Positive and negative regulation of a transfected chimeric tyrosine aminotransferase gene: effect of copy number

To define a selective system for the study of rat tyrosine aminotransferase (TAT; EC 2.6.1.5) gene expression, we have introduced into cultured cells the selectable bacterial gene gpt linked to TAT gene flanking sequences. After integration in host cell DNA, the chimeric gene exhibits the same pattern of regulation as the TAT gene. In hepatoma cells, its expression is induced after glucocorticoid hormone treatment and repressed after fusion with fibroblasts. In fibroblasts, the chimeric gene is not expressed. The correct pattern of regulation is lost when the number of integrated copies is high. At copy number above 10, the transfected gene responds poorly to glucocorticoids in hepatoma cells. At copy number above 50, the gene is expressed in fibroblasts. Another gene present in the same construction and controlled by the SV40 early promoter and enhancer is positively regulated by glucocorticoids in hepatoma cells but not after fusion with fibroblasts. These data indicate that in hybrid cells, both TAT promoter and glucocorticoid-responsive elements are negatively regulated.     

Activity of the rat liver-specific aldolase B promoter is restrained by HNF3.

Although it contains binding sites for HNF1, NFY and C/EBP/DBP, the proximal promoter of the aldolase B gene is surprisingly weak when tested by transient transfection in differentiated hepatoma cells. This low activity could be due to overlapping between HNF1 and HNF3 binding sites in element PAB, from -127 to -103 bp with respect to the cap site. Replacement of the PAB region by a consensus HNF1 binding site unable to bind HNF3, results in a 30 fold activation of the promoter, in accordance with the hypothesis that activity of the wild-type promoter is normally restrained by HNF3 binding to PAB competitively with HNF1. Consistently, transactivation of the wild-type promoter by excess HNF1 is very high, most likely due to the displacement of HNF3, while the construct with the exclusive HNF1 binding site is weakly transactivated by HNF1. The inhibitory effect of HNF3 on HNF1-dependent transactivation is clearly due to competition between these two factors for binding to mutually exclusive, overlapping sites; indeed, when HNF1 and HNF3 sites are contiguous and not overlapping, the resulting promoter is as active as the one containing an exclusive HNF1 binding site. A construct in which PAB has been replaced by an exclusive HNF3 binding site is weakly expressed and is insensitive to HNF3 hyperexpression. DBP-dependent transactivation, finally, is independent of the nature of the element present in the PAB region.     

Hierarchy and positive/negative interplays of the hepatocyte nuclear factors HNF-1, -3 and -4 in the liver-specific enhancer for the human alpha-1-microglobulin/bikunin precursor.

Alpha-1-microglobulin and bikunin are two plasma glycoproteins encoded by an alpha-1-microglobulin/bikunin precursor (AMBP) gene. The strict liver-specific expression of the AMBP gene is controlled by a potent enhancer made of six clustered boxes numbered 1-6 that have been reported to be proven or potential binding sites for the hepatocyte-enriched nuclear factors HNF-1, -4, -3, -1, -3, -4, respectively. In the present study, electromobility shift assays of wild-type or mutated probes demonstrated that the boxes 1-5 have a binding capacity for their cognate HNF protein. Box 5 is also a target for another, as yet unidentified, factor. A functional analysis of the wild-type or mutated enhancer, driving its homologous promoter and a reporter CAT gene in the HepG2 hepatoma cell line, demonstrated that all six boxes participate in the enhancer activity, with the primary influence of box 4 (HNF-1) and box 2 (HNF-4). A similar analysis in the HNF-free CHO cell line co-transfected with one or several HNF factors further demonstrated various interplays between boxes: box 3 (HNF-3 alpha and beta) has a negative influence over the major HNF-4 box 2 as well as a positive influence over the major HNF-1 box 4.     

vHNF1 is a homeoprotein that activates transcription and forms heterodimers with HNF1.

vHNF1 and HNF1 are two nuclear proteins that bind to an essential element in the promoter proximal sequences of albumin and of many other liver-specific genes. HNF1 predominates in hepatocytes but is absent in dedifferentiated hepatoma cells. These cells contain vHNF1 but fail to express most of the liver traits. In the present work we have isolated cDNA clones for vHNF1 and found that it is a homeoprotein homologous to HNF1 in regions important for DNA binding. Unexpectedly, vHNF1 transactivated the albumin promoter in transfection experiments. Like the HNF1 mRNA, the vHNF1 message was found in kidney, liver and intestine although in different proportions. The fact that vHNF1 and HNF1 readily form heterodimers in vitro and the biochemical characterization of vHNF1/HNF1 heterodimers in nuclear extracts of kidney, liver and several cell lines, strongly argue that such heterodimers exist in vivo. Our results raise the possibility that heterodimerization between homeoproteins could be a common phenomenon in higher eukaryotes, which may have implications in the regulatory network sustained between these factors.     

Transcriptional control of the human aldehyde dehydrogenase 2 promoter by hepatocyte nuclear factor 4: inhibition by cyclic AMP and COUP transcription factors.

An important regulatory element (designated FP330-3') of the ALDH2 promoter mediates activation by hepatocyte nuclear factor 4 (HNF4). This activation of promoter constructs containing this element by HNF4 was reduced by nearly half by 8-Br-cAMP in H4IIEC3 cells, an effect that was blocked by inhibitors of protein kinase A (PKA). Cotransfection assays showed that COUP-TF I, ARP-1, or PPARdelta suppressed the ability of HNF4 to activate the reporter. The repression was potentiated by 8-Br-cAMP. Electrophoretic mobility shift assays revealed that treatment of hepatoma cells or cultured rat hepatocytes with 1 mM 8-Br-cAMP or glucagon reduced binding of FP330-3' by HNF4 by half. In vitro phosphorylation of HNF4 by PKA decreased binding to FP330-3'. Fasting reduced the ALDH2 protein level in liver and kidney, two tissues expressing HNF4, but not heart. These data suggest that ALDH2 expression can be suppressed by cAMP, most likely through phosphorylation of HNF4 by PKA, and this may account for the reduction in enzyme protein during fasting.