| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Department of Microbiology and Immunology, University of Illinois College of Medicine (J.G.C., S.K., B.S.P.), Chicago, Illinois 60612; Indiana University School of Medicine (G.S.S.), Evansville, Indiana 47712; and University of Michigan (J.R.B.), Ann Arbor, Michigan 48109
Address all correspondence and requests for reprints to: Dr. Bellur S. Prabhakar (MC 790), Department of Microbiology and Immunology, E-705 Medical Sciences Building, 835 South Wolcott Avenue, Chicago, Illinois 60612.
Abstract
Patients with Graves disease have autoantibodies that bind to the TSH receptor and stimulate the thyroid, leading to hyperthyroidism. Earlier studies have shown that the ectodomain of the glycosylated human TSH receptor contains epitopes that could adsorb these pathogenic antibodies. Further studies with mutated cDNAs, chimeric proteins, peptides, and antipeptide antibodies suggested that alterations in the conformation of the protein could lead to loss of reactivity, and that thyroid-stimulating antibodies interact with the N-terminal region of the TSH receptor. Although many of these studies provided valuable insights, they were somewhat inconclusive due to limitations inherent to each of the approaches. In an attempt to further define regions within the TSH receptor with which thyroid-stimulating antibodies interact, we expressed seven recombinant TSH receptor fragments in insect cells and tested them for their ability to neutralize TSH binding inhibitory Igs and thyroid-stimulating antibody activity in the sera of patients with Graves disease. The fragments containing amino acids 22305 were able to neutralize the TSH binding inhibitory Ig activity, whereas a fragment containing amino acids 54254 was able to neutralize the thyroid-stimulating antibodies. Fragments containing additional amino acids, flanking residues 54254, failed to neutralize the thyroid-stimulating antibody activity, suggesting that thyroid-stimulating antibody epitopes are masked. Our studies show that thyroid autoantibodies, with different functional properties, bind to distinct conformational epitopes on the TSH receptor.
THE TSH RECEPTOR (TSHR) is an important autoantigen in many autoimmune thyroid diseases, including Graves disease (GD) (reviewed in Refs. 1, 2, 3, 4). GD is characterized by the presence of thyroid-stimulating antibodies (TSAb). The TSAb bind to the TSHR and mimic the action of TSH, leading to hyperthyroidism. Cloning and sequencing of a cDNA encoding the human TSHR showed that it consists of 764 amino acids (aa) with a 415-aa extracellular domain containing six potential glycosylation sites, seven hydrophobic membrane spanning domains, and one cytoplasmic domain (5, 6, 7, 8, 9, 10). Several laboratories, including our own, have shown that the ectodomain of TSHR (ETSHR) is sufficient for TSH and TSAb binding (11, 12, 13, 14, 15, 16, 17). Earlier studies had shown that glycosylated ETSHR (ETSHR-gp) is necessary for reactivity with TSAb (15, 16). Moreover, studies using mutagenesis of TSHR, synthetic peptides, and chimeric proteins (TSHR/LH-CGr) have indicated that TSAb bind to the N-terminal region of the ETSHR (reviewed in Refs. 1, 2, 3, 4). However, in many instances these studies were unable to account for either lack of native conformation or glycosylation due to limitations inherent to each of the approaches. Therefore, the objective of the current study was to use the predicted structural information (18) to design and generate recombinant fragments of ETSHR-gp and identify functional domains to which TSAb bind.
Materials and Methods
Source of sera
Sera were collected from patients with GD. Diagnoses were based
on appropriate standard clinical and laboratory criteria. From the GD
sera collected, seven sera, from patients designated 17, were
selected based on their elevated thyroid receptor antibody [TSH
binding inhibitory Ig (TBII)] values (Fig. 4
, graph A). The TBII
values were determined using a commercially available TBII assay
(Kronus, Dana Point, CA). Sera were also obtained from
healthy volunteers with no family history of autoimmune thyroid
disease. Sera were stored at -20 C until used.
|
Sf9 insect cells (Life Technologies, Inc., Gaithersburg, MD) were maintained in TNM-FH medium (Sigma, St. Louis, MO) supplemented with 10% FBS (Sigma) and antibiotic/antimycotic agents.
