Kong YM, Banga JP 2004 Superiority of thyroid peroxidase DNA over protein immunization in replicating human thyroid autoimmunity in HLA-DRB1*0301 (DR3) transgenic mice. Clin Exp Immunol 137:503–512 6. Ng HP, Banga JP, Kung AW 2004 Development of a murine model of autoimmune thyroiditis induced with homologous mouse thyroid peroxidase. Endocrinology 145:809 – 816 7. Ng HP, Kung AW 2006 Induction of autoimmune thyroiditis and hypothyroidism by immunization of immunoactive T cell epitope of thyroid peroxidase. Endocrinology 147:3085–3092 8. Ruf J, Toubert ME, Czarnocka B, Durand-Gorde JM, Ferrand M, Carayon P 1989 Relationship between immunological structure and biochemical properties of human thyroid peroxidase. Endocrinology 125:1211–1218 9. Czarnocka B, Janota-Bzowski M, McIntosh RS, Asgar MS, Watson PF, Kemp EH, Carayon P, Weetman AP 1997 Immunoglobulin Gk antithyroid peroxidase antibodies in Hashimoto’s thyroiditis: epitope mapping analysis. J Clin Endocrinol Metab 82:2639 –2644 10. Chazenbalk GD, Portolano S, Russo D, Hutchison JS, Rapoport B, McLachlan S 1993 Human organ-specific autoimmune disease. Molecular cloning and expression of an autoantibody gene repertoire for a major autoantigen reveals an antigenic immunodominant region and restricted immunoglobulin gene usage in the target organ. J Clin Invest 92:62–74 11. Jastrzebska-Bohaterewicz E, Gardas A 2004 Proportion of antibodies to the A and B immunodominant regions of thyroid peroxidase in Graves and Hashimoto disease. Autoimmunity 37:211–216 12. Taurog A, Dorris ML, Doerge DR 1996 Mechanism of simultaneous iodination and coupling catalyzed by thyroid peroxidase. Arch Biochem Biophys 330: 24 –32 13. Libert F, Ruel J, Ludgate M, Swillens S, Alexander N, Vassart G, Dinsart C 1987 Thyroperoxidase, an auto-antigen with a mosaic structure made of nuclear and mitochondrial gene modules. EMBO J 6:4193– 4196 14. Bresson D, Rebuffat SA, Peraldi-Roux S 2005 Localization of the immunodominant region on human thyroid peroxidase in autoimmune thyroid diseases: an update. J Autoimmune Dis 2:2 15. Blanchin S, Estienne V, Guo J, Rapoport B, McLachlan SM, Carayon P, Ruf J 2002 Human thyroperoxidase folds in one complex B-cell immunodominant region. Biochem Biophys Res Commun 295:1118 –11124

