The transcription factor Elf5 plays an important role in mammary gland development. However, because of the embryonic lethality of Elf5 straight knockout mice, prior studies have been limited to experiments with Elf5 haploinsufficient animals, overexpression systems or transplants. Here, we have utilized K14-Cre to generate mammary-gland specific Elf5 conditional knockout mice. During pregnancy, Elf5-null mammary epithelium completely failed to initiate alveologenesis, and a characteristic of virgin ductal epithelial cells persisted postpartum. We demonstrate that the loss of Elf5 leads to the absence of alveolar secretory markers confirming previous published data. Interestingly, the developmental block due to a lack of Elf5 could not be restored by multiple gestations. Elf5-null mammary epithelial cells also display disorganized cell structures as evident by altered cell polarities, which might be the cause for collapsed lumina. We observe reduced levels of Stat5 and attenuated Stat5 activity as measured by p-Stat5 levels both in Elf5-null mammary glands as well as cultured mammary epithelial cells. This data suggests that the failure of alveolar and lactogenic differentiation due to the loss of Elf5 is mediated in part due to impaired Stat5 activity. In support of this hypothesis, we show by ChIP experiments that Stat5a promoter contains a conserved Elf5-binding site that is occupied by Elf5 in mammary glands. Mammary epithelia lacking Elf5 exhibited downregulation of several other critical genes involved in alveologenesis, suggesting Elf5 as a master regulator in alveolar development. We propose a model for Elf5-mediated alveolar development, in which Elf5 regulates the expression of key mediators of the PrlR/Jak2/Stat5 signaling pathway.

Activation of the protein kinase A (PKA) signaling system is necessary for FSH-induced granulosa cell differentiation, but it is not known whether activation of PKA is sufficient to account for the complex pattern of gene expression that occurs during this process. We addressed this question by infecting granulosa cells with a lentiviral vector that directs the expression of a constitutively active mutant of PKA (PKA-CQR) and compared the cellular responses to PKA-CQR with cells stimulated by FSH. Expression of PKA-CQR in undifferentiated granulosa cells resulted in the induction of both estrogen and progesterone production in the absence of cAMP. The stimulatory effects of both PKA-CQR and FSH on estrogen and progesterone production were suppressed by the PKA inhibitor H-89 and were mimicked by PKA-selective cAMP agonists. mRNA levels for P450scc and 3beta-HSD were induced to a similar extent by FSH and PKA-CQR, whereas mRNA levels for P450arom and the LHr were induced to a greater extent by FSH. Microarray analysis of gene expression profiles revealed that the majority of genes appeared to be comparably regulated by FSH and PKA-CQR but that some genes appear to be induced to a greater extent by FSH than by PKA-CQR. These results indicate that the PKA signaling pathway is sufficient to account for the induction of most genes (as identified by microarray analysis), including those of the progesterone biosynthetic pathway during granulosa cell differentiation. However, optimal induction of aromatase, the LHr, and other genes by FSH appears to require activation of additional signaling pathways.

Granulosa cells express the closely related orphan nuclear receptors steroidogenic factor-1 (SF-1) and liver receptor homolog-1 (LRH-1). To determine whether SF-1 and LRH-1 have differential effects on steroid production, we compared the effects of overexpressing LRH-1 and SF-1 on estrogen and progesterone production by undifferentiated rat granulosa cells. Adenovirus mediated overexpression of LRH-1 or SF-1 had qualitatively similar effects. Neither LRH-1 nor SF-1 alone stimulated estrogen or progesterone production, but when combined with FSH and testosterone, each significantly augmented progesterone production and mRNAs for cholesterol side-chain cleavage enzyme and 3beta-hydroxysteroid dehydrogenase above that observed with FSH alone, with SF-1 being more effective than LRH-1. LRH-1 did not augment FSH-stimulated estrogen production, whereas SF-1 produced only a slight ( approximately 30%) augmentation of FSH-stimulated estrogen production. The stimulatory actions of both were reduced by overexpression of dosage-sensitive sex reversal, adrenal hypoplasia congenita, critical region on the X chromosome, gene 1. Expression of either LRH-1 or SF-1 together with constitutively active protein kinase B in the absence of FSH stimulated progesterone production and mRNAs for 3beta-hydroxysteroid dehydrogenase and cholesterol side-chain cleavage enzyme but did not stimulate estrogen production or mRNA for aromatase. These findings demonstrate that LRH-1 and SF-1 have qualitatively similar actions on FSH-stimulated estrogen and progesterone production, which would suggest that these factors may have overlapping actions in the regulation of steroidogenesis that accompanies granulosa cell differentiation.

Utilization of heme as an iron source by Bradyrhizobium japonicum involves induction of the outer membrane heme receptor gene hmuR and other genes within the heme utilization locus. Here, we discovered the hmuP gene located upstream of hmuR and transcribed divergently from it along with hmuTUV. hmuP encodes a small protein that accumulated under iron limitation and is transcriptionally controlled by the global iron-responsive regulator Irr, as were all genes within the heme utilization locus. Cross-linking and immunoprecipitation experiments showed that Irr occupies the hmuR-hmuP promoter in vivo. An hmuP mutant did not grow on heme as an iron source, but retained the ability to use ferric chloride. Correspondingly, induction of hmuR mRNA under iron limitation was severely diminished in an hmuP strain, but other genes within the Irr regulon were unaffected. HmuP occupied the hmuR-hmuP promoter, and thus it plays a direct regulatory role in gene expression. HmuP was not required for Irr occupancy, nor was ectopic expression of hmuP from an Irr-independent promoter sufficient to induce the hmuR gene. Thus, both HmuP and Irr occupancy are necessary for hmuR induction. We suggest that HmuP is a coactivator of Irr-dependent expression of hmuR.

BackgroundThe ETS transcription factor Elf5 (also known as ESE-2) is highly expressed in the mammary gland and plays an important role in its development and differentiation. Indeed studies in mice have illustrated an essential role for Elf5 in directing alveologenesis during pregnancy. Although the molecular mechanisms that underlie the developmental block in Elf5 null mammary glands are beginning to be unraveled, this investigation has been hampered by limited information about the identity of Elf5-target genes. To address this shortcoming, in this study we have performed ChIP-cloning experiments to identify the specific genomic segments that are occupied by Elf5 in pregnant mouse mammary glands.ResultsSequencing and genomic localization of cis-regulatory regions bound by Elf5 in vivo has identified several potential target genes covering broad functional categories. A subset of these target genes demonstrates higher expression levels in Elf5-null mammary glands suggesting a repressive functional role for this transcription factor. Here we focus on one putative target of Elf5, the Ccnd2 gene that appeared in our screen. We identify a novel Elf5-binding segment upstream of the Ccnd2 gene and demonstrate that Elf5 can transcriptionally repress Ccnd2 by directly binding to the proximal promoter region. Finally, using Elf5-null mammary epithelial cells and mammary glands, we show that loss of Elf5 in vivo leads to up regulation of Ccnd2 and an altered expression pattern in luminal cells.ConclusionsIdentification of Elf5-targets is an essential first step in elucidating the transcriptional landscape that is shaped by this important regulator. Our studies offer new toolbox in examining the biological role of Elf5 in mammary gland development and differentiation.