Search

Adaptation of liver to nutritional signals is regulated by several transcription factors that are modulated by intracellular metabolites. Here, we demonstrate a transcription factor network under the control of hepatocyte nuclear factor 4alpha (HNF4alpha) that coordinates the reciprocal expression of fatty acid transport and metabolizing enzymes during fasting and feeding conditions. Hes6 is identified as a novel HNF4alpha target, which in normally fed animals, together with HNF4alpha, maintains PPARgamma expression at low levels and represses several PPARalpha-regulated genes. During fasting, Hes6 expression is diminished, and peroxisome proliferator-activated receptor alpha (PPARalpha) replaces the HNF4alpha/Hes6 complex on regulatory regions of target genes to activate transcription. Gene expression and promoter occupancy analyses confirmed that HNF4alpha is a direct activator of the Pparalpha gene in vivo and that its expression is subject to feedback regulation by PPARalpha and Hes6 proteins. These results establish the fundamental role of dynamic regulatory interactions between HNF4alpha, Hes6, PPARalpha, and PPARgamma in the coordinated expression of genes involved in fatty acid transport and metabolism.

The promoter fragment -1550/+129 of the human hepatic triglyceride lipase (HTGL) gene drives the expression of the CAT gene in HepG2 cells, albeit at very low levels. Transient transfections in HepG2 and HeLa cells of 5' deletion constructs indicated that the regulatory elements that control this expression are located in the proximal region of the gene. DNase I footprint analysis with DNA fragments spanning the region -483 to +129 and rat liver nuclear extracts identified eight protected regions, four upstream of the transcription initiation site (A, -28 to -75; B, -96 to -106; C, -118 to -158; D, -185 to -255) and four in the first exon of the gene (E1, -5 to +20; E2, +36 to +55; E3, +58 to +83; E4, +86 to +107). DNA binding and footprinting analysis demonstrated that the region -75 to -43 within footprint A binds to the liver-specific transcription factor HNF1. The region +28 to +129 contains a functional negative regulatory element (NRE) since deletion of this region results in a 17-fold increase in CAT activity. The NRE can act independent of orientation and position and repress transcription driven by heterologous promoters. DNA binding assays using native and fractionated liver nuclear extracts identified two transcription factors that bind to element E2 and also to element E3. A dinucleotide mutation in element E2 which causes derepression of the HTGL gene by 10-fold also abolishes the binding of these two activities. Transfection experiments showed that deletion of the NRE allows expression of reporter constructs in HeLa cells, indicating that the NRE may play a determinant role for the expression of HTGL gene in hepatic cells.

We have shown previously that the hepatic and intestinal transcription of the human apolipoprotein A-II (apoA-II) gene in cell cultures is controlled by a complex set of regulatory elements A-N [Chambaz, et al. (1991) J. Biol. Chem. 266, 11676-11685; Cardot, et al. (1991) J. Biol. Chem. 266, 24460-24470]. In the present communication, we have assessed the functional importance of each of the regulatory elements. In addition, we have used DNA binding and competition assays and protein fractionation to identify the hepatic nuclear activities which are involved in the regulation of the human apoA-II gene. Such activities may be of general importance for the regulation of liver-specific genes. The DNA binding and competition analysis showed that the regulatory elements M, D, and F bind new activities which have not been identified in apolipoprotein or other liver-specific promoters. These activities have been designated AIIM1 and AIIM2 for element M, AIID1 and AIID2 for element D, and AIIF2 for element F. The activity AIIM2 is present in liver, but absent in CaCo-2 cells. A set of regulatory elements binds activities which resemble liver-enriched or ubiquitous factors previously shown to play important roles in the regulation of their target genes. Thus, element I binds to activities related to NF1, and elements L, C, D, G, AB, and F bind to C/EBP alpha as well as other heat-stable activities. The affinity of the bacterially expressed C/EBP alpha for the various apoA-II regulatory regions follows the order: AIIL approximately AIIC > AIID > AIIF > AIIG > AIIAB. Protein fractionation showed that element J binds at least three hepatic nuclear activities and is also recognized by members of the nuclear receptor family, HNF4, EAR2, EAR3, and ARP1. Another liver-enriched factor, HNF1, was shown previously to bind to element H. Despite the importance of HNF1, HNF4, NF1, and C/EBP alpha in the regulation of numerous other target genes, deletion of the HNF1, NF1, and HNF4 and several C/EBP binding sites did not drastically affect the hepatic transcription of the apoA-II gene. Rather, the hepatic and intestinal transcription is affected severely by deletion of elements A, B, K, L, and N. In addition, the intestinal transcription is affected by deletion of elements C, J, and M. The in vivo physiological importance of these elements will require analysis of their function in transgenic animals. This analysis establishes the organization of several nuclear activities on the human apoA-II promoter.(ABSTRACT TRUNCATED AT 400 WORDS)

The plasma levels of apo B and apo E, and the level of hepatic and intestinal mRNA coding for these apolipoproteins were investigated in weanling male rats pair-fed for 6 wk with a control or copper-deficient diet. Plasma cholesterol, triglycerides, and phospholipids were significantly increased, and plasma apo B and apo E levels were also markedly increased in copper-deficient rats as compared to control rats. Copper deficiency significantly increased triglyceride levels and decreased cholesterol levels in the liver. No major differences in the levels of hepatic and intestinal apo B and apo E mRNA occurred between control and copper-deficient rats. These data imply that hypertriglyceridemia dn hypercholesterolemia owing to the copper deficiency are not accompanied by modifications in the gene expression at the mRNA level in the liver and intestine of the apolipoproteins studied.

