DNA sequences required for expression of the LH beta promoter in primary cultures of rat pituitary cells

A novel activation function for NAB proteins in EGR-dependent transcription of the luteinizing hormone beta gene

The EGR1/NGFI-A transcription factor directly activates the luteinizing hormone beta (LHbeta) subunit promoter, and female mice lacking EGR1 are infertile due to LHbeta deficiency. The NGFI-A-binding proteins NAB1 and NAB2 are corepressors of EGR1/NGFI-A and of the related proteins EGR2/Krox20 and EGR3. Here we report that at certain promoters, including LHbeta, NAB proteins display a novel ability to stimulate EGR-directed transcription. NAB coactivation requires the conserved NCD2 protein domain, previously implicated in NAB corepression, is strictly dependent upon EGR binding to the LHbeta proximal promoter and is independent of EGR activation domains. Furthermore, we report that NAB-activated promoters such as LHbeta contain EGR consensus sites that are fewer in number and lower in binding affinity than those found at NAB-repressed promoters such as basic fibroblast growth factor. Analysis of mutant and synthetic promoters confirms that both the strength and multiplicity of EGR-binding sites influence the transcriptional outcome of NAB recruitment. These results suggest a novel means by which EGR target genes could be differentially regulated in cells where EGR and NAB proteins are coexpressed.

The equine luteinizing hormone beta-subunit promoter contains two functional steroidogenic factor-1 response elements

The requirements for basal expression of the LH beta-subunit promoter in pituitary gonadotropes are largely unknown. We have used the equine (e) LHbeta subunit promoter as a model to unravel the combinatorial code required for gonadotrope expression. Through the use of 5'-deletion mutagenesis, a region between -185 and -100 of the eLHbeta promoter was shown to play a critical role in maintaining basal promoter activity in alphaT3-1 and LbetaT2 cells. This region encompasses the steroidogenic factor-1 (SF-1) binding site that has been reported to have a functional role in expression of the LHbeta promoter in other species. We have also identified an additional SF-1 site at -55 to -48. Binding of SF-1 to both sites was confirmed by electrophoretic mobility shift assays. Mutations within these sites, either individually or in combination, did not attenuate basal activity of the eLHbeta promoter in alphaT3-1 cells, but did diminish promoter activity in LbetaT2 cells. Interestingly, cotransfection with an expression vector encoding SF-1 induced eLHbeta promoter activity, and this induction was abrogated by mutations within the SF-1 sites in alphaT3-1 cells. Block replacement mutagenesis was performed on the -185/-100 region of the eLHbeta promoter to identify DNA response elements responsible for maintaining basal promoter activity. From this analysis, two regions emerged as being important: a distal 31-bp segment (-181 to -150) and an element located immediately 3' to the distal SF-1 site (-119 to -106). It is hypothesized that these two regions as well as the SF-1 sites represent regulatory elements that contribute to a combinatorial code involved in targeting expression of the eLHbeta promoter to gonadotropes.

Identification of cis-acting deoxyribonucleic acid elements that mediate gonadotropin-releasing hormone stimulation of the rat luteinizing hormone beta-subunit gene

GnRH plays a critical role in reproductive development and function by regulating the biosynthesis and secretion of the pituitary gonadotropins, LH and FSH. Although it is known that GnRH induces gonadotropin subunit gene transcription, the mechanism by which this occurs has not been elucidated. Studies have been hindered by the lack of available cell lines that express the LH and FSH subunit genes and respond to GnRH. We have transfected the rat pituitary GH3 cell line with the rat GnRH receptor complementary DNA. These cells, when cotransfected with regulatory regions of the LH or FSH subunit genes fused to a luciferase reporter gene, respond to GnRH with an increase in promoter activity comparable to that seen in primary rat pituitary cells. In this study, we have used this cell model to identify cis-acting elements of the LHbeta gene that mediate stimulation by GnRH. Analysis of a series of 5'-deletion and internal deletion constructs has revealed two regions of the rat LHbeta gene promoter involved in mediating the response to GnRH, region A (-490/-352) and region B (-207/-82). Fusion of region A upstream of a heterologous minimal promoter linked to the luciferase gene conferred GnRH responsiveness to the promoter, whereas region B did not. However, the presence of both regions A and B conferred a greater GnRH response than region A alone. Electrophoretic mobility shift assay revealed the presence of a protein(s) binding to region A using GH3 as well as alphaT3-1 nuclear extracts. Thus, region A (-490/-352) confers GnRH responsiveness to the LHbeta subunit gene and binds to a protein(s) present in pituitary cell lines. DNA sequences in region B (-207/-82) also contribute to GnRH responsiveness. The identification of putative GnRH response elements in the rat LHbeta gene promoter will aid in elucidation of the mechanisms of regulation of gene expression by GnRH.

Luteinizing hormone deficiency and female infertility in mice lacking the transcription factor NGFI-A (Egr-1)

The immediate-early transcription factor NGFI-A (also called Egr-1, zif/268, or Krox-24) is thought to couple extracellular signals to changes in gene expression. Although activins and inhibins regulate follicle-stimulating hormone (FSH) synthesis, no factor has been identified that exclusively regulates luteinizing hormone (LH) synthesis. An analysis of NGFI-A-deficient mice derived from embryonic stem cells demonstrated female infertility that was secondary to LH-beta deficiency. Ovariectomy led to increased amounts of FSH-beta but not LH-beta messenger RNA, which suggested a pituitary defect. A conserved, canonical NGFI-A site in the LH-beta promoter was required for synergistic activation by NGFI-A and steroidogenic factor-1 (SF-1). NGFI-A apparently influences female reproductive capacity through its regulation of LH-beta transcription.

The 1994 Stevenson Award Lecture. Follicle-stimulating hormone and luteinizing hormone: a tale of two gonadotropins

Although most gonadotropes synthesize both luteinizing hormone and follicle-stimulating hormone, the transcription, content, and secretion rates of the two gonadotropins can be separated. The signals external to the gonadotropic cells that appear to be important in the differential regulation are gonadotropin-releasing hormone pulse frequency (high pulse frequency favors luteinizing hormone), steroid feedback (works on both but induces a more powerful negative feedback on luteinizing hormone), and gonadal peptide feedback (activin increases follicle-stimulating hormone; inhibin and follistatin decrease it). We know very little about the pathways within the gonadotropes that favor one gonadotropin rather than another. It is expected that the cloning of the genes for both gonadotropins and the use of specific cell lines and transfections will lead to elucidation of these pathways.

Characterization of the ovine LH beta-subunit gene: the promoter directs gonadotrope-specific expression in transgenic mice

The alpha- and beta-subunits of the gonadotropin hormones are expressed in the gonadotrope cells of the anterior pituitary. There are no adequate in vitro systems for the analysis of beta-subunit gene expression. In this study, therefore, transgenic mice have been used to investigate the regulation of expression of the ovine luteinizing hormone beta-gene (oLH beta) in vivo. oLH beta was isolated, characterized, and 1.9 kb of the promoter fused to the bacterial reporter chloramphenicol acetyl-transferase (CAT). Three lines of transgenic mice were generated. CAT enzyme was detected in the pituitary of two lines, whereas the third line did not express. Measurement of endogenous luteinizing hormone and follicle stimulating hormone levels in both expressing lines revealed small differences when compared to controls, but these did not affect the fertility of the animals. Immunostaining of the anterior pituitary revealed that the oLH beta CAT transgene was expressed specifically in gonadotrope cells.

Expression of luteinising hormone-beta subunit chloramphenicol acetyltransferase (LH-beta-CAT) fusion gene in rat pituitary cells: induction by cyclic 3'-adenosine monophosphate (cAMP)

In this study we determined the activity of the rat luteinising hormone-beta gene promoter in a heterologous rat pituitary cell line (GH3 cells). 1.7 kb of LH-beta 5' flanking sequence and the first 5 bp of the 5' untranslated region were ligated to the chloramphenicol acetyltransferase (CAT) receptor gene (LH-beta-CAT) and transiently transfected by calcium phosphate precipitation into subconfluent cultures of GH3 cells. Basal low-level CAT activity was only detected in GH3 cells, being absent in two non-pituitary cell lines (BeWo and HeLa) RNase analysis revealed that mRNA from transfected GH3 cells protected a fragment of labelled antisense probe of correct size for transcription initiation from the LH-beta CAP site, confirming that promoter activity reflected correctly initiated LH-beta-CAT fusion gene transcripts. CAT activity was consistently induced by an average of 3-5-fold from the full-length 1.7 kb promoter, in a dose- and time-dependent manner, by forskolin, dibutyryl cAMP, and 8-bromo cAMP implying presence of a cAMP-responsive cis-acting domain in the LH-beta promoter region. Transfection of deletion mutants delta-615-CAT, delta-385-CAT and delta-250-CAT each reduced forskolin inducibility to 1.7-fold but did not abolish induction completely suggesting a domain between -1.7 and -0.6 kb contained a cAMP-responsive element(s) (CRE). Further deletion of LH-beta 5' flanking sequences to delta-85-CAT restored forskolin induction to wild-type levels (3-5-fold), suggesting the presence of a weak inhibitory element between -600 and -85 kb, and a cAMP-responsive domain in the proximal promoter region. The LH-beta promoter does not contain perfect tandem repeat palindromic CRE DNA sequences, though there are several octanucleotide sequences differing by only 1 bp from AP-2 binding sites, the consensus CRE, and the vasointestinal peptide gene CRE. Although these data suggest that the LH-beta gene is cAMP responsive this is likely mediated by several and complex protein interactions with multiple DNA sequences in the proximal and distal LH-beta promoter enhancer.