NF-kappaB is involved in the induction of the rat hepatic alpha1-acid glycoprotein gene by phenobarbital

Phenobarbital, a classical inducer of the drug-metabolizing cytochrome P450 genes, induces alpha1-acid glycoprotein gene expression through a PB-responsive element (PBRE) located at position -142 to -126 from the transcriptional start site. The aim of this study was to investigate nuclear protein binding to the PBRE sequence after PB treatment. Cycloheximide treatment showed that de novo protein synthesis was not required for PB to induce AGP gene expression, pointing to post-translational modifications. Studies of the DNA-protein complex with the PBRE showed that phosphorylation status is a key regulator of the binding capacity of transactivating proteins involved in PB transcriptional activation. This DNA-protein complex, analyzed by southwestern blotting and UV cross-linking, involves three nuclear factors with molecular weights of 43, 52, and 65 kDa. Supershift and competition experiments showed that the 43-kDa factor can be related to C/EBPalpha and the 52- and 65-kDa factors to the two subunits of NF-kappaB.

Growth hormone inhibits rat liver alpha-1-acid glycoprotein gene expression in vivo and in vitro

The gene encoding alpha-1-acid glycoprotein (AGP), one of the major acute-phase proteins, is positively controlled at the transcriptional level by cytokines (interleukin-1 [IL-1], IL-6, and tumor necrosis factor alpha) and glucocorticoids. Here, we show that growth hormone (GH) treatment of isolated rat hepatocytes in vitro reduces AGP messenger RNA (mRNA) expression. AGP gene expression remained inducible by IL-1, IL-6, and phenobarbital (PB) in GH-treated hepatocytes. Interestingly, the repressive effect of GH on AGP gene expression was also observed in vivo: liver AGP mRNA content was strongly increased in hypophysectomized rats, and GH treatment of these animals led to a decrease in mRNA to levels lower than those in untreated control animals. Moreover, the inhibitory effect of GH mainly occurs at the transcriptional level and can be observed as little as 0.5 hours after GH adding in vitro to isolated hepatocytes. These results show negative regulation of AGP gene expression and strongly suggest that GH is a major endogenous regulator of constitutive AGP gene expression. Moreover, transfection assays showed that the region of the AGP promoter located at position -147 to -123 is involved in AGP gene regulation by GH. Furthermore, GH deeply modifies the pattern of nuclear protein binding to this region. GH treatment of hypophysectomized rats led to the release of proteins of 42 to 45 and 80 kd and to the binding of proteins of 48 to 50 and 90 kd.

Phenobarbital induction of alpha 1-acid glycoprotein in primary rat hepatocyte cultures

The serum level of rat alpha 1-acid glycoprotein is significantly increased by treatment with phenobarbital, and in vivo studies have shown that phenobarbital seems to act mainly at the transcriptional level. To show the direct mediating effect of phenobarbital on alpha 1-acid glycoprotein gene expression, we investigated the ability of primary cultured rat hepatocytes to respond to in vitro phenobarbital administration. Phenobarbital increased both alpha 1 acid glycoprotein secretion and corresponding mRNA levels in primary rat hepatocytes cultured on matrigel. Used in combination with interleukin-1, interleukin-6 and dexamethasone, phenobarbital had an additive or synergistic effect on alpha 1-acid glycoprotein synthesis. These results show that (a) phenobarbital acts directly on hepatocytes by increasing alpha 1-acid glycoprotein gene expression and (b) this effect is mediated by a specific mechanism independent of pathways involved in alpha 1-acid glycoprotein induction by interleukin-1, interleukin-6 and glucocorticoids.

Transcriptional regulation of rat alpha 1-acid glycoprotein gene by phenobarbital

Phenobarbital induces gene transcription of both cytochrome P450IIB (the barbiturate-inducible cytochrome P450 in mammals) and alpha 1-acid glycoprotein, one of the major acute-phase proteins in rats and humans. Analysis of the 5'-regulatory sequences of cytochrome P450IIB and alpha 1-acid glycoprotein genes in rats revealed the presence of a consensus sequence of 10 base pairs, termed the phenobarbital-responsive element or Barbie box, located in a region extending from positions -136 to -127 from the transcription start site of the alpha 1-acid glycoprotein gene. A 17-base pair oligonucleotide probe specific for alpha 1-acid glycoprotein and including the consensus sequence showed, in mobility shift assays, slight binding to liver nuclear protein from untreated animals. This binding was strongly and specifically increased with protein extracts from phenobarbital-treated rats. Transfection of rat primary hepatocytes with the pAGPcat construct induced basal expression of chloramphenicol acetyltransferase activity, which was increased by phenobarbital and dexamethasone treatment of cells. Induction of chloramphenicol acetyltransferase activity by phenobarbital was abolished when hepatocytes were transfected by constructs with a mutation or deletion of the Barbie box sequence. These results strongly suggest that the Barbie box sequence is involved in alpha 1-acid glycoprotein gene regulation by phenobarbital.

Induction of rat alpha-1-acid glycoprotein by phenobarbital is independent of a general acute-phase response

Phenobarbital (PB) induces transcription of the alpha 1-acid glycoprotein (AGP) gene, one of the major positive acute-phase proteins, the expression of which is controlled by a specific combination of glucocorticoids and cytokines. This raises questions as to the involvement of glucocorticoids and cytokine pathways in the PB-mediated effect on AGP gene expression. We found that the pattern of whole-serum proteins in PB-treated rats differed markedly from that observed during a typical acute inflammatory response (in turpentine-treated rats): levels of some positive acute-phase proteins (APP) increased slightly (alpha 1-acid glycoprotein, haptoglobin, hemopexin and T-kininogen), while levels of alpha 2 macroglobulin, the most sensitive marker of the acute-phase reaction, decreased. Among the negative APP, neither albumin nor prealbumin decreased while CBG increased. The cytokines involved in AGP gene regulation (mainly IL1, IL6 and TNF alpha) do not therefore seem to mediate the effect of PB on acute-phase protein expression. Glucocorticoid involvement is also ruled out by the observed enhancement of the effect of PB on AGP expression in adrenalectomized animals. Our results suggest that phenobarbital acts on AGP expression by a mechanism independent of the inflammatory pathway.