| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Institut für Experimentelle Endokrinologie und Endokrinologisches Forschungs-Centrum der Charité EnForCé Charité Universitätsmedizin Berlin Humboldt-Universität zu Berlin D-10098 Berlin, Germany
Address all correspondence and requests for reprints to: Prof. Dr. J. Köhrle, Institut für Experimentelle Endokrinologie und Endokrinologisches Forschungs-Centrum der Charité EnForCé, Campus Charité Mitte, Charité Universitätsmedizin Berlin, Humboldt-Universität zu Berlin, Schumannstrasse 20/21, D-10098 Berlin, Germany. E-mail: josef.koehrle{at}charite.de.
Thyroid cancers constitute the most frequent endocrine neoplasia with an incidence of one to four new cases per 100,000 persons per year, remarkably with 2- to 3-fold higher prevalence in women (1). An aggressive papillary form of childhood thyroid cancer emerged after the Chernobyl atomic reactor accident in radiation-exposed children (2, 3). Diagnosis and treatment of thyroid cancer have achieved high levels of medical and technical standard (4). Although some forms of thyroid tumors (papillary microcarcinoma or papillary thyroid cancer) show quite good prognosis or might even remain undetected or indolent (5), others such as anaplastic tumors or medullary thyroid carcinoma have a rather poor prognosis (4, 6).
So far, only few animal models faithfully recapitulating the development of the various forms of thyroid cancers have been available (see Table 1
). In this issue, Kato et al. (7) present a mouse model that convincingly mimics the full spectrum of the human follicular form of thyroid cancer. Their achievement is based on a tricky, but successful, double genetic manipulation of the expression of the thyroid hormone receptor TRß. The TRß isoform is one of the two major targets of the thyromimetically active hormone T3. They created their heterozygous mouse model by crossbreeding the knockout mice (TRß–/–) of the wild-type (wt) TRß with the knock-in (TRßPV/PV) of the human T3 receptor mutant TRßPV, which was originally identified in a patient suffering from thyroid hormone resistance syndrome. Due to a frameshift mutation in exon 10, encoding the ligand binding domain, TRßPV is unable to bind the regulatory active thyroid hormone T3 and thus, in its unliganded form, acts as a dominant-negative transcription factor at T3-responsive elements of several T3-regulated genes (8, 9). These animals, which rapidly develop follicular thyroid cancer, carry one gene-targeted, inactivated TRß allele, unable to create a TRß transcript, and express from the second allele of this locus the strong, dominant-negative TRßPV.
|
The thyroid gland may give rise to several neoplasias. Parafollicular, calcitonin-producing C-cells can develop to adenoma or medullary thyroid carcinoma, either familial or sporadic, and frequently mutations of the RET oncogene are found in these tumors (6). Depending on the type and location of mutations of this gene, medullary thyroid carcinoma tumors undergo different progression and require adapted clinical diagnostic procedures, management, and treatment also of relatives carrying the mutation. The thyroid hormone-producing follicular epithelial cells are considered to be the precursors of four other forms of thyroid neoplasia: thyroid adenoma, follicular, papillary, and anaplastic carcinoma as well as some variants therefrom such as Hürthle cell carcinoma. The latter are characterized by some peculiar histomorphological features (abundant mitochondria) and lack of accumulation of radioiodide. Two families of oncogenes have been particularly associated with differentiated thyroid tumors, RET (12) and RAS (13), whereas constitutively activating mutations of the TSH receptor gene and of G proteins (gsp) are associated with (benign) thyroid adenoma (14, 15). Several other mutations of oncogenes and tumor suppressor genes contribute to progression to dedifferentiated anaplastic carcinoma (p53, Rb) (15, 16). Many papillary thyroid carcinoma show rearrangements of the RET oncogene with several fusion partners (H4, ELE1) leading to activation of tyrosine kinases and increased proliferation of thyrocytes that might be accompanied by altered expression of genes of growth factor-receptor networks (NTRK, MET).
Current opinion prevails that adenoma with activating TSH receptor mutations do not progress to carcinoma unless a second tumor-causing hit occurs in the context of a major proliferation pressure such as severe iodine deficiency (14, 17). Furthermore, the follicular carcinoma develop along different molecular routes [RAS oncogene (13)] compared with papillary carcinoma. Still unclear is the multistage development of anaplastic thyroid tumors, which might originate by dedifferentiation of follicular or papillary tumors subsequent to acquisition of additional mutations in the p53 (16) or Rb tumor suppressor genes, activation of oncogenes or alterations in growth factors or their receptors. Recently, a novel type of chromosomal rearrangement [PAX-8/PPAR
(peroxisome proliferator-activated receptor )] was identified in differentiated thyroid tumors. This leads to a fusion protein between one of the three-key transcription factors, Pax8, relevant for thyroid development, with the nuclear receptor PPAR, which is involved in control of proliferation and differentiation (18). Tumors expressing these proteins might provide useful targets for therapeutic intervention with novel PPAR ligands.
Over several decades of intensive basic, clinically oriented, and therapeutic research, various experimental models have been developed focusing on control of thyroid development, differentiation, growth, and proliferation, which are modulated by iodine deficiency, goitrogens, carcinogens, or continuous stimulation of the gland via its master regulator, the pituitary-derived TSH, or under pathological conditions, stimulating thyroid autoantibodies in Basedow’s/Graves’ autoimmune thyroid disease, Therefore, it came as a major surprise that the thyroid hormone receptor itself acts as a tumor suppressor gene for maintenance and control of thyrocytes. Similarly surprising is the observation that the expression of the TRßPV/PV mutant, initially associated with the thyroid hormone resistance syndrome, results in development of thyroid cancer if targeted to thyrocytes. These tumors exhibit all essential stages from hyperplasia, proliferation, capsular and vascular invasion, metastatic spread to known target tissues of thyroid cancer such as lung, and eventually undergo anaplastic dedifferentiation. The metastases still express characteristic features of thyrocytes such as thyroglobulin, which is the most useful marker of metastatic proliferation in thyroid cancer patients after surgical or radioiodine-induced removal of the lesioned gland.
None of the previously available animal models for thyroid cancer up to now mimicked this multistep tumorigenesis sequence so precisely, and rapidly, as this model. But, more than that, this seems to represent not only a useful rodent model, but is also relevant for humans because TR mutations have previously also been identified in several tumors of parenchymal or epithelial organs such as liver, kidney, and also the thyroid (8, 9, 19, 20). Why, and this remains still open, do tumors develop at first (and only?) in the thyroids of this mouse model then? Will there also be primary liver, kidney, and other tumors detected, in addition to metastatic spreads of the primary thyroid tumor, during the aging process of these mice?
How can it happen that the thyrocyte, virtually representing the one and only source of the prohormone T4 in the body and normally producing, generating, and liberating a significant portion (20–30%) of L-T3, goes astray with a combination of an inactivated TRß locus, normally encoding the receptor for exactly this ligand, and a TRß mutant, acting as repressor at T3-responsive elements but deficient in control by T3? Will there be further disruptions in the powerful prereceptor control of ligand availability and cellular ligand homeostasis, which is exerted by the three intracellular deiodinase selenoenzymes (21) and cell-specific thyroid hormone transporters such as the recently identified MCT8 protein or members of the OATP (organic anion transporter gene) family (22)? Interestingly, the type I 5'-deiodinase is rapidly down-regulated in thyroid (and other tumors) and has been discussed as a differentiation marker for thyroid (and other parenchymal) tumors (23).
