Thyroid hormone-responsive pituitary hyperplasia independent of somatostatin receptor 2.

Mice homozygous for the targeted disruption of the glycoprotein hormone alpha-subunit (alphaGsu) display hypertrophy and hyperplasia of the anterior pituitary thyrotropes. Thyrotrope hyperplasia results in tumors in aged alphaGsu(-/-) mice. These adenomatous pituitaries can grow independently as intrascapular transplants in hypothyroid mice, suggesting that they have progressed beyond simple hyperplasia. We used magnetic resonance imaging to follow the growth and regression of thyrotrope adenomatous hyperplasia in response to thyroid hormone treatment and discovered that the tumors retain thyroid hormone responsiveness. Somatostatin (SMST) and its diverse receptors have been implicated in cell proliferation and tumorigenesis. To test the involvement of SMST receptor 2 (SMSTR2) in pituitary tumor progression and thyroid hormone responsiveness in alphaGsu(-/-) mutants, we generated Smstr2(-/-), alphaGsu(-/-) mice. Smstr2(-/-), alphaGsu(-/-) mice develop hyperplasia of thyrotropes, similar to alphaGsu(-/-) mutants, demonstrating that SMSTR2 is dispensable for the development of pituitary adenomatous hyperplasia. Thyrotrope hyperplasia in Smstr2(-/-), alphaGsu(-/-) mice regresses in response to T4 treatment, suggesting that SMSTR2 is not required in the T4 feedback loop regulating TSH secretion.

Functional interactions of an upstream enhancer of the mouse glycoprotein hormone alpha-subunit gene with proximal promoter sequences.

Transcription of the glycoprotein hormone alpha-subunit gene in the pituitary is governed by different promoter elements in thyrotropes and gonadotropes. We recently identified an upstream enhancer that directs a high level of cell type specific expression in transgenic mice and stimulates proximal promoter activity in cultured alphaTSH and alphaT3 cells. To assess the contribution of promoter sequences that functionally interact with the enhancer, we mutated two proximal elements shown to be important in both thyrotrope and gonadotrope cells. Disruption of the pituitary glycoprotein hormone basal element (PGBE), which binds a LIM homeodomain protein, resulted in a decrease in basal promoter activity in both alphaTSH and alphaT3 cells. Enhancer function was completely abolished by the PGBE site mutation in alphaT3 gonadotropes, whereas some stimulatory activity remained in alphaTSH thyrotropes. Mutation of the gonadotrope specific element (GSE), which binds SF1 and is important for basal activity in gonadotropes and TRH response in thyrotropes, resulted in declines in basal and enhanced promoter activity only in alphaT3 cells and not in alphaTSH cells. Despite this decrease in enhanced activity, the GSE mutated promoter still retained some enhancer stimulated activity, suggesting that the PGBE site still functionally interacts in the absence of an intact GSE. This mutation had no effect in alphaTSH cells. These data suggest that although the enhancer works in both cell types it exhibits cell type specific functional characteristics.

Cell-specific expression of the mouse glycoprotein hormone alpha-subunit gene requires multiple interacting DNA elements in transgenic mice and cultured cells.

The glycoprotein hormone alpha-subunit gene is expressed and differentially regulated in pituitary gonadotropes and thyrotropes. Previous gene expression studies suggested that cell specificity may be regulated by distinct DNA elements. We have identified an enhancer region between -4.6 and -3.7 kb that is critical for high level expression in both gonadotrope and thyrotrope cells of transgenic mice. Fusion of the enhancer to -341/+43 mouse alpha-subunit promoter results in appropriate pituitary cell specificity and transgene expression levels that are similar to levels observed with the intact -4.6 kb/+43 construct. Deletion of sequences between -341 and -297 resuited in a loss of high level expression and cell specificity, exhibited by ectopic transgene activation in GH-, ACTH-, and PRL-producing pituitary cells as well as in other peripheral tissues. Consistent with these results, transient cell transfection studies demonstrated that the enhancer stimulated activity of a -341/+43 alpha-promoter in both alphaTSH and alphaT3 cells, but it did not enhance alpha-promoter activity significantly in CV-1 cells. Removal of sequences between -341 and -297 allowed the enhancer to function in heterologous cells. Loss of high level expression and cell specificity may be due to loss of sequences required for binding of the LIM homeoproteins or the alpha-basal element 1. These data demonstrate that the enhancer requires participation by both proximal and distal sequences for high level expression and suggests that sequences from -341 to -297 are critical for restricting expression to the anterior pituitary.

