Two forms of human thyroid peroxidase cDNAs were isolated from a Xgtll cDNA library, prepared from Graves disease thyroid tissue mRNA, by use of oligonucleotides. The longest complete cDNA, designated phTPO-1, has 3048 nucleotides and an open reading frame consisting of 933 amino acids, which would encode a protein with a molecular weight of 103,026. Five potential asparagine-linked glycosylation sites are found in the deduced amino acid sequence. The second peroxidase cDNA, designated phTPO-2, is almost identical to phTPO-1 beginning 605 base pairs downstream except that it contains 1-base-pair difference and lacks 171 base pairs in the middle of the sequence. This results in a loss of 57 amino acids corresponding to a molecular weight of 6282. Interestingly, this 171-nucleotide sequence has GT and AG at its 5' and 3' boundaries, respectively, that are in good agreement with donor and acceptor splice site consensus sequences. Using specific oligonucleotide probes for the mRNAs derived from the cDNA sequences hTPO-1 and hTPO-2, we show that both are expressed in all thyroid tissues examined and the relative level of two mRNAs is different in each sample. These results suggest that two thyroid peroxidase proteins might be generated through alternate splicing of the same gene. By using somatic cell hybrid lines, the thyroid peroxidase gene was mapped to the short arm of human chromosome 2.

A survey of goitre was made in the goitrous regions on the coast of Hokkaido, the northern island of Japan. Prevalence of goitre was confirmed in Hidaka coast and Rishiri Island. All goitrous patients were clinically euthyroid. The usual diet of the inhabitants of these districts consisted of a large quantity of iodine-rich seaweeds. Urinary excretion of iodine in five patients exceeded 20 mg per day. Studies of 131I and 127I metabolism were performed both during ingestion and after restriction of seaweed. When the patients were taking their usual diet, the mean thyroidal 131I uptake in 57 patients was 9.6% at 3 hours and 11.7% at 24 hours. In five of seven patients plasma inorganic iodine and thyroidal iodine space were markedly increased. Significant discharge of thyroidal 131I followed administration of thiocyanate. After withdrawal of seaweed from their usual diet, the plasma inorganic iodine was below 2 μg/100 ml but the thyroidal stable iodine uptake was higher than normal, depending on increase in thyroidal 131I clearance rate. No discharge was shown by thiocyanate block. Plasma PBI and thyronine-iodine level and serum T3 resin uptake were within the normal range. Radiochromatography of the thyroid tissue of the goitrous patients showed an increase in MIT/DIT ratio and a decrease in T3 + T4 proportion. No evidence for peripheral defect in DIT-131I deiodination was obtained. In a few patients restriction of seaweed induced a marked decrease in the size of goitre. The major cause of the endemic coast goitre seems to be excessive and longstanding intake of iodine from seaweed, and the similarities of iodine metabolism between the endemic coast goitre and iodide goitre arc discussed.

To clarify whether apoptosis of thyroid follicular epithelial cells occurs at the tissue level in autoimmune thyroiditis, 17 specimens of thyroid tissues with Hashimoto's thyroiditis were stained for fragmented DNA. Almost all nuclei of follicular epithelial cells forming atrophic thyroid follicles surrounded by mononuclear cell infiltration and fibrosis showed positive staining. With increasing distance from lymphoid cell follicles, the percentage of follicular epithelial cells with DNA fragmentation-positive nuclei decreased (30-80%). Electron microscopic study revealed the existence of epithelial cells with shrunk and condensed nuclei. The frequency of those cells in different areas was almost compatible with that of cells with fragmentation-positive nuclei. These findings suggest that apoptosis plays an important role in the thyroid tissue injury in autoimmune thyroiditis.

Staining for dipeptidyl aminopeptidase IV (DAP IV [EC: 3.4.14.5]) activity was applied to aspiration biopsy specimens or imprint preparations of surgical biopsy specimens from thyroid tumors. Material was obtained from 55 patients with histologically proven thyroid diseases: 9 with papillary carcinoma, 5 with follicular carcinoma, 11 with follicular adenoma, 13 with adenomatous goiter, 8 with Hashimoto's thyroiditis, and 9 with other benign conditions. Most tumor cells, follicular lumina in cell clusters, and intranuclear inclusions were strongly positive for DAP IV in all examples of papillary or follicular carcinoma. In contrast, only a few epithelial cells were labeled for DAP IV in follicular adenoma and adenomatous goiter. Some Hürthle cells in Hashimoto's thyroiditis also were positive for DAP IV. When a DAP IV scoring system based on the percentage of positive cells and staining intensity was used, all benign tissues except one (from a follicular adenoma) were found to have extremely low scores. These results indicate that staining for DAP IV activity is a simple but useful tool to aid in distinguishing benign from malignant thyroid neoplasms.

