Search

Exposure of rat thyroid cells for 1 week to a temperature-sensitive variant of Kirsten murine sarcoma virus (KiMSV) Ras inactivated the thyroglobulin promoter (pTg). Cellular dedifferentiation was paralleled by the loss of the thyroid-specific trans-acting factor, TgTF1, which binds to pTg. When Ras was denatured by shifting cells to 39 degrees C, TgTF1 binding and pTg function recovered rapidly without the synthesis of new protein. TgTF1 could be reactivated in vitro by treating nuclear extracts with protein kinase A. After 4 weeks of exposure to the oncogene, denaturation of Ras no longer restored TgTF1 binding or reactivated pTg. Incubation of nuclear extracts with protein kinase A likewise did not reactivate TgTF1. Cells chronically exposed to Ras did, however, yield differentiated clones after treatment with 5-azacytidine. We suggest that Ras induces dedifferentiation in two sequential steps: (1) Ras reduces PKA activity; TgTF1 (or an auxiliary protein) becomes dephosphorylated, and binding to pTg is abolished. (2) The effects of Ras become imprinted by methylation, possibly of the TgTF1 gene.

Thyroglobulin gene expression was repressed in a rat thyroid cell line transformed with Kirsten murine sarcoma virus. Expression of a dominant selectable marker driven by the thyroglobulin promoter was also inhibited. Somatic cell hybridization of transformed and differentiated thyroid cells resulted in extinction of thyroglobulin gene expression. When transformed cells carrying a dominant selectable marker driven by the thyroglobulin promoter were fused to differentiated cells and expression of this marker was selected, we obtained stable hybrid cell lines expressing both the endogenous and the exogenous thyroglobulin promoters. Although the expression of v-ras remained unchanged compared with expression in the parental transformed cells, transformation was suppressed in the hybrid cell lines. The other thyroid differentiation markers, iodide uptake and thyroid-stimulating hormone-dependent growth, were inhibited in all the hybrids tested. We show that activity of the thyroglobulin promoter correlates with the presence of a thyroid nuclear factor that binds the promoter at position -60 from the transcription start site. Loss of this factor accompanies the extinction of thyroglobulin gene expression in hybrids selected for expression of a non-thyroid-specific promoter.

We have isolated the entire gene for rat thyroglobulin, the precursor for thyroid hormone biosynthesis. The gene is at least 170,000 base pairs (bp) long; 9000 bp of coding information are distributed in 42 exons of homogeneous size (150-200 bp) except for two exons of 1100 and 620 bp. The sequences coding for two major thyroxine-forming sites are localized in exons 2 and 39. These two sequences do not show any homology either at the DNA or at the protein-sequence level, even though they code for sites highly specialized for the same function. Furthermore, both the 3' and the 5' end of the thyroglobulin structural gene appear to be made of repetitive units, which again do not show any homology. On the basis of these observations, we propose that the thyroglobulin gene arose by shuffling of at least two segments, with different evolutionary histories, each of which already contained introns.

The entire rat thyroglobulin mRNA sequence (about 8500 nucleotides) has been cloned in five recombinant plasmids containing overlapping cDNA inserts. The 3' end of the mRNA is precisely defined by the poly (A) tail found in the furthest 3' end clone. Evidence that most of the 5' end is cloned come from size considerations and from a primer extension experiment. At the 3' end of the mRNA only one long open reading frame is present in the sequence of 3018 nucleotides that has been established. In the deduced protein sequence we have localized two thyroxine-forming sites in a region containing a high concentration of tyrosine residues.

Transformed rat thyroid cells fail to express thyroglobulin. Cells transformed with a Kirsten murine sarcoma virus carrying a temperature-sensitive ras allele lose their transformation phenotype when shifted to the nonpermissive (39 degrees C) temperature. The thyroglobulin promoter, however, remains inactive. Similarly, transfection of these cells with a thyroglobulin promoter fused to a neomycin resistance reporter gene does not produce clones resistant to G418. Treatment of the transfected cells with the DNA demethylating agent 5-azacytidine reactivates the thyroglobulin promoter and yields stable G418-resistant clones. We show that thyroglobulin promoter activity is correlated with the presence of a thyroid-specific nuclear factor, TgTF1. TgTF1 cannot be detected in transformed cells but reappears after treatment with 5-azacytidine at 39 degrees C. Restoration of Ras activity at 33 degrees C leads to the rapid loss of TgTF1 and G418 resistance.

A secondary structure prediction has been made using the available primary sequence data of the proposed carboxy-terminal of rat thyroglobulin. The model predicts 22% alfa-helix, 28% beta-structure and 17% beta turns. Out of the 8 possible carbohydrate acceptor-sites (Asn-x-Ser/Thr), 3 (residues 136, 368, 782) are associated with peptide sequences which favour the formation of beta-turn or loop-structures and are located in high hydrophilic regions. The entire sequence is predicted to be made up of two domains: one of them is highly structured, contains the hormonogenic sites, a cluster of tyrosines and at least one carbohydrate acceptor site.

Two plasmids containing rat thyroglobulin cDNA sequences have been constructed and characterized. A plasmid with a 500-bp insert (pRT6) was isolated and identified as thyroglobulin-specific on the basis of the tissue specificity of the inserted sequence and of its ability to retain thyroglobulin mRNA on a nitrocellulose filter. The cDNA insert in pRT6 was subsequently used to screen a rat thyroid cDNA library constructed with large cDNA. A plasmid was found containing a 1700-bp insert. The polarity and the fidelity of the insert is demonstrated by S1 mapping.

We report the structural organization of an 80 Kb segment of rat DNA, which encodes for about 40% of Thyroglobulin mRNA at the 3' end. The codogenic information included in this segment is splitted in 17 exons of homogeneous size (about 200 bp). The seven exons at the extreme 3' end have been precisely defined by DNA sequence analysis. No clear sequence homology is found among the exons, even though their coding capacity is quite similar, from 55 to 63 aminoacids residues. We located 2 hormonogenic (T4 forming) sites on the extreme 3' end of the gene in different exons. The DNA sequence coding for these functional sites shows a 70% homology in a 50 nucleotides segment. In addition we found a remnant of this sequence in other exons of the gene. Two large introns have been found on the 3' end of the gene: one is 17 Kb and the other one is more than 30 Kb long. On the basis of these findings and of preliminary studies on the remaining 5' end of the gene, we can predict that the minimum length of the rat TGB gene will be 150 Kb, which makes this gene the largest so far identified eukaryotic gene. We propose in addition that the 3' end exons arose by duplication of a common ancestor.