Activation of the glycoprotein hormone alpha-subunit gene promoter in thyrotropes.

Pituitary expression of the glycoprotein hormone alpha-subunit gene localizes to thyrotropes and gonadotropes. Factors that activate the alpha-subunit promoter exclusively in thyrotropes have not been described. In the current studies, alpha-subunit promoter activity was compared in alphaTSH thyrotropic and alphaT3 gonadotropic cells. Deletional analysis revealed a more important contribution of the -453/-381 region in thyrotropic cells and the -341/-297 region in gonadotropic cells, while other deletions had similar effects. Mutational analysis revealed some regions that were functionally similar, while others were active only in thyrotropic cells, such as the regions - 434/ - 421, - 398/ - 385, - 376/- 364 and - 363/- 351, that decreased activity by approximately 2-fold each. Combined mutation of the regions - 434/ - 421, - 398/ - 385 and - 376/ - 364 decreased activity by 5-fold in thyrotropic cells and not at all in gonadotropic cells. Cotransfection with Ptx1 in CV-1 cells resulted in a strong stimulation of alpha-subunit promoter activity, that was significantly diminished with mutation of the - 398/ - 385 region that disrupts the Ptx1 binding site, suggesting that this factor may play a role in thyrotrope-specific activation. This analysis provides basic information important to identify common and unique factors contributing to the cell-specific expression of the alpha-subunit gene.

Msx1 is present in thyrotropic cells and binds to a consensus site on the glycoprotein hormone alpha-subunit promoter.

Our studies are aimed at identifying the transcription factors that activate the glycoprotein hormone alpha-subunit promoter. Therefore, we performed a Southwestern screening of a thyrotropic (alphaTSH) cDNA expression library, using the region of the promoter from -490 to -310 that contains sequences critical for expression in thyrotrope cells. A clone was isolated corresponding to part of the coding sequence of Msx1, which is a homeodomain-containing transcription factor that has been found to play an important role in the development of limb buds and craniofacial structures. Northern blot analysis, using the cloned Msx1 cDNA fragment as a probe, demonstrated that alpha-subunit-expressing thyrotrope cells (alphaTSH cells and TtT97 tumors) contained Msx1 RNA transcripts of 2.2 kb, while somatomammotrope (GH3) cells that do not produce the alpha-subunit had barely detectable levels. The presence of Msx1 protein was demonstrated by Western blot analysis in alphaTSH cells. We also demonstrated that transcripts encoding the closely related Msx2 factor were not detectable by Northern blot analysis in either thyrotrope or somatomammotrope-derived cells. Subfragments of the region from -490 to -310 of the alpha-subunit promoter were used in a Southwestern blot assay using bacterially produced Msx1 and demonstrated that binding was localized specifically to the region from -449 to -421. Deoxyribonuclease I protection analysis, using purified Msx1 homeodomain, demonstrated structurally induced differences in DNA digestion patterns between -436 and -413, and sequence analysis of this region revealed a direct repeat of the sequence GXAATTG, which is similar to the Msx1 consensus-binding site. When nucleotides at both sites were mutated, Msx1 binding was dramatically reduced, and the activity of an alpha-subunit promoter construct decreased by approximately 50% in transfected thyrotrope (alphaTSH) cells. These studies suggest that Msx1 may play a role in the expression of the alpha-subunit gene in thyrotrope cells.

Reconstitution of triiodothyronine inhibition in non-triiodothyronine-responsive thyrotropic tumor cells using transfected thyroid hormone receptor isoforms.