Serum autoantibodies to thyroid peroxidase (TPO) in patients with thyroid autoimmune diseases were studied by micro-ELISA and immunoblotting. Twenty-four patients, 15 with Graves' disease and 9 with Hashimoto's thyroiditis, whose serum titers were greater than 3200 on the microsomal hemagglutination test (except for 1 patient with a titer of 800) had autoantibodies to TPO. Both immunoglobulin G and M classes of autoantibodies were detected, with the former being more prominent. When TPO and thyroid microsomes were used as a target in a competitive binding inhibition test, the results suggested that TPO was a major thyroid microsomal antigen. On the other hand, immunoblotting analysis showed 3-4 bands in the 45-60K region stained by patients' sera in addition to human TPO with mol wt of 100K and 107K; only the latter 2 bands stained with antiporcine TPO antibody. In the majority of sera, TPO bands were clearer than others, although some sera showed the clearest band with a mol wt of 55K. These results indicate that patients with autoimmune thyroid disease often have autoantibodies to TPO that can be detected by micro-ELISA and immunoblotting, and that TPO is a major component of the thyroid microsomal antigen.

In this report we show that CD26 (dipeptidyl peptidase IV/DPP IV) is a novel molecular marker for differentiated thyroid carcinoma. Northern-blot analysis of 22 various thyroid tissues revealed that CD26 is a more specific marker of differentiated thyroid carcinoma than 3 proto-oncogenes previously reported to increase mRNA expression in thyroid carcinomas: c-met, c-erbB-2 and EGF-R. A comparative study of 3 CD26 assays, Northern blotting, immunohistochemical staining and activity staining clearly showed that CD26 enzyme activity staining is the most specific assay for differentiated thyroid carcinoma, yet the easiest to perform. Activity staining of 216 thyroid tissues detected CD26 in all 52 papillary carcinomas and all 5 follicular carcinomas, while all 58 cases of Graves' disease were CD26-negative. Among benign neoplasms, 54 of 55 adenomatous goiters and 29 of 33 follicular adenomas were CD26 negative. Staining intensity of the enzyme activity was relative to the degree of CD26 mRNA expression. Southern-blot study showed no gene amplification or major translocation of the CD26 gene in 7 papillary carcinomas examined. Based on this study, ectopic expression of CD26 in differentiated thyroid carcinomas is thought to be mainly caused by increased CD26 mRNA expression. In conclusion, CD26 activity staining is a simple, specific assay which should be added to the usual pathological examinations in order to distinguish differentiated thyroid carcinomas from benign thyroid diseases.

Fibrinogen interactions with vascular endothelial cells are implicated in various physiological and pathophysiological events, including angiogenesis and wound healing. We have shown previously that integrin alpha(5)beta(1) is a fibrinogen receptor on endothelial cells [Suehiro, K., Gailit, J., and Plow, E.F. (1997) J. Biol. Chem. 272, 5360-5366]. In the present study, we have characterized fibrinogen interactions with purified alpha(5)beta(1) and have identified the recognition sequence in fibrinogen for alpha(5)beta(1). The binding of fibrinogen to immobilized alpha(5)beta(1) was selectively supported by Mn(2+). Fibrinogen bound to purified alpha(5)beta(1) in a time-dependent, specific, and saturable manner in the presence of Mn(2+), and the binding was blocked completely by Arg-Gly-Asp (RGD)-containing peptides and by anti-alpha(5) and anti-alpha(5)beta(1) monoclonal antibodies. A monoclonal antibody directed to the C-terminal RGD sequence at Aalpha572-574 significantly inhibited the binding of fibrinogen to alpha(5)beta(1), whereas monoclonal antibodies directed to either the N-terminal RGD sequence at Aalpha95-97 or the C-terminus of the gamma-chain did not. Furthermore, substituting RGE for RGD at position Aalpha95-97 in recombinant fibrinogen had a minimal effect on binding, whereas substituting RGE for RGD at position Aalpha572-574 decreased binding by 90%. These results demonstrate that the C-terminal RGD sequence at Aalpha572-574 is required for the interaction of fibrinogen with alpha(5)beta(1).

An H2O2-generating fraction was prepared from porcine thyroid homogenate by differential and Percoll-density gradient centrifugations. The fraction consisted of mainly fragmented plasma membranes as judged by marker enzyme analysis and electron microscopy. The fraction produced H2O2 by reaction with NADPH only in the presence of Ca2+. The Ca2+ concentration for half-maximal activation (KCa) was about 0.1 microM and the Hill coefficient was 2. Sr2+ also activated the reaction whereas Mn2+, Zn2+, and Cd2+ inhibited it. The reaction was enhanced about twice by addition of ATP but not ADP, and inhibited by addition of hexokinase together with glucose to remove ATP. The Km value for NADPH was 35 microM and was less than 1/12 that for NADH. The NADPH oxidation rate was measured and the KCa and the Km were similar to those for the H2O2 production. The stoichiometry between the oxidation and the H2O2 formation was essentially 1. Superoxide dismutase (SOD) and KCN did not affect H2O2 production. The fraction catalyzed NADPH-cytochrome c reduction but the activity was SOD-insensitive. These results suggest that H2O2 was not generated through superoxide anion formation. NADPH-dichloroindophenol (DCIP) reductase activity was also observed and DCIP inhibited the production of H2O2. The cytochrome c and DCIP reductase activities were not influenced by Ca2+ or ATP. A unique electron transport system regulated by Ca2+ and ATP exists in the thyroid plasma membrane that produces H2O2. The concentrations of Ca2+ and ATP in thyroid cells may regulate hormone synthesis through activation of the production of H2O2, a substrate for peroxidase.