Construction of transfer vector
Results from computer modeling were used to design seven
recombinant ETSHR-gp fragments in which structural conformation was
likely to be maintained (18). The primers used to generate
the corresponding cDNA fragments, and the outlines of these fragments
are shown in Figs. 1
and 2
, respectively. Using ETSHR cDNA as the
template and appropriate sense and antisense primers, seven cDNAs
(ETSHR fragments 17) were created employing PCR. The cDNAs encoding
ETSHR fragments 17 were cloned in the proper orientation into the
unique EcoRI site of the transfer vector pAcGP67A
(PharMingen, San Diego, CA).
|
|
Seven different recombinant viruses were generated by cotransfecting Sf9 cells with 0.5 µg linearized Baculogold viral DNA (PharMingen) and 2 µg pAcGP67A DNA containing a cDNA encoding each of the seven ETSHR fragments according to the manufacturers protocol. Media containing the recombinant viruses were harvested after 5 d and amplified twice to obtain higher titers of virus. Amplification occurs via infecting monolayers of Sf9 cells with previously harvested medium. The media from this reinfection was harvested after 5 d and stored at -20 C. This amplification process increases the viral titers found in the medium. Generation of recombinant ETSHR-gp virus has been described previously (16).
Expression of ETSHR fragments 17 and ETSHR-gp
For recombinant protein production we followed a previously published protocol (13). Monolayers of Sf9 cells were infected with ETSHR 17 or ETSHR-gp viruses. Cells were harvested, and proteins were separated by SDS-PAGE and either stained with Coomassie blue or subjected to Western blot analysis. For Western blotting, proteins separated by SDS-PAGE were electrophorectically transferred onto nitrocellulose membranes and stained with ETSHR-gp-specific Abs.
Neutralization of TBII activity in patients sera
A modified TBII assay was used to test the ability of ETSHR-gp and ETSHR fragments 17 to neutralize the TBII activity in GD patients sera (16). Fifty microliters of GD and control sera were incubated for 2 h at room temperature with 3 x 106 insect cells expressing ETSHR-gp, ETSHR fragments 17, or uninfected Sf9 cells and then tested for their TBII activity using a commercially available RRA (Kronus). Insect cells (3 x 106) were used in this assay to ensure an excess antigen to antibody ratio.
Neutralization of the biological activity of TSAb in patients sera
Chinese hamster ovary (CHO) cells permanently transfected with a full-length human TSHR cDNA were provided by Drs. K Tahara and L. D. Kohn, NIDDK, NIH (Bethesda, MD) (19). The CHO cells expressing TSHR were grown to confluence in 96-well plates. Of the 7 sera originally selected, 4 sera (patients 14) possessed elevated TSAb activity based on cAMP assay and were used in this experiment. Ten microliters of GD patients sera were incubated with 2 x 106 insect cells expressing ETSHR-gp or ETSHR fragments 17 for 1 h at room temperature. Insect cells (2 x 106) were used to ensure an excess antigen to antibody ratio. To test the biological activity of TSAb, sera that were incubated with or without the recombinant proteins were added to duplicate wells in a hypotonic HBSS containing 0.5 mM 3-isobutylmethylxanthine and incubated at 37 C in 5% CO2 for 3 h (16). The final volume in all wells was 100 µl/well. Subsequently, supernatants were collected, diluted 1:20, and assayed for cAMP using a commercially available RIA kit (DuPont, Boston, MA).
Results
Expression of ETSHR fragments 17
The Sf9 cells were infected with various recombinant viruses
described above and analyzed by SDS-PAGE for the production of
recombinant proteins. Figure 3
shows a
Coomassie blue-stained gel and a Western blot analysis of the
recombinant proteins. Major protein bands were observed for ETSHR
fragments 17 at approximately 6, 37, 39, 34, 35, 26, and 22 kDa,
respectively. The expected molecular masses for ETSHR fragments 17
based on their aa sequence are 5.9, 25.5, 31.1, 22.0, 25.2, 17.7, and
12.2 kDa, respectively. The larger molecular masses observed for ETSHR
fragments 27 are most likely due to glycosylation, as each of these
fragments contains one or more potential glycosylation sites.
|
A routinely used commercial RRA measures the ability of antibodies
to bind to TSHR on porcine thyroid membrane and block the binding of
125I-labeled TSH. Using this assay, we tested the
recombinant proteins for their ability to neutralize autoantibodies to
TSHR in the sera of patients with GD. As shown in Fig. 4
, TBII activity was significantly
neutralized with ETSHR-gp. ETSHR fragments 2 and 3 showed varying, but
considerable, neutralization of the TBII activities. In contrast, ETSHR
fragments 1 and 47 failed to neutralize the TBII activities in any of
the seven sera tested with the exception of patient 5 with fragment 7.
This assay was repeated using patient sera 1 and 2, and similar results
were observed.