The triiodothyronine (T3) inhibitory effect on the thyrotropin (TSH)beta- and alpha-subunit genes is believed to be mediated by binding of T3 to specific nuclear receptors that are present in various isoforms. alphaTSH cells, which are derived from a pure alpha-subunit secreting thyrotropic tumor, contain the same nuclear factors that are important for alpha-subunit gene expression in TSH-expressing T3-responsive thyrotropic cells (TtT97). However, as in the parent tumor, alpha-subunit expression in alphaTSH cells was not inhibited by T3, despite the presence of high-affinity nuclear T3 receptors (TRs) with a similar number of sites per cell as in TtT97. When transcripts coding for the different TR isoforms from the MGH101A tumor were analyzed by Northern blot, TR alpha1 was present, as well as the non-T3-binding variant alpha2, but transcripts encoding the opposite strand Rev-ErbAa were not detectable and neither TR beta1 nor TR beta2 mRNAs were detectable, whereas all these transcripts were detectable in TtT97 tumors. Similar findings were observed in alphaTSH cells, where TR beta1 transcripts were barely detectable in Northern blots and TR beta2 transcripts were detectable only by RT-PCR. The TR beta gene locus is present and unrearranged in the tumor genome. In transient transfection studies conducted in alphaTSH cells overexpression of either TR beta1, TR beta2, or TR alpha1 reconstituted T3-inhibition of the alpha-subunit promoter down to 40% to 50% of control. We conclude that the relative lack of TR beta gene expression correlates with unresponsiveness to T3. The alphaTSH cell line represents a unique model in which to dissect the mechanism of T3 inhibition.

Thyroid hormone-induced expression of specific somatostatin receptor subtypes correlates with involution of the TtT-97 murine thyrotrope tumor.

Following the protracted hypothyroid state, treatment with thyroid hormone will induce a decline in TSH and reduce thyrotrope hyperplasia. Somatostatin is a hypothalamic peptide that has been implicated in the negative regulation of TSH secretion in the thyrotrope. Moreover, analogs of native somatostatin have potent TSH-reducing and growth-retarding effects on human thyrotropinomas. The TtT-97 tumor is an in vivo murine thyrotropic model that has retained its physiological response to thyroid hormone. This study investigates the regulation of somatostatin receptor subtypes in this tumor. TtT-97 tumors, actively growing in hypothyroid mice, did not express any significant somatostatin receptor messenger RNA (mRNA) or protein. T4 administration resulted in a reduction in TSH beta mRNA expression and a marked degree of tumor involution. Analysis of residual tumors from thyroid hormone-treated mice showed the specific up-regulation of SSTR1 and SSTR5 mRNA subtypes and the appearance of abundant, high affinity SSTR receptor-binding sites within the tumor. Thus, the TtT-97 tumor provides a thyrotrope-specific model in which to study the regulation of somatostatin receptor subtypes by thyroid hormone and correlate this expression with both antisecretory and antiproliferative effects.

Enhancer-mediated high level expression of mouse pituitary glycoprotein hormone alpha-subunit transgene in thyrotropes, gonadotropes, and developing pituitary gland.

The pituitary hormones LH, FSH, and TSH are heterodimers composed of a common alpha-subunit and unique beta-subunits. We demonstrate that 4.6, 2.7, 1.49 or 0.48 kilobases (kb) mouse alpha-subunit 5'-flanking sequences are sufficient for transgene expression in both gonadotropes and thyrotropes but not in inappropriate pituitary cells. In contrast, transgenes with bovine or human alpha-subunit flanking sequences have been shown to confer reporter gene expression only to gonadotrope cells, suggesting that the elements regulating cell-specific expression may differ between species. Equal levels of reporter gene expression were conferred by 5.0 and 0.48 kb in transiently transfected thyrotrope tumor-derived cells. In contrast, in transgenic mice, high level expression was only obtained with 4.6 kb 5'-flanking sequences, indicating the presence of an enhancer element between 4.6 and 2.7 kb. The 4.6 kb of 5'-flanking sequences are sufficient for both hormonal and developmental regulation of transgene expression. Mice rendered hypothyroid by radiothyroidectomy had significantly higher levels of transgene expression than either hyperthyroid or euthyroid animals. The temporal and spatial pattern of transgene expression in Rathke's pouch paralleled that of the endogenous gene; the onset of transgene expression occurred by embryonic day 9.5. Low level expression of both the transgene and the endogenous alpha-subunit gene were detected in some unexpected peripheral sites, such as the embryonic extraocular and olfactory regions, suggesting that alpha-subunit may have a more diverse role in development than previously considered.

Analysis of Pit-1 in regulating mouse TSH beta promoter activity in thyrotropes.

