Thyroid iodide transporter: local sequence homologies with thyroid autoantigens

Identification of Potentially Cross-Reactive Peanut-Lupine Proteins by Computer-Assisted Search for Amino Acid Sequence Homology

BACKGROUND

Allergic cross-reactions are an issue of major concern because of implications for public health. The molecular basis of cross-allergy is the similarity of epitopes belonging to proteins of different organisms. Lupine is an emerging cause of food allergy, which has become a 'hot topic' because of recent large-scale introduction into processed foods and frequent cross-reactions with other members of the legume family. However, no lupine allergen has been characterized thus far. Prompted by a recently reported case of peanut-lupine cross-allergy, we wished to identify the possible cross-reactive allergen(s) between the two vegetal species.

METHODS

We used computer-aided amino acid sequence comparison, a well-established technique for the study of protein homology, and followed the FAO/WHO guidelines for the identification of potential allergens. We also performed a three-dimensional modeling of the suspected cross-reactive proteins to compare their molecular surfaces.

RESULTS

We found a highly significant sequence homology and molecular similarity between allergen Ara h 8 of peanut and pathogenesis-related protein PR-10 of white lupine. Another protein of lupine, the beta-conglutin precursor, was found to be significantly homologous to the Ara h 1 allergen of peanut. The molecular surfaces of Ara h 8 and PR-10 were remarkably similar.

CONCLUSIONS

Our in silico data allow to predict the allergenicity of PR-10 and beta-conglutin precursor of white lupine according to FAO/WHO guidelines. Amino acid sequence homology also suggests that these proteins could be responsible, at least in part, for some of the allergic cross-reactions between peanut and lupine reported in the literature.

Homologies of the thyroid sodium-iodide symporter with bacterial and viral proteins

We have demonstrated that Na+/I- symporter (NIS), a novel thyroid autoantigen, has local amino acid sequence homologies with the other thyroid autoantigens: Thyroglobulin (Tg), thyroid peroxidase (TPO) and thyrotropin receptor (TSH-R). These homologies concern the 4th, 5th, 6th extracellular loop and the beginning of the intracellular tail. We have expanded our studies and found that there are significant local homologies with other 11 proteins, most of them of bacterial or viral origin (e.g., Streptococcus or Herpes). These homologies concern the 2nd and 4th extracellular loop, and both the beginning and the end of the intracellular tail. These 11 homologies were retrieved by a computer-assisted search and extracted out of a database containing almost 300,000 amino acid sequences. These homologies were of magnitude greater than those concerning the three thyroid autoantigens [identities=51.1+/-7.3% vs 25.3+/-7.8% (mean+/-SD), p<0.001; similarities=70.6+/-10.7% vs 43.3+/-8.5%; p<0.001]. In addition, extensive, not local, homology was found with a number of unknown proteins from invertebrates (Drosophila melanogaster and Caenorhabditis elegans) and bacteria such as Bacillus subtilis and Xanthobacter. Previously, we had found that NIS has no extensive homology with Tg or TPO or TSH-R. This is the first demonstration of both extensive and local homologies between one thyroid autoantigen (NIS) and microbiological proteins. Taken together with data of the literature on the homologies between other thyroid antigens (Tg, TPO, TSH-R) and bacteria, the homologies we have now found reinforce the view that both bacterial and viral infections may trigger autoimmune thyroid diseases.