Neutralization of TSAb activity in GD sera
The TBII assay detects anti-TSHR antibodies that may or may not
have biological activities and are not useful for discriminating
between blocking (often found in patients with primary myxedema) or
stimulatory (found in patients with GD) antibodies. Therefore, we
carried out the bioassay to identify fragments to which stimulatory
antibodies bind. When CHO cells expressing TSHR were incubated with
either TSH or stimulatory antibodies, they produced elevated levels of
cAMP. To evaluate ETSHR fragments for their ability to neutralize the
TSAb activity, GD patients sera were incubated with ETSHR-gp or ETSHR
fragments 17. Subsequently, these sera were tested for their ability
to stimulate cAMP production. As shown in Fig. 5
, incubation with either ETSHR-gp or
ETSHR fragment 4 resulted in maximal neutralization of the TSAb
activity in three of four and, to a lesser extent, in one of four sera.
These results showed that ETSHR fragment 4 has TSAb-reactive epitopes.
It is interesting to note that treatment of sera with ETSHR fragments
1, 5, 6, and 7 enhanced TSAb activity to varying degrees, with ETSHR
fragment 7 showing maximal enhancement. This assay was repeated using
both patient sera 3 and 4, and similar results were observed.
|
Many investigators have attempted to finely define the different functional epitopes of TSHR involved in TSH and autoantibody binding and have suggested that TSAb bind to the N-terminal region of the ETSHR (reviewed in Refs. 1, 2, 3, 4). However, in some instances, these studies were unable to account for either lack of native conformation or glycosylation due to limitations inherent to each of the approaches. Therefore, the objective of the current study was to exploit the predicted three-dimensional structure of the TSHR to generate recombinant fragments of the human ETSHR-gp. The fragments were selected for the following reasons. ETSHR fragment 1 was selected because aa 2276 contained the unique insertion site and preceded the leucine-rich repeat (LRR) domain and the potential glycosylation sites. ETSHR fragment 2 (aa 22254) consists of the unique insertion site and the LRR. ETSHR fragment 3 (aa 22305) consists of the unique insertion, LRR domain, and all potential glycosylation sites. ETSHR fragment 4 (aa 54254) consists of only the LRR domain. ETSHR fragment 5 (aa 76305) consists of all potential glycosylation sites. ETSHR fragment 6 consists of the post-LLR domain. ETSHR fragment 7 consists of domains present beyond all glycosylation sites and represents the most hydrophilic (thus most immunogenic) region of the ETSHR-gp (18). These fragments are expected to retain their domain conformation and be glycosylated (when glycosylation sites are available). This, we reasoned, would be better than using relatively short synthetic peptides that are not glycosylated and most likely lack conformation. Therefore, the fragments were expressed and employed to identify domains to which TSHR autoantibodies bind.
We expressed ETSHR-gp and seven ETSHR fragments (i.e. ETSHR 17; aa 2276, 22254, 22305, 54254, 76305, 255416, and 305416, respectively) spanning the entire ETSHR-gp. We then tested for their ability to neutralize the TBII activity of GD patients sera. As shown previously, ETSHR- gp almost completely neutralized the TBII activity in six of seven sera, and neutralized over 50% in one of seven sera (16, 19). Similarly, ETSHR fragments 2 and 3 were also able to neutralize the TBII activity, but in some cases to a lesser extent than that seen with ETSHR-gp. Although, ETSHR fragment 3 was slightly more effective than ETSHR fragment 2, our results indicate that aa 22254 may be sufficient to significantly neutralize the TBII activity. These data are in agreement with earlier studies that showed that recombinant TSHR-261 (aa 1261) and TSHR-289 (aa 1289) expressed in CHO cells neutralized the TBII activity in the sera of Graves patients (20, 21). Our results also suggest that aa 254305 (present on ETSHR fragment 3 but not on ETSHR fragment 2) provide either additional epitopes or further enhance the conformation required for more effective autoantibody binding.
Interestingly, ETSHR fragment 1 (aa 2276) did not react with the sera, suggesting that this region is not critical. However, ETSHR fragment 5 (aa 76305) which is similar to ETSHR fragment 3 (aa 22305), except that it does not contain aa 2276, did not react with the sera either. These results suggest that aa 2276 are essential for appropriate folding of the protein required for the formation of TBII binding epitope(s). Together, our results show that a significant part of the TBII activity is due to interactions with epitopes that reside within aa 22254 of the human ETSHR-gp.
In our earlier study, using TSHR LH/CGR chimeric proteins, we showed that MC2 and MC2+4 (in which aa 89164 and aa 89164, 260333 of TSHR are replaced with the corresponding aa from LH/CGR, respectively) were able to neutralize TBII activity of patient sera. However, replacement of aa 888 (MC1), 165259 (MC3), or 260333 (MC4) completely abrogated the ability of the protein to neutralize the TBII activity in patient sera. This would suggest that aa 89164 are not important for detecting TBII activity (22). Combined with our current results, it would appear that the predominant TBII binding region consists of aa 22305 (to a lesser extent aa 22254); however, aa 89164 within this region do not appear to be critical. Full resolution of this would require structural studies.