TSH beta gene expression is restricted to pituitary thyrotropes. Since Pit-1 is present in these cells, we characterized Pit-1 RNA and protein in thyrotropes, and tested its function in regulating TSH beta promoter activity. We demonstrate that both TtT-97 thyrotropic tumors and pituitaries contain four Pit-1 transcripts of 3.2, 2.6, 2.4, and 1.9 kb, respectively. Only two transcripts of 2.7 and 2.1 kb were detected in alpha TSH cells, a thyrotrope derived cell that no longer expresses TSH beta. Western analysis revealed Pit-1 protein in TtT-97 cells but not in alpha TSH cells. DNase I protection assays localized Pit-1 binding to three areas of the mouse TSH beta promoter. However, basal TSH beta promoter activity was minimally stimulated when alpha TSH cells or TtT-97 thyrotropes were co-transfected with mouse Pit-1 and a mTSH beta luciferase construct. These studies suggest that Pit-1 is not limiting for cell-specific expression of the TSH beta gene in thyrotrope-derived cells and implies that additional thyrotropic factors are likely required.

Effect of thyroid hormone on T3-receptor mRNA levels and growth of thyrotropic tumors.

Thyrotropic tumors (TtT97) contain mRNA transcripts for three T3-receptor (TR) isoforms, alpha 1, beta 1 and beta 2, and a non-receptor alpha 2-variant. We administered T4 (5 mg/l of drinking water) for one month to TtT97-bearing mice, to examine its effect on tumor growth and tumor TR isoform steady-state mRNA levels. Baseline mice were killed at the start of the experiment, and placebo mice were maintained hypothyroid. The treated tumors were 30-35% smaller than the baseline tumors (p = NS), while the placebo tumors were 2- to 7-fold larger than the baseline tumors (p < 0.05). TR beta 1 mRNA increased 5- to 6-fold, while TR beta 2 mRNA decreased by 76%. TR alpha 1 and the alpha 2-variant decreased by 52% and 70%, respectively. Therefore, the tumors decreased their growth rate in response to T4 administration, and increased the ratio of TR beta 1 to TR beta 2 mRNA. This raises the intriguing possibility of a correlation between the relative abundance of the TR beta isoforms and tumor growth.

A cell line that produces the glycoprotein hormone alpha-subunit contains specific nuclear factors similar to those present in thyrotropes.

A unique characteristics of thyrotrope-specific gene expression is the coordinated expression and regulation of the alpha- and beta-subunits of TSH. A cell line (alpha TSH) derived from the transplantable mouse thyrotropic tumor MGH101A, which no longer expresses the TSH beta-subunit gene but continues to secrete large amounts of alpha-subunit, was used as a model to study alpha-subunit gene expression independent from the TSH beta-subunit gene and was compared with the expression in TSH-secreting TtT97 tumors. Transient transfection studies showed a striking similarity in the activity of 5' deletions of the mouse alpha-subunit gene promoter in both alpha TSH and TtT97 cells and localized two regions important for expression that spanned 100 base pairs, from -480 to -417 and from -417 to -381. These regions were found to have no activity in nonthyrotrope pituitary GH4 cells and L-cell fibroblasts. Analysis of the alpha-subunit 5' flanking DNA interactions with alpha TSH and TtT97 nuclear extracts showed two DNase I protected sequences, from -474 to -452 and from -447 to -400, both of which colocalized with the functionally important regions. Gel retardation analysis demonstrated the specificity of these interactions, and a similar migration of the DNA-protein complexes suggested that protein factors were similar in the two cell types. We conclude that the nuclear factors necessary for alpha-subunit expression in thyrotropes are retained in alpha TSH cells. Moreover, since alpha TSH cells do not express the TSH beta-subunit gene, the factors that determine the expression of the alpha-subunit may not be sufficient for TSH beta-subunit gene expression.

Thyrotrope expression and thyroid hormone inhibition map to different regions of the mouse glycoprotein hormone alpha-subunit gene promoter.

