Regulation of glycoprotein hormone alpha-subunit gene expression

The appropriate, regulated expression of the glycoprotein hormone subunit genes is required to enable the biosynthesis of luteinizing hormone, follicle-stimulating hormone, thyroid-stimulating hormone, and chorionic gonadotropin. We have focused our attention on mechanisms mediating regulated transcription of the common alpha-subunit gene. Our studies have examined both the signaling mechanisms and the DNA elements and transcription factors that are important for alpha-subunit expression. Our initial efforts involved characterization of DNA elements of the alpha-subunit gene important for basal and GnRH-stimulated expression. Clustered point mutation analysis identified two different, unrelated sequences that play a role in alpha-subunit transcription. When tested as multiple copies on a minimal promoter, one of these elements was sufficient to permit a response to GnRH, while the other enhanced basal expression. Therefore, we designated these DNA elements as the GnRH-response element (GnRH-RE) and the pituitary glycoprotein hormone basal element (PGBE), respectively. The GnRH-RE contains a consensus binding site for the Ets family of transcription factors. As several Ets factors have been shown to mediate transcriptional responses to the mitogen-activated protein kinase (MAPK) pathway, we investigated the possibility that GnRH effects on alpha-subunit transcription may involve the MAPK cascade. We found that GnRH can indeed activate MAPK and that MAPK activation is sufficient and necessary for transcriptional activation of the alpha-subunit gene. Efforts to further characterize proteins that interact with the PGBE led to the cloning of a LIM-homeodomain transcription factor designated LH-2. Recombinant LH-2 selectively binds to the PGBE in vitro. Transfection experiments have shown that an expression vector for LH-2 can activate the alpha-subunit promoter in heterologous cells. LH-2 appears to be a component of the endogenous factors that bind to the PGBE. Thus, LH-2 appears to be an excellent candidate as a factor responsible for basal expression of the alpha-subunit gene. Overall, these studies have contributed to identification of molecular components important for regulated expression of the glycoprotein hormone alpha-subunit gene.

Activation of the glycoprotein hormone alpha-subunit promoter by a LIM-homeodomain transcription factor

Recently, a pituitary-specific enhancer was identified within the 5' flanking region of the mouse glycoprotein hormone alpha-subunit gene. This enhancer is active in pituitary cells of the gonadotrope and thyrotrope lineages and has been designated the pituitary glycoprotein hormone basal element (PGBE). In the present studies, we sought to isolate and characterize proteins which interact with the PGBE. Mutagenesis experiments identified a 14-bp imperfect palindrome that is required for binding of a factor which is present in cells of gonadotrope and thyrotrope lineages but not in other cells. Screening of a mouse cDNA library with a DNA probe containing the imperfect palindrome resulted in the isolation of a LIM-homeodomain transcription factor. The cDNA predicts a mouse protein which is 94% identical to the recently described rat LIM-homeodomain protein LH-2. LH-2 contains two zinc fingers (LIM domain) and a consensus homeodomain. Hybridization analysis revealed relatively high expression of LH-2 mRNA in the central nervous system and in pituitary cells of the gonadotrope and thyrotrope lineages. Lower or nondetectable levels of LH-2 mRNA were found in other pituitary cells and tissues, including placental cells. Recombinant LH-2 homeodomain was found to selectively bind to the previously identified imperfect palindrome in the PGBE. Point mutations in the PGBE resulted in parallel losses in the binding of a nuclear factor from a cell line of the gonadotrope lineage and recombinant LH-2-binding activity. Use of an antibody to LH-2 provided evidence that endogenous PGBE-binding activity from cells of the gonadotrope lineage involves a protein which is immunologically related to LH-2. Expression of LH-2 in two heterologous cell types resulted in activation of a reporter gene containing the mouse alpha promoter. These data suggest that the LIM-homeodomain factor LH-2 plays a role in stimulating tissue-specific expression of the mouse glycoprotein hormone alpha subunit. The finding that a LIM-homeodomain protein can stimulate expression of one of the earliest markers of pituitary differentiation raises the possibility that this factor plays a role in cell lineage determination in the pituitary.