16. Guo J, Rapoport B, McLachlan SM 2002 Localisation of the thyroid peroxidase autoantibody immunodominant region to a junctional region containing portions of the domains homologous to complement control protein and myeloperoxidase J Biol Chem 277:40189 – 40195 17. Bresson D, Cerutti M, Devauchelle G, Pugniere M, Roquet F, Bes C, Bossard C, Chardes T, Peraldi-Roux S 2003 Localisation of the discontinuous immunodominant region recognised by human thyroperoxidase autoantibodies in autoimmune thyroid disease. J Biol Chem 278:9560 –9569 18. Grennan Jones F, Ziemnicka K, Sanders J, Wolstenholme A, Fiera R, Furmaniak J, Rees Smith B 1999 Analysis of autoantibody epitopes on human thyroid peroxidase. Autoimmunity 30:157–169 19. Xiong Z, Farilla L, Guo J, McLachlan S, Rapoport B 2001 Does the autoantibody immunodominant region on thyroid peroxidase include amino acid residues 742–771? Thyroid 11:227–231 20. Gardas A, Sohi M, Sutton BJ, McGregor AM and Banga JP 1997 Purification and crystallisation of the human autoantigen thyroid peroxidase. Biochem Biophys Res Commun 237:366 –370 21. Hendry E, Taylor G, Ziemnicka K, Grennan-Jones F, Furmaniak J, Rees Smith B 1999 Recombinant human thyroperoxidase expressed in insect cells is soluble at high concentrations and forms diffracting crystals. J Endocrinol 160:R13–R15 22. Hobby P, Gardas A, Radomski R, McGregor AM, Banga JP, Sutton BJ 2000 Identification of an immunodominant region recognised by human autoantibodies in a three dimensional model of thyroid peroxidase. Endocrinology 141:2018 –2026 23. Gardas A, Watson PF, Hobby P, Smith A, Weetman AP, Sutton BJ, Banga JP 2000 Human thyroid peroxidase: mapping of autoantibodies, conformational epitopes to the enzyme surface. Redox Report 5:237–241 24. Go´ra M, Gardas A, Wiktorowicz W, Hobby P, Watson PF, Weetman AP, Sutton BJ, Banga JP 2004 Evaluation of conformational epitopes on thyroid peroxidase by antipeptide antibody binding and mutagenesis. Clin Exp Immunol 136:137–144 25. Go´ra M, Gardas A, Watson PF, Hobby P, Weetman AP, Sutton BJ, Banga JP 2004 Key residues contributing to dominant conformational autoantigenic epitopes on thyroid peroxidase identified by mutagenesis. Biochem Biophys Res Commun 320:795– 801 26. Bresson D, Pugniere M, Roquet F, Rebuffat SA, N-Guyen B, Cerutti M, Guo J, McLachlan SM, Rapoport B, Estienne V, Ruf J, Chardes T, Peraldi-Roux S 2004 Directed mutagenesis in region 713–720 of human thyroperoxidase assigns 713KFPED717 residues as being involved in the B domain of the discontinuous immunodominant region recognized by human autoantibodies. J Biol Chem 279:39058 –39067 27. Bresson D, Rebuffat SA, Nguyen B, Banga JP, Gardas A, Peraldi-Roux S 2005 New insights into the conformational dominant epitopes on thyroid peroxidase recognized by human autoantibodies. Endocrinology 146:2834 –2844 28. Rebuffat SA, Bresson D, Nguyen B, Peraldi-Roux S 2006 The key residues in the immunodominant region 353–363 of human thyroid peroxidase were identified. Int Immunol 18:1091–1099

Dubska et al. • Autoreactive Residues on TPO

29. Ho SN, Hunt HD, Horton RM, Pullen JK, Pease LR 1989 Site-directed mutagenesis by overlap extension using the polymerase chain reaction. Gene 77:51–59 30. McIntosh RS, Asghar MS, Kemp EH, Watson PF, Gardas A, Banga JP, Weetman AP 1997 Analysis of immunoglobulin G kappa antithyroid peroxidase antibodies from different tissues in Hashimoto’s thyroiditis. J Clin Endocrinol Metab 82:3818 –3825 31. Ewins DL, Barnett PS, Tomlinson RW, McGregor AM, Banga JP 1992 Mapping epitope specificities of monoclonal antibodies to thyroid peroxidase using recombinant antigen preparations. Autoimmunity 11:141–149 32. Fenna R, Zeng J, Davey C 1995 Structure of the green heme in myeloperoxidase. Arch Biochem Biophys 316:653– 656 33. Fiedler TJ, Davey CA, Fenna RE 2000 X-ray crystal structure and characterization of halide-binding sites of human myeloperoxidase at 1.8 A resolution. J Biol Chem 275:11964 –11971 34. Schwede T, Kopp J, Guex N, and Peitsch MC 2003 SWISS-MODEL: an automated protein homology-modeling server. Nucleic Acids Res 31:3381– 3385

Endocrinology, December 2006, 147(12):5995– 6003

6003

35. Vriend G 1990 A molecular modeling and drug design program. J Mol Graph 8:52–56 36. Chakrabarti P, Janin J 2002 Dissecting protein-protein recognition sites. Proteins 47:334 –343 37. Van Regenmortel MHV 1996 Mapping epitope structure and activity: from one-dimensional prediction to four-dimensional description of antigenic specificity. Methods 9:465– 472 38. Jones S, Thornton JM 1996 Principles of protein-protein interactions. Proc Natl Acad Sci USA 93:13–20 39. Davies DR, Cohen GH 1996 Interactions of protein antigens with antibodies. Proc Natl Acad Sci USA 93:7–12 40. Lo Conte L, Chothia C, Janin J 1999 The atomic structure of protein-protein recognition sites. J Mol Biol 285:2177–2198 41. Guo J, Yan XM, McLachlan SM, Rapoport B 2001 Search for the autoantibody immunodominant region on thyroid peroxidase: epitopic footprinting with a human monoclonal autoantibody locates a facet on the native antigen containing a highly conformational epitope. J Immunol 166:1327–1333

Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.