The alpha-subunit gene of the glycoprotein hormones is normally expressed in pituitary thyrotropes and gonadotropes and in placental cells. Thus, this gene must contain elements that mediate expression and hormonal responses in different cell types. The localization of DNA regions important for expression and regulation of the alpha-subunit gene in thyrotrope cells has not previously been reported. In these studies luciferase expression constructs containing 1700 basepairs of 5' flanking DNA derived from the mouse alpha-subunit gene were introduced by electroporation into freshly dispersed cells from TSH-producing mouse pituitary tumors (TtT 97). This promoter functioned with greater efficiency in thyrotropes than in nonthyrotrope pituitary GH4 cells and L-cell fibroblasts. Primer extension confirmed that transcription from the alpha-subunit constructs initiated at the same site as the endogenous gene. Studies using 5' truncations showed a progressive loss of alpha-subunit promoter activity in thyrotropes between -480 and -120, with regions upstream of -254 contributing substantially to expression in thyrotrope cells. Thyroid hormone inhibited alpha-subunit promoter activity in a dose-dependent fashion, although in vivo treatment of tumors with thyroid hormone before transfection was necessary to achieve maximal inhibition. Thyroid hormone inhibition of alpha-subunit promoter activity also occurred in GH4 cells, but no effect was observed in L-cells. Studies using 5' truncations localized a region responsible for thyroid hormone inhibition between -62 and +43, encompassing the TATA sequence and the transcriptional initiation site. When this region was compared to the thyroid hormone inhibitory regions of the alpha-subunit genes from other species and the mouse TSH beta-subunit gene, a 6-basepair motif, 5' (G/A)GTG(G/A)G 3', emerged as a possible consensus sequence for a thyroid hormone inhibitory element.

TSH subunit gene promoters from a murine alpha-subunit producing tumor function normally.

The murine thyrotropic MGH101A tumor is characterized by absent thyrotropin (TSH) beta gene expression and altered thyroid hormone (T3) regulation of the alpha-subunit. Comparison of the promoter structures of both alpha and TSH beta subunit genes from MGH101A with the promoter in expressing TtT-97 thyrotropes revealed no detectable differences. Transfection of the TSH beta promoter from MGH101A linked to luciferase showed minimal expression in primary or cloned MGH101A cells, or L-cells. However, a 6- to 10-fold increase in expression was exhibited in transfected thyrotropes. For the alpha gene, promoter activity was highest in thyrotropes and in cloned MGH101A cells, 5-fold lower in MGH101A tumors, and 10-fold lower in L-cells. Both promoters were not substantially affected by T3 treatment in MGH101A cells. In thyrotropes, promoter activity was inhibited 62.5% and 57.7% by 10 nM T3 treatment for the TSH beta and alpha genes, respectively. DNase I protection showed that factors from TtT-97 but not from MGH101A cells interacted with regions in the TSH beta promoter, while nuclear extracts from each tumor demonstrated at least one protein-DNA interaction with the alpha-subunit promoter. These studies suggest that the molecular defects in the MGH101A tumor are related to the absence of trans-acting factors and are not a result of altered primary gene structure.

Identification of cis-acting promoter elements important for expression of the mouse glycoprotein hormone alpha-subunit gene in thyrotropes.

The glycoprotein hormone alpha-subunit gene is expressed in a cell-specific manner in the anterior pituitary and placenta. Previous studies have shown that the region between -178 to -111 is indispensable for placental-specific expression of the human alpha-subunit gene. Using gene transfer techniques with chimeric luciferase plasmids, this report identifies regions of the mouse alpha-subunit promoter that are important for transcriptional activation in primary thyrotropic cells. Transient expression of a series of 5' flanking DNA deletions resulted in stepwise reductions of basal promoter activity between -480 to -417 (4-fold), -254 to -177 (5-fold), and -177 to -120 (3.5-fold). DNase-I protection analysis with nuclear extracts from thyrotropic tumor cells revealed specific protein-DNA interactions within each of these functionally defined regions. These were mapped to positions -474 to -452, -447 to -419, -213 to -170, and -158 to -101 within the 5' flanking region. In contrast, in mouse fibroblast L-cells no significant difference in alpha-subunit promoter activity was found by deleting the region from -480 to -177. However, a 3-fold decrease, similar to that found in primary thyrotropes, was found by deleting the region from -177 to -120. Further, a smaller region between -138 and -122 was the only area detected by the DNase-I protection assay using L-cell nuclear extracts. Thus, several cis-acting promoter elements located up-stream of position -177 are important for expression in thyrotropes. These elements also bind nuclear factors present in thyrotropes but not in nonpituitary fibroblasts and, therefore, differ from those mediating expression of the human alpha-subunit gene in the placenta.