Two different DNA elements mediate gonadotropin releasing hormone effects on expression of the glycoprotein hormone alpha-subunit gene

Transient transfection studies using gonadotrope-derived, alpha T3-1 cells were used to determine the DNA sequences of the mouse glycoprotein hormone alpha-subunit gene that mediate the transcriptional response to gonadotropin releasing hormone (GnRH). The roles of phorbol esters and cyclic AMP in mediating the GnRH response were also investigated. The initial studies demonstrated that a construct containing approximately 500 base pairs of alpha-subunit flanking sequence was sufficient to mediate responses to a GnRH agonist (GnRHa), phorbol myristate acetate (PMA) and a cAMP analog. Responses to combinations of cAMP and GnRHa or cAMP and PMA were approximately additive, whereas the response to the combination of GnRHa and PMA was similar to that seen with either of the agents alone. Cotransfection studies with an expression vector for the heat-stable inhibitor of the cAMP-dependent protein kinase demonstrated that GnRHa and PMA responses are not dependent on the cAMP-dependent kinase. Deletion analysis indicated that sequences between -507 and -205 were involved in mediating responses to GnRHa and PMA. To determine if this region alone could support responses to these agents, the -507 to -205 region was linked to a minimal promoter and tested in transient transfections. The results demonstrated that this region supports responses to GnRHa, PMA, and cAMP. Clustered point mutations of this region were used to further characterize sequences involved in the GnRH response. Mutations in two regions, one at positions -406 to -399 and one at positions -337 to -330, resulted in decreased responses to GnRH and PMA. There is no obvious sequence similarity between the two regions that are required for the GnRH response. An enhancer test demonstrated that multimers of the -416 to -385 region were able to function as a GnRH-responsive element when linked to a minimal promoter, although a single copy of this region was not sufficient to permit a response to GnRH. In contrast, multimers of the -344 to -300 region did not permit a response to GnRH, but enhanced basal transcription. These findings are consistent with the identification of a two-component GnRH response unit, which probably involves the functional cooperation of two different transcription factors. The observation that GnRH responsiveness appears to co-localize with PMA responsiveness suggests that GnRH effects on the alpha-subunit transcription are likely mediated by the protein kinase C pathway.

Phosphorylation of cyclic adenosine 3',5'-monophosphate (cAMP) response element-binding protein isoforms by the cAMP-dependent protein kinase

The cAMP response element-binding protein (CREB) mediates transcriptional activation of genes in response to the cAMP signal transduction pathway. There are two different isoforms of CREB, which are generated by alternative RNA splicing. There is evidence that the two isoforms may have different biological activities. As the longer isoform (CREB341) contains a potential phosphorylation site that is not present in the shorter isoform (CREB327), we examined the possible differential phosphorylation of the two CREB isoforms. Recombinant CREB was prepared and used as substrate for phosphorylation by the cAMP-dependent protein kinase in vitro. Phosphopeptide mapping and mutagenesis studies demonstrated that CREB341 contains two sites, serine 133 and serine 98, that can be phosphorylated in vitro by the catalytic subunit of the cAMP-dependent protein kinase. In contrast, CREB327 contains only a single phosphorylation site at serine 119 (equivalent position to serine 133 in CREB341). A kinase titration experiment demonstrated that serine 98 of CREB341 was phosphorylated only at relatively high concentrations of the cAMP-dependent protein kinase. Transient transfection studies were used to test for any possible function of the phosphorylation of serine 98 of CREB341. These studies used GAL4-CREB fusion proteins. We found that mutation of serine 98 to alanine (which would block phosphorylation) has little or no effect on the ability of the CREB fusion protein to activate transcription. These findings suggest that differences in the biological activity of the two CREB isoforms are probably not mediated by differential phosphorylation by the cAMP-dependent protein kinase.

Analysis of DNA sequences required for pituitary-specific expression of the glycoprotein hormone alpha-subunit gene

Transient transfection studies have been used to determine the DNA sequences of the glycoprotein hormone alpha-subunit gene that are required for tissue-specific expression. In the initial phase of these studies, a variant mouse alpha gene was identified which contains a fully palindromic cAMP response element (CRE). The corresponding region of a previously cloned and sequenced mouse alpha gene contains a single point mutation that disrupts the symmetrical nature of this element. DNase footprint studies demonstrate that the fully palindromic CRE binds the CRE-binding protein with much higher affinity than the imperfect palindrome. Transfection experiments using both mouse alpha gene variants demonstrate differences in basal and cAMP-induced expression. Studies of the cAMP response of the human alpha gene indicated that this gene contains sequences other than the known CRE that are sufficient to permit a transcriptional response to cAMP in both placental and pituitary cells. Expression of human and mouse alpha-subunit genes has been examined in cells of the gonadotrope, thyrotrope, and trophoblast lineages to identify DNA sequences that mediate selective transcription of the alpha gene in these cells. The results demonstrate that sequences between about -500 and -200 are important for expression in the pituitary, but not the placenta. Clustered point mutations were used to further characterize sequences required for expression in the pituitary. Two regions, one at positions -445 to -438 and one at positions -337 to -330, were required for expression in cells of the gonadotrope lineage. One of these regions, at -337 to -330, is also important for expression in thyrotropes. When linked to a minimal promoter, multiple copies of the -344 to -300 region had transcriptional enhancer activity in gonadotropes and thyrotropes, but not in several other cell types. These results are consistent with a model involving different combinations of regulatory elements that determine cell-specific alpha expression in gonadotropes and thyrotropes.

Characterization of a non-tissue-specific, 3',5'-cyclic adenosine monophosphate-responsive element in the proximal region of the rat prolactin gene

The nature of a DNA element located in the -100 to -85 region of the rat PRL gene has been characterized. Previous studies demonstrated that this region may contribute to basal and hormonally regulated expression of the PRL gene. As this region contains a sequence with similarity to a consensus cAMP-responsive element (CRE), a possible role for the cAMP response element binding protein (CREB) has been explored. A point mutation which made the PRL CRE-like sequence less like a consensus CRE had little effect on basal or cAMP-stimulated expression of a PRL-luciferase reporter gene. DNase footprint studies demonstrated that the proximal region of the PRL gene does not contain a high affinity CREB binding site. Mobility shift experiments demonstrated that the major GH3 nuclear protein which interacts with the -100 to -85 region of the PRL gene in vitro is not CREB. Transfection of a dominant inhibitor of CREB action had little or no effect on expression of an indicator gene containing the PRL proximal region. Thus, the PRL proximal region does not contain a high affinity CREB binding site, and it is unlikely that CREB plays a major role in expression of the PRL gene. The functional capabilities of the -100 to -85 region of the PRL gene were then tested in a transfection assay. Synthetic multimers of this region were found to be sufficient to permit a transcriptional response to cAMP or TRH in GH3 cells and cAMP in Rat-1 cells.(ABSTRACT TRUNCATED AT 250 WORDS)

Regulation of the alpha inhibin gene by cyclic adenosine 3',5'-monophosphate after transfection into rat granulosa cells

Inhibin gene expression in the ovary is stimulated by FSH, which uses cAMP as an intracellular second messenger. To examine further the transcriptional regulation of the alpha inhibin gene by FSH and cAMP, we have isolated and characterized a genomic clone that contains the entire rat alpha inhibin gene. Sequence analysis of the alpha inhibin promoter region revealed several potential cAMP response elements (CREs) and transcription factor AP2-binding sites that might mediate cAMP regulation. To determine the functional importance of these sequences, fusion genes including the alpha inhibin 5' flanking region linked to a luciferase reporter gene were transiently transfected into primary granulosa cells isolated from immature rats. These fusion genes were both expressed and regulated by the adenylyl cyclase activator forskolin in transfected granulosa cells. Analysis of a series of 5' deletion mutants indicated that a construct containing as little as 170 basepairs up-stream of the alpha inhibin start site, which includes a single imperfect CRE and no AP2 sites, was regulated by forskolin. DNAse footprinting was used to demonstrate that bacterially expressed CRE-binding protein (CREB) binds to this CRE located 122 basepairs up-stream of the alpha inhibin gene transcriptional start site. To investigate further the role of this CRE in alpha inhibin gene expression, site-specific mutagenesis of the CRE was performed. The alpha inhibin promoter containing a mutated CRE was not regulated by forskolin in granulosa cells and did not bind the CREB protein. Interestingly, mutation of the CRE also substantially reduced basal expression of the alpha inhibin promoter. Lastly, a gel mobility shift assay was used to examine CRE-binding proteins from granulosa cell extracts. Granulosa cells contain a protein that specifically interacts with CRE-containing oligonucleotides or with the alpha inhibin promoter and that is recognized by antibodies against the CREB protein. Our results suggest that CREB or related transcription factors play an important role in both basal and cAMP-regulated expression of the alpha inhibin gene in ovarian granulosa cells.

Increased bronchovascular permeability after allergen exposure in sensitive asthmatics

To investigate the possibility that an increase in bronchovascular permeability is associated with allergen exposure in sensitive asthmatics we evaluated the amounts of serum proteins in bronchoalveolar lavage (BAL) effluents before and after local challenge with allergen. After exposure of sensitive asthmatic airways (n = 15) to allergen significant increases in total protein compared with controls were observed: 0.08 +/- 0.01 mg/ml in control airways and 0.13 +/- 0.02 mg/ml in challenged airways; P less than 0.05. The greatest changes induced by allergen exposure involved small-molecular-weight proteins (less than 345,000) and an inverse correlation was observed between log molecular weight and percent increase in the concentrations of the specific proteins; r = -0.61. BAL-serum distribution coefficients of serum proteins in airway fluids reflected a greater diffusability of low-molecular-weight proteins immediately after allergen exposure. We also evaluated the movement of serum proteins into lung after local allergen exposure using intravenously administered 99mTc-albumin (n = 10) and found an immediate 3.8-fold increase in amounts of radioactive albumin in allergen exposed airways compared with airways exposed to diluent. Most of the radioactivity was recovered in the first 5 ml of aliquot withdrawn, suggesting a marked increase in the permeability of the bronchial (large airway) vascular-epithelial membrane. An increase in serum proteins was also observed in BAL fluid of asthmatics 2-4 h after aerosol challenge (n = 4), including all proteins in the molecular weight range 45,000-900,000. These studies suggest that allergen exposure in sensitive asthmatics causes an acute increase in bronchovascular permeability to serum proteins.

A novel bioassay for lactogenic activity: demonstration that prolactin cells differ from one another in bio- and immuno-potencies of secreted hormone

A reverse hemolytic plaque assay for detecting casein release from individual mammary cells in culture was developed as a bioassay for PRL. Treatment with rat PRL caused dose-related increases in the percentage of mammary cells that released casein and the average size of casein plaques that formed. The assay exhibited exquisite sensitivity (156 fg rat PRL per assay slide) and could be used to evaluate the biopotency of PRL released from individual cells. By combining this "plaque bioassay" with a standard plaque assay for measuring the secretion of immunoreactive PRL, it was possible to compare the bio- and immuno-potencies of hormone released from the same pituitary cells. The results of three separate studies revealed major differences among PRL secretors in these potency estimates. Given the existence of PRL variants with different biological and immunological efficacies, these findings suggest that PRL cells differ from one another in the molecular form(s) of hormone released.

Complement activation in cystic fibrosis respiratory fluids: in vivo and in vitro generation of C5a and chemotactic activity

Experiments performed in vitro have demonstrated that leukocyte neutral proteases produce an important mediator of inflammation, C5a, by proteolysis of the C5 component of the complement system. Cystic fibrosis (CF) lung fluids were characterized by high levels of neutrophils (39% of total cells versus 2% in normals) and contained significantly elevated amounts of elastolytic activity (mean 17.7 ng/micrograms total protein) compared to the lung fluids obtained from normal volunteers (0.2 ng elastolytic activity/micrograms protein, p = 0.001). The objective of these studies was to determine if complement activation and complement-derived chemotactic activity are present in CF lung fluids. C3c peptide representing activation of C3 could not be identified in the bronchial-alveolar lung lavage fluids of normal subjects but was readily identified by means of crossed immunoelectrophoresis in CF lung fluids (n = 9, mean 49% of C3); the mean level of C3 was decreased in CF lung specimens. Chemotactic activity was significantly elevated in lung fluids of the CF patients when compared to normal lung fluids. Using gel-filtration chromatography and a sensitive radioimmunoassay the chemotaxin present in CF specimens was identified as the anaphylatoxin C5a. C5a levels in the bronchial-alveolar lavage fluids of CF patients was inversely related to volume in liters expired in 1 s of a forced expiratory maneuver expressed as a percent of vital capacity determined from a forced expiratory maneuver (r = -0.72). Because there was a direct relationship between the total elastolytic activity present in CF airways and the concentration of C5a (r = 0.97, p = 0.03), it was postulated that airway proteases with elastolytic activity also cleave C5, nonimmunologically producing C5a. Detailed inhibition assays revealed that much of the total elastolytic activity had the inhibition profile of a serine proteinase. The levels of the serine proteinases were closely correlated with the numbers of neutrophilic leukocytes present per ml of lavage fluid (r = 0.7, p = 0.05). However, inhibitors of leukocyte serine proteases did not prevent the generation of additional chemotactic activity and the proteolysis of radiolabeled C5 substrate was not prevented by inhibitors of neutrophil elastase. Although the purified metalloelastase of Pseudomonas aeruginosa was active on cell-bound and free C5 yielding C5a, inhibition of this bacterial protease in CF lung fluids only partially blocked cleavage of the alpha- and beta-chains of C5.(ABSTRACT TRUNCATED AT 400 WORDS)