Next, we tested the ability of the ETSHR-gp and its fragments to neutralize the ability of TSAb to induce cAMP production. To carry out these studies we selected four different sera containing relatively high levels of TSAb activity (patients 14). Both ETSHR-gp and ETSHR fragment 4 were able to neutralize the TSAb activity. This suggested that aa 54254 represent the core residues containing the epitopes to which TSAb bind. However, ETSHR fragments 2 and 3, which contain all of the amino acids present in ETSHR fragment 4, largely failed to neutralize the TSAb activity. Intuitively, one would have expected these fragments also to neutralize the TSAb activity, but they did not. This raises the possibility that amino acids flanking ETSHR fragment 4 (aa 2253 present in both ETSHR fragments 2 and 3, and aa 255305 present in ETSHR fragment 3) somehow obscure the TSAb epitope(s). Other studies using cells transfected with cDNAs encoding TSHR-LH/CGR chimeras have shown that aa 23165 of human TSHR are important for both TBII and TSAb activities, and aa 90165 are important for TSAb activity (23, 24, 25, 26, 27, 28, 29, 30). Our current data are in general agreement with these earlier results.
Antigenic sites on a complex protein can be broadly divided into dominant, subdominant, and cryptic epitopes (31, 32, 33, 34). Epitopes against which an immune response can be readily elicited are considered immunodominant. In contrast, induction of an immune response against certain epitopes is very difficult. These regions are considered to be immunologically cryptic, and the remaining epitopes are considered to be subdominant. Usually, immune responses against dominant and subdominant epitopes, but not against cryptic epitopes of self-antigens, can be readily induced in experimental animals. Despite high levels of antibodies against self-antigens (i.e. autoimmunity), these animals show no signs of autoimmune disease. This is the case in mice immunized with different insect cell-derived TSHR preparations. These mice develop very high titers of antibodies with significant TBII activity, but fail to show either severe hyperthyroidism or high levels of TSAbs (35, 36, 37, 38).
It is generally accepted that most pathogenic epitopes on autoantigens are cryptic, and if the immune response directed against dominant epitopes eventually turns against the cryptic epitopes, most likely through immunological repertoire spreading, then the disease ensues. If this were true for the TSHR, then aa 55254 probably represent an immunologically cryptic region, and by removing surrounding amino acids, this region becomes exposed and is capable of reacting with TSAbs. The presence of flanking residues on the N-terminal side (i.e. aa 2253 in ETSHR fragments 2 and 3) reduces TSAb reactivity, with this reactivity further abrogated by the presence of additional flanking residues on the C-terminal end of the fragment (i.e. aa 255305 in ETSHR fragment 3). If this were the case, then how could B cells recognize the cryptic regions and produce TSAbs that can react against intact TSHR in vivo? Although we do not know the answer, a likely explanation is that autoantibodies to TSHR are heterogeneous. Perhaps nonpathogenic antibodies that can bind to some of the dominant epitopes appear first and alter the conformation of the TSHR protein, leading to exposure of cryptic epitopes. This would allow subsequent TSAb production as well as binding.
Recently, we studied the evolution of antibody response against the TSHR in a mouse model for Graves disease (39). These studies showed that initially ELISA-positive but TBII-negative Abs appear, followed by the appearance of TBII-positive Abs with subsequent production of TSAbs (39). Evolution of TSAbs could be primarily due to repertoire spreading, although affinity maturation of antibodies can also contribute. Studies are underway to determine which of the two mechanisms is relevant for TSAb production and to see whether one can more readily induce TSAbs in experimental animals using ETSHR fragment 4 relative to other fragments.
Acknowledgments
Footnotes
This work was supported in part by an Alpha Omega Alpha Student Research Fellowship and Grants DK-47417 and DK-44972 from the NIH.
Abbreviations: aa, Amino acid; ETSHR, ectodomain of TSHR; ETSHR-gp, glycosylated ETSHR; GD, Graves disease; LRR, leucine-rich repeat; TBII, TSH binding inhibitory Ig; TSHR, TSH receptor; TSAb, thyroid-stimulating antibodies.
Received February 12, 2001.
Accepted May 9, 2001.
References
This article has been cited by other articles:
![]() |
J. Van Sande, M. J. Costa, C. Massart, S. Swillens, S. Costagliola, J. Orgiazzi, and J. E. Dumont Kinetics of Thyrotropin-Stimulating Hormone (TSH) and Thyroid-Stimulating Antibody Binding and Action on the TSH Receptor in Intact TSH Receptor-Expressing CHO Cells J. Clin. Endocrinol. Metab., November 1, 2003; 88(11): 5366 - 5374. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |