| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
TRH-TSH-THYROID |
Institute of Endocrine Sciences, University of Milano, Ospedale Maggiore IRCCS, Milano 20122, Italy
Address all correspondence and requests for reprints to: Anna Spada, M.D., Istituto di Scienze Endocrine Ospedale Maggiore, IRCCS, Via Francesco Sforza 35, 20122 Milano, Italy. E-mail: anna.spada{at}unimi.it
 |
Introduction |
|---|
 Top Introduction References |
|---|
As far as the pituitary is concerned, three recent studies report gene expression profile of human and rat pituitary in physiological and pathological conditions using cDNA microarray analysis (10, 11, 12). The first study analyzes gene expression in the pituitary from young normal rats and old rats with spontaneously occurring pituitary tumors (10). Out of the 588 cDNAs screened in the pituitary tissue, representing about 1% of the entire rat genome, 79 genes were detected in both young and old animals. In keeping with the well known age-dependent changes in hormonal secretion pattern, some genes showed marked differential expression probably linked to senescence. For example, GH gene displayed higher level of expression in the young pituitary, in agreement with the decrease in somatotrope activity by aging, while the opposite occurred for PRL. Among the genes with the highest difference in expression level, the old pituitary showed 17-fold increased levels of galanin, which is known to have a PRL promoting activity and is frequently expressed in human prolactinomas (13, 14), whereas in the young pituitary the largest overexpression was for gluthatione S transferase M2. Considering that about 60–85% of rats aged over 24 months has a spontaneous pituitary tumor, it is worth noting that the Myc-Max-interacting tumor suppressor gene, which is frequently inactivated in experimental rat tumors (15), resulted to be highly expressed in the pituitary from young rats, whereas the protooncogene c-erbA was highly expressed in old animals (10). Although this study provides important information on the age-dependent changes of gene expression in the pituitary, the use of this animal model prevents the possibility to discriminate gene expression profile associated to senescence from that associated to pituitary tumorigenesis.
The second study investigates the pattern of gene expression in human pituitary (11). By using cDNA microarray analysis, the expression profile of 7,075 genes was monitored in normal pituitary and nonfunctioning, PRL-, GH-, and ACTH-secreting adenomas. The study shows that 128 genes were differentially expressed by 2.0-fold or more in tumoral vs. normal pituitary and that 3 genes, which are involved in carcinogenesis in other tissues, were uniquely expressed among the tumor subtypes. In particular, folate receptor gene appeared to be overexpressed in nonfunctioning adenomas and underexpressed in PRL- and GH-secreting adenomas, ornithine decarboxylase gene was overexpressed in GH-secreting adenomas and underexpressed in ACTH-secreting adenomas, and C-mer protooncogene tyrosine kinase gene was overexpressed in ACTH-secreting adenomas and underexpressed in PRL-secreting adenomas (11). Although the precise role of these genes in pituitary carcinogenesis is still unknown, the detection of candidate genes emerging from cDNA expression studies provides a rationale for investigating the possible oncogenic potential of these genes in the pituitary.
The third study (12) appears in the current issue of Endocrinology and reports gene expression profile in the TtT-97 tumor, a well differentiated mouse pituitary adenoma that synthesizes and secretes TSH and has retained a physiological response to thyroid hormone. The tumor is propagated by dispersed TtT-97 tumor tissue injection into radio-thyroidectomized mice. The implanted tumor progressively enlarges and secretes increasing amounts of TSH. When mice with established tumors are supplemented with thyroid hormone, there is a reduction of TSH secretion and a gradual involution of the tumor, both events being reversible upon withdrawal of thyroid hormone. This animal model is reminiscent of the so-called feedback tumors occurring in hypothyroid patients. Indeed, the possibility of increased size of the pituitary in primary hypothyroidism in humans was recognized for the first time by Niepce in 1851 at the autopsy of a cretin (16). A systematic autopsy study of 64 patients with primary hypothyroidism showed diffuse and nodular thyrotrope hyperplasia in 69% and 25% of cases, respectively, with a true adenoma in 18% (17). Interestingly, as it occurs in TtT-97 tumors, thyroid hormone replacement is generally efficient in reversing pituitary hyperplasia and adenoma in humans (17). Although all this evidence points to a direct causal relationship between thyroid hormone and thyrotroph proliferation, the molecular mechanisms responsible for the formation of pituitary tumors in the absence of thyroid hormone, and their shrinkage and disappearance following replacement therapy, are currently unknown (17). The effects of thyroid hormone are mediated by T3 nuclear receptors 
and ß (TR and TRß) that are encoded by two different genes and expressed as several isoforms. TRs bind to thyroid hormone-responsive elements in the promoter of target genes, whose transcription can be positively or negatively regulated. In recent years, using cDNA microarrays, the number of thyroid hormone-regulated genes has been documented to be much higher than previously thought. By studying the effect of thyroid hormone in the rat hepatocyte, 55 of 2,225 genes, representing approximately 10% of the liver transcriptosome, were found to be importantly affected by thyroid hormone (18). Consistent with some of the known metabolic effects of thyroid hormones, T3 affected gene expression for a diverse range of cellular pathways and functions, including gluconeogenesis, lipogenesis, insulin signaling, and apoptosis. Moreover, thyroid hormone was effective in inducing the expression of genes involved in cell proliferation, such as Bcl3, which serves as a coactivator for several nuclear factors, the glycosylphosphatidyl-inositol-linked protein B61, which binds Eck receptor protein-tyrosine kinase and acts as a mitogen for hepatocyte, and Kip 1p, which participates in the segregation of chromosomes during mitosis. All these effects are consistent with previous reports indicating an important role of thyroid hormone in inducing hepatocyte proliferation and cell survival (18).
The gene expression profile induced by thyroid hormone in the liver is expected to be different from that induced in the pituitary where thyroid hormone acts as the main negative regulator of thyrotroph function and proliferation. Indeed, a new pattern of pituitary gene expression has been provided by monitoring gene profile in the TtT-97 tumor (12, 19, 20). In this tumor, the nature of the genes involved in thyroid hormone-induced growth arrest was previously investigated by Ridgway’s group using conventional RNA Northern blot analysis of candidate genes (19, 20). From that study, it appeared that TtT-97 tumors actively growing under conditions of profound hypothyroidism do not express somatostatin receptor mRNA and that the reduction of tumor expansion induced by thyroid hormone is associated with the specific up-regulation of somatostatin type 1 and type 5 receptor (sst1 and sst5) expression (20). By the same experimental approach, a thyroid hormone-induced up-regulation of TRß1 accompanied by a reduction of TRß2 and TR mRNA was also observed (19). The study published in this issue of Endocrinology (12) further characterizes thyroid hormone induced changes of gene expression in TtT-97 tumors, by expanding the number of analyzed genes. The expression profile of more than 1,000 genes in tumors obtained from mice 24 h after vehicle or thyroid hormone administration was evaluated by combining cDNA microarray and Northern blot analysis (12). Of the 1,176 genes analyzed, 7 resulted to be up-regulated and 40 down-regulated by thyroid hormone, whereas a good proportion of genes, particularly those within the cell cycle regulator category, fell below the limits of detectability on the microarray and were, therefore, analyzed by Northern blot. The study confirms that sst5 is direct target for thyroid hormone because the transcript was up-regulated within few hours (12, 20). The somatostatinergic pathway has been demonstrated to be relevant for pituitary tumorigenesis also in humans. Indeed, low levels of sst2 and sst5 transcripts, together with rare mutations preventing the antiproliferative action of sst5, have been reported in a subset of GH-secreting adenomas characterized by a poor responsiveness to somatostatin analogs (21, 22). Although the levels of sst transcripts have not been quantified in human TSH-secreting adenomas, the observation that somatostatin analogs are effective in reducing or even normalizing TSH secretion in the majority of patients and in shrinking the tumor mass in a half (17) suggests that the expression of sst may contribute to maintain the low growth rate characterizing human pituitary tumors and that their up-regulation may be associated with growth arrest in experimental tumors, such as TtT-97 (12).
Another gene found to be up-regulated by thyroid hormone was TRß (12, 19). The role of defective negative feedback due to TR abnormalities in pituitary tumorigenesis has been recently put forward. In particular, a decreased expression of both TR and TRß has been observed in two TSH-secreting adenomas, whereas in another tumor an aberrant alternative splicing variant of TRß2 mRNA-lacking hormone binding domain and impairing the thyroid hormone-dependent inhibition of TSHß and -subunit transcription has been identified (23, 24). These abnormalities support the hypothesis that the disruption of TRs signaling may be an additional mechanism for defective negative regulation of both TSH secretion and thyrotrope growth, and are consistent with the observation that up-regulation of TRs is an early event associated with TtT-97 tumor involution induced by thyroid hormone (12).
Moreover, in TtT-97 tumors, thyroid hormone decreased the expression of genes encoding receptors and ligands with mitogenic potential. In particular, a significant decrease of TRH receptor 1 (TRHR1) transcript was caused by thyroid hormone (12). This observation is in keeping with the view that alterations of hypothalamic neurohormone signaling may be relevant to pituitary tumor formation (1, 2, 3, 4). This phenomenon is well documented for GHRH and CRH, which, if overproduced, cause acromegaly and Cushing’s disease, respectively, while a role for TRH in excess in thyrotrope proliferation remains to be demonstrated (1, 3, 4). However, the large majority of human pituitary adenomas abnormally express TRH receptors, which, although intact in their coding sequence (25), activate intracellular transduction pathways with mitogenic potential, such as the PKC-dependent pathway. The report of a direct association between thyrotrope tumor involution and TRHR1 gene down-regulation strongly supports a role of TRH in pituitary cell growth.
In addition to down-regulating TRHR1, thyroid hormone caused a significant reduction of the neurotrophin brain-derived neurotrophic factor (BDNF), as well as its receptor trkB in TtT-97 tumor, in analogy with what was observed in the central nervous system (26). BDNF is a neuronal survival factor widely expressed in the hypothalamus and pituitary. In particular, this factor is specifically present in thyrotropes where it probably colocalizes with nerve growth factor (27). As far as the known effects of these neurotrophins are concerned, much evidence suggests that nerve growth factor cooperates in directing lactotrope differentiation by stimulating PRL synthesis and D2 dopaminergic receptor expression, whereas the functional significance of BDNF in the pituitary is still unknown (4). Interestingly, in the hypothalamus BDNF is required for the early expression of TRH neurons and in the cerebral cortex this factor induces somatostatin gene expression, probably by up-regulating VIP expression (28, 29).
Among the different genes regulated by thyroid hormone, a few other genes showed an early response, consistent with a transcriptional effect of thyroid hormone, and were selectively expressed in thyrotropes. In particular, within 2 h thyroid hormone up-regulated neuronatin, a member of the "proteolipid" class of proteins that functions as regulator of ion channels during brain development and down-regulated PB cadherin, a member of the cell adhesion proteins responsible for the Ca2+-dependent cell adhesion (30, 31). Although their action on thyrotrope growth is still undefined, neuronatin and PB cadeherin, together with sst5, TRHR1, BDNF, and trkB, are excellent candidates for thyrotrope-restricted primary targets for thyroid hormone.
The study also reports the levels of expression of cell cycle regulators, such as cyclin-dependent kinases (cdks), their regulatory activators (cyclin), and inhibitors (cdkis), which control the phosphorylation state of retinoblastoma protein Rb. From gene expression analysis, a model for thyroid hormone-induced inhibition of cell growth is envisaged whereby the expression of cyclinA/cdk2 complexes, which are required for maintenance of S phase progression via Rb phosphorylation, is down-regulated and cdk inhibitors, such as p15, are up-regulated. This pattern of expression is consistent with the general view that Rb phosphorylation is the key step in cell cycle initiation and progression, whereas Rb dephosphorylation results in growth arrest. Moreover, transcript levels of other cell cycle proteins, such as the transcription factor E2F1, which is activated by phosphorylated Rb, thus resulting in DNA replication, showed a decreased expression. However, the genes coding these cell cycle proteins are not direct targets for thyroid hormone because changes in transcript levels were late effects. This is consistent with the notion that thyroid hormone inhibits the proliferation of a specific cell type, i.e. thyrotrope, while the growth of most cells that express both TRß and cell cycle regulators is not affected by thyroid hormone.
Finally, other genes not previously reported to be expressed in the pituitary or to be the target for thyroid hormone were identified using cDNA microarray strategy. These genes belong to the categories of cell adhesion proteins (such as cadherin 4 and 
-catenin), intracellular transducers (such as calmodulin, adenylyl cyclase type 6, and the nonclassical tyrosine kinase ryk), extracellular signaling (such as -subunit of inhibin, chromogranin B and C), transcription factors (such as neurogenin 3), apoptosis-associated proteins (such as DAD1), and metabolic pathways (such as prohormone convertase PC1). Most of them were down-regulated by thyroid hormone.
This study enhances our awareness of the large repertoire of genes that are regulated by thyroid hormone in the pituitary and provides new insights into the molecular mechanisms that may ultimately lead to pituitary tumor formation. This study confirms that proteins involved in cell signaling are key components of the multistep process of pituitary tumorigenesis as growth arrest induced by thyroid hormone seem to involve primarily the transcription of genes coding for growth factors and receptors. Indeed, in this reversible model of thyrotrope tumorigenesis tumor involution occurred via induction of receptors that mediate inhibitory signals, i.e. sst5 and TRß, as well as repression of specific growth factors and their receptors, i.e. BDNF, TRHR1, trkB. Moreover, this study confirms that cDNA microarray methodology is a powerful strategy to identify and to select for future studies genes previously ascribed to other domains of interest. Further research will tell us whether the new candidate genes emerged from cDNA expression study play a role on thyrotrope proliferation. Nonetheless, it is worth noting that this strategy resulted to be disappointingly inadequate to identify important genes relevant to cell proliferation that were below the levels of detectability.
|
Footnotes |
|---|
Received November 28, 2001.
Accepted for publication November 30, 2001.
|
References |
|---|
|
TopIntroductionReferences |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
