Article révisé par les pairs
Résumé : Thyroperoxidase is a membrane-bound, heme-containing enzyme which catalyses iodination of thyroglobulin and coupling of resulting iodotyrosines to produce thyroid hormone. In addition to the full length molecule of 933 amino acids (TPO1), Northern blotting and sequencing have revealed several shorter transcripts. The most abundant is a species lacking 171 nucleotides in which the alternative splicing results in the deletion of codons 533-590 in exon 10 (TPO2). Evidence for TPO2 transcripts being translated into a protein is lacking, but in Western blots TPO invariably appears as a doublet of 110 and 105 kDa. In the present study we have produced two recombinant fusion proteins for: (i) the 57 amino acids which are spliced out in TPO2 and (ii) for the 20 amino acids which bridge the splice site (10 amino acids on both sides). Both recombinant fragments have been produced in the pMAL-cRI vector as a maltose-binding protein (MBP) fusion, permitting their purification from a bacterial lysate on an amylose column. Rabbits have been immunized by intradermal injection of 500 micrograms of fusion protein, initially in complete Freund's adjuvant followed by two boosts, at 2-week intervals, in incomplete Freund's adjuvant. The resulting high titre immune sera (IS) were reactive with the relevant immunising antigens, when tested by ELISA. Depletion of each serum by passage through an MBP-CNBr Sepharose column allowed purification of antibodies against the relevant peptides, as demonstrated by ELISA with the appropriate fusion protein and MBP. This demonstrates that we have produced specific polyclonal antibodies for the 57 amino acids unique to TPO1 and for the amino acid segment bridging the splice site, found in TPO2. These polyclonal antibodies were used in Western blotting experiments with normal and Graves' thyroid membranes, in reducing and non-reducing conditions. Monoclonal 47/C21 which recognises a linear epitope (amino acids residues 710-722) common to TPO1 and TPO2 was used as a control. In non-reducing conditions, we observed a broad signal at 105-110 kDa, which appeared to comprise two bands, with both polyclonal antibodies and the monoclonal. There was no difference in the image between the normal and the Graves' thyroid. In reducing conditions, the broad signal resolved clearly into two distinct bands, one at 105 and the other at 110 kDa. Once again we observed exactly the same pattern of reactivity with all three antibodies both in normal and Graves' glands. We conclude that the TPO doublet is not the consequence of translation of TPO2.