Differential autoantibody responses to thyroid peroxidase in patients with Graves' disease and Hashimoto's thyroiditis.

In order to examine the specificity of the autoantibody response to thyroid peroxidase (TPO, previously identified as thyroid microsomal antigen) in autoimmune thyroid disease, we examined reactivity of sera from 45 Hashimoto's and 48 Graves' patients to native thyroid microsomes, denatured and reduced human TPO and several recombinant fragments of human TPO corresponding to amino acids 457-933 of the native protein. Both Graves' and Hashimoto's sera bound native, denatured and reduced TPO at significantly greater rate than normal controls, and no differences were noted between the two disorders in binding to these forms of the autoantigen. Binding was also noted to two recombinant fragments of TPO, corresponding to amino acids 513-633 and 633-933 in TPO. The frequency of autoantibodies to the TPO AA(633-933) region was not significantly different in Hashimoto's vs. Graves' disease patients (58% vs. 65% respectively), and appeared to relate to evidence of glandular inflammation in the Graves' patients (presence of anti-thyroglobulin antibodies and elevated anti-microsomal antibody levels). In contrast, antibodies to the TPO AA(513-633) fragment were significantly more common and of higher titer in Hashimoto's vs. Graves' disease patients, and did not correlate with any measure of glandular inflammation. These results identify two specific regions of TPO autoantibody binding and indicate that there are differences in the autoantibody response to TPO in Hashimoto's and Graves' diseases.

Identification of localized autoantibody epitopes in thyroid peroxidase.

Recent reports have disagreed on the nature of the autoantibody epitopes in thyroid peroxidase (TPO). We used immunoprecipitation of recombinant human TPO constructs to determine if localized autoantibody binding sites exist in this autoantigen. In vitro transcription and translation of TPO cDNA fragments yielded 35S-labeled products consisting of either full-length protein (933 amino acids) or N-terminal peptides of 631, 455, and 120 amino acids. Immunoprecipitates analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and autoradiography revealed that the Hashimoto's sera consistently precipitated the full-length and the 631 amino acid products, but not the shorter N-terminal peptides. An additional construct resulting in a full-length TPO peptide with an internal deletion of amino acids 4-455 was also made, and this product was also precipitated by the Hashimoto's sera. A fusion protein consisting of maltose binding protein followed by amino acids 456-933 of human TPO was produced in Escherichia coli and subjected to Western blot analysis using the Hashimoto's sera. The Hashimoto's sera reacted with the MalTose binding protein TPO (MBP/TPO) fusion protein, but not a control fusion protein (MBP/LacZ alpha). Together, these results indicate the presence of localized autoantibody epitopes in the portion of the human TPO molecule from amino acids 456 to 933, with at least one binding site located between amino acids 456 and 631.