Search

In the present study, we analyzed the influence of retinoic acids on the expression of alpha-1 acid glycoprotein (AGP). We show that in rat primary hepatocytes, 9-cis retinoic acid and all-trans retinoic acid increase AGP gene expression at the transcriptional level. Transient transfections of rat primary hepatocytes with a reporter construct driven by the rat AGP gene promoter indicated that retinoids regulate AGP gene expression via the -763/-138 region of the AGP promoter. Furthermore, cotransfection experiments with retinoic acid receptor alpha (RARalpha) and retinoid X receptor alpha (RXRalpha) expression vectors in NIH3T3 cells demonstrated that both RXRalpha/RXRalpha homodimer and RXRalpha/RARalpha heterodimer are competent for ligand-induced transactivation of the AGP promoter. Unilateral deletion and site-directed mutagenesis identified two retinoic-acid responsive elements (RARE), RARE-I and RARE-II, which interestingly correspond to a direct repeat of two TGACCT-related hexanucleotides separated by a single bp only (DR1-type response element). Cotransfection assays showed that RXRalpha and RARalpha activate AGP gene transcription through these two elements either as a homodimer (RXRalpha/RXRalpha) or as a heterodimer (RXRalpha/RARalpha). The RXRalpha/RXRalpha homodimer acts most efficiently through the RARE-I response element to promote AGP transactivation, whereas the RXRalpha/RARalpha heterodimer mediates transactivation better via the RARE-II responsive element.

Responsiveness to cytokine-mediated acute inflammatory stimuli of the highly differentiated and polarized WIF-B hybrid cell line was studied by measuring the induction of alpha 1-acid glycoprotein and alpha 2-macroglobulin mRNAs after interleukin-1, interleukin-6 and tumor necrosis factor-alpha treatments in the presence of dexamethasone. Compared with their Fao parent, WIF-B cells were 10 times more responsive to 24-h interleukin-6 induction regarding alpha 2-macroglobulin induction. At variance from the response measured in Fao cells, the late effects of interleukin-6 treatment confirmed the higher sensitivity of WIF-B cells to this cytokine as a 72-h treatment as 10 times more effective than a 24-h treatment at inducting alpha 1-acid glycoprotein mRNA. These findings highlight the hepatocyte differentiation of WIF-B cells compared with other hepatoma cell lines, with respect to the regulation of acute-phase protein gene expression. They also make WIF-B cells a convenient model to study the molecular effects of interleukin-6 in terms of transduction and/or transcription, and the many cross-talks that occur during the regulation of acute-phase protein gene expression.

The effect of D-galactosamine on protein N-glycosylation was studied in rat hepatocyte primary cultures for alpha 1-antitrypsin (three complex type oligosaccharide chains) and alpha 1-acid glycoprotein (six complex type oligosaccharide chains). D-Galactosamine at a concentration of 4 mM inhibited partially de novo N-glycosylation leading to the formation of alpha 1-antitrypsin lacking one to two and of alpha 1-acid glycoprotein lacking one to five of its carbohydrate side chains. In addition D-galactosamine interfered with oligosaccharide processing, leading to the formation of some carbohydrate side chains remaining in an endoglucosaminidase H sensitive, i.e., not completely processed, form. D-Galactosamine impaired the secretion of alpha 1-antitrypsin and of alpha 1-acid glycoprotein but did not inhibit the secretion of the unglycosylated albumin. The inhibitory effect of D-galactosamine on de novo glycosylation as well as on oligosaccharide processing lasted for at least 24 h after it had been removed from the cells. D-Galactosamine impaired the glycosylation of alpha 1-antitrypsin only in hepatocytes, but not in human monocytes. Furthermore, D-galactosamine did not impair the N- and O-glycosylation of interleukin-6 in human monocytes and in MRC 5 fibroblasts. The results indicate that the effect of D-galactosamine on protein glycosylation is restricted to D-galactosamine metabolizing hepatocytes and is not exerted by the drug itself but by its metabolites.

The relationship of protein glycosylation to the externalization of glucocorticoid inducible alpha1-acid glycoprotein and mouse mammary tumor virus glycoproteins was examined in M1.54, a clonal population of mouse mammary tumor virus-infected rat hepatoma cells. Multiple freeze-thaw of isolated microsomes revealed that while alpha 1-acid glycoprotein is carried through the cell as a soluble component of vesicles, extracellular viral glycoproteins are initially membrane-associated. At concentrations of tunicamycin that specifically inhibited N-linked protein glycosylation, alpha 1-acid glycoprotein fractionated between the cellular and extracellular compartments. Thus, approximately one half of the newly synthesized, nonglycosylated (22,000 Mr) alpha 1-acid glycoprotein was rapidly secreted with kinetics similar to its glycosylated counterpart (release half-time of 60 min), while the remaining species first localized in an undefined intracellular compartment prior to its slow secretion (release half-time of 24 h). The same distribution of nonglycosylated alpha 1-acid glycoprotein was observed at various absolute levels of polypeptide, suggesting that this was not due simply to the saturation of an efficient secretory pathway at high polypeptide levels. In contrast to alpha 1-acid glycoprotein, no labeled viral antigens were released by tunicamycin-treated M1.54, while a nonglycosylated viral precursor glycopolyprotein was expressed intracellularly. Taken together, these results suggest that carbohydrate attachment strongly regulates the externalization of both alpha 1-acid glycoprotein and mouse mammary tumor virus species, which represent two distinct classes of extracellular glycoproteins.