In this context, studies from the era before the cloning of T3 receptors are of interest, which indicated that thyroid status is linked to tumorigenesis and growth particularly of solid tumors (24, 25). In hypothyroid animal models, less tumors develop by chemical induction and fewer grafts grow after xenotransplantation, whereas hyperthyroidism favors tumorigenesis and tumor growth. Might it be that adequate T3 ligand occupation, at least of TRß maintains and favors the functional state of (terminally) differentiated epithelial and parenchymal cells? This might additionally require cooperation with adequate differentiation signals exerted by retinoids and their receptors (26, 27, 28), which form heterodimers with TR.? Do the data of the models presented by the group of Cheng et al. (7, 10) imply that TR
alone cannot act as tumor suppressor on its own? Is an adequate T3 supply to or inside the parenchymal cell, combined with proper TRß-mediated signaling, essential to prevent growth of a tumorigenic clone, to warrant a certain rate of apoptosis, and to maintain the differentiated features of thyrocytes?
Systematic cDNA (micro)array analysis of gene expression profiles from several tissues from the TRPV/PV mutant mice in comparison to hyperthyroid wt mice shed some light on signaling pathways potentially involved in the tumorigenic effect(s) of the TRßPV/PV mutant gene (29). These results indicated that several antiproliferative and growth arrest genes as well as pathways induced by T3 are down-regulated in TRßPV/PV mutant mice (e.g. the APC, dkk3, and p21 genes). In addition, expression of immune response genes is decreased in TRßPV/PV mutant mice. These constellations favor tumorigenesis and remind of the transforming action of v-erbA, the retroviral oncogene related to the TR family. Another aspect relevant to this yin and yang type of control of proliferation vs. differentiation might be the observation, that TRßPV heterodimerizes with TR1 and retinoid X receptor (11, 29) and also binds to PPAR response elements. Presumably, these interactions impair or prohibit the usual differentiating function of these nuclear receptors.
However, loss of PPAR expression is not tumorigenic per se and needs second hits to induce tumors, but PPAR +/– mice are more susceptible to (chemo-)induced cancers (11), compatible with observations that overexpression of PPAR or administration of PPAR agonists suppresses tumorigenesis. In their TRPV/PV mutant mice model, Ying et al. (11) found repression of PPAR mRNA and demonstrated impaired transcriptional activity of PPAR ligand in cotransfection experiments with TRPV.
The impression arises that loss of wt TRß removes one essential brake on the course toward uncontrolled proliferation mediated via the introduction of the TRßPV gene. However, how do we reconcile these tumorigenic events with the clinical picture of thyroid hormone resistance in the heterozygous patient from whom this mutant was isolated? A further puzzle is the observation that expression of dominant-negative mutants of TR1 in mice, under the control of ubiquitously expressed strong promoters, does not lead to (thyroid) cancer development but results in very low frequency of the transgene expression, suggesting embryonic lethality (30) and reminding of the so far unsuccessful search for TR mutants in patients with the thyroid hormone resistance syndrome.
In summary, follicular thyroid cancer rapidly develops in a novel mouse model by introducing the TRßPV/PV mutant on the background of the inactivated TRß wt locus. The demonstration of a tumor suppressor role for the TRß isoform opens new therapeutic perspectives for the antiproliferative, differentiation-promoting function of the TRß-isoform and its native ligand T3 as well as for TRß isoform-selective novel T3 analogs. This tumor suppressor function presumably cannot be exerted by the TR isotype receptor. Silencing of the TRß gene by hypermethylation or by chromatin-modifying agents might help to elucidate the mechanism(s) involved in its tumor suppressor function. Further studies need to identify whether other signaling partners and cascades are involved in mediating this tumor suppressive action in addition to nuclear receptor (heterodimerization) partners of the retinoid X receptor and PPAR families. Whether antiproliferative, apoptosis-promoting, and differentiation-maintaining effects contribute to the rapid follicular thyroid carcinogenesis can now be revealed in this novel mouse model.
 |
Footnotes |
|---|
Received July 19, 2004.
Accepted for publication August 5, 2004.
|
References |
|---|
 Top References |
|---|
during thyroid carcinogenesis. Cancer Res 63:5274–52801 fusion oncogene in human thyroid carcinoma. Science [Erratum (2000) 289:1474] 289:1357–1360
1 in mice. Endocr J 50:561–570[Medline]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