The effect of nutritional factors on apolipoprotein gene expression by rat liver were studied. Dietary carbohydrates or fatty acids regulate the expression of apo E gene, by altering either gene transcription or mRNA stability. Conversely, apo A1 regulation occurs at a post transcriptional level. In vivo and in vitro experiments gave contradictory results concerning apo B gene expression. The more dramatic changes in plasma lipids and apolipoproteins are obtained under dietary fish oil. Hepatocytes from fish oil-fed rats retain for several days modification in fatty acid metabolism, i.e. a shift in oleic acid channeling towards oxidation at the expense of esterification and a reduced ability to synthesize and secrete triacylglycerol. These modifications are paralleled with a decrease in the synthesis and in the secretion of apo Bs. Hepatocytes from fish oil fed rats secrete degradative forms of apo B which might result from either a sluggish VLDL synthesis and secretion or a more specific effect of n-3 long chain polyunsaturated fatty acid peroxidative products. Hepatocytes from fish oil fed rats exhibit a reduced ability to synthesize cholesterol, associated with a decrease in apo A1 synthesis and secretion without any modification in apo A1 mRNA. In contrast, the hepatocytes exhibit a concomitent decrease in apo E synthesis and secretion and in cellular apo E mRNA levels.

The apolipoprotein A-II (apoA-II) gene regulatory region -911 to +29 strongly promotes the transcription of the promotorless chloramphenicol acetyltransferase (CAT) gene in cells of hepatic (HepG2) and intestinal (CaCo2) origin but not in HeLa cells. Deletion of nucleotides -911 to -860 decreased the hepatic and intestinal transcription to 7% and 18% of control, respectively. Further progressive deletions extending to nucleotides -614, -440, -230, and -80 abolished both hepatic and intestinal transcription, indicating that the distal promoter region -911 to -614 contains regulatory elements that are essential for intestinal and hepatic transcription. An internal deletion of the -614 to -230 region decreased hepatic transcription 60% while it increased intestinal transcription 140% of control indicating that the elements which control hepatic and intestinal transcription of the apoA-II gene may be different within this regulatory region. DNase I footprinting analysis with rat liver nuclear extracts identified 14 protected regions: A, -40 to -33; B, -65 to -42; C, -126 to -110; D, -276 to -255; E, -377 to -364; F, -404 to -384; G, -468 to -455; H, -573 to -554; I, -706 to -680; J, -734 to -716; K, -760 to -743; L, -803 to -773; M, -853 to -829, and N, -903 to -879, as the DNA binding sites for nuclear factors. Five of the regions (B, C, G, H, and K) bind to heat-stable factors. DNA binding gel electrophoretic assays indicated that region N (-903 to -879), which is essential for efficient transcription, binds predominantly a nuclear activity designated AIIN3. This activity is present in cells of hepatic and intestinal origin but absent in HeLa cells. Similar analysis showed that region H (-573 to -554) binds to the liver-specific factor HNF1/LFB1. Deletion of this region decreased hepatic and intestinal transcription 80 and 64% of control, respectively, suggesting that HNF1/LFB1 or a related activity contributes to optimal transcription but is not essential for the tissue-specific expression of the human apoA-II gene.

A 3-week fish oil diet induced in weanling rats a decrease in plasma lipids and liver triacylglycerol, and an increase in insulinemia, compared to a corn oil diet. At the same time, plasma apolipoprotein (apo) A-I was slightly lower and plasma heavy apo B/light apo B ratio was higher in fish-oil-fed than in corn-oil-fed rats. Hepatocytes obtained from fish-oil-fed and corn-oil-fed rats were used to examine how fish oil affects lipid and apolipoprotein synthesis and secretion. Primary culture of hepatocytes from fish-oil-fed rats displayed a lower ability to synthesize and secrete triacylglycerol than hepatocytes from corn-fed rats, as measured by mass determination or [U-14C]glycerol incorporation. Hepatocytes from fish-oil-fed rats exhibited a lower synthesis of cholesterol, measured by [14C]acetate incorporation, than hepatocytes from corn-oil-fed rats. This impairment was associated with an increase in beta-oxidation, a higher channeling of oleic acid into phospholipids, and a lower triacylglycerol/diacylglycerol ratio in hepatocytes from fish-oil-fed rats than in hepatocytes from corn-oil-fed rats. Incorporation of [35S]methionine into secreted apoB was reduced in hepatocytes from fish-oil-fed rats, but was not paralleled by a decrease in apo B mRNA. The appearance of degradative forms of apo B suggest an increase in apo B degradation in hepatocytes from fish-oil-fed rats. Incorporation of [35S]methionine into cellular and secreted apo A-I was lower in hepatocytes from fish-oil-fed rats than in hepatocytes from corn-oil-fed rats, and was not paralleled by any difference in the apo A-I mRNA level. Finally, [35S]methionine incorporation into cellular and secreted forms of apo E and apo A-I mRNA were reduced in hepatocytes from fish-oil-fed rats, compared with hepatocytes from corn-oil-fed rats. These combined data show that fish oil diet reduces triacylglycerol synthesis and secretion and affects apo B synthesis at a post-transcriptional level, and reduces cholesterol synthesis and affects apo E and apo A-I synthesis at a transcriptional and a post-transcriptional level.

Primary cultures of adult rat hepatocytes were used to compare the uptake and esterification of essential polyunsaturated fatty acids (18:2, 20:3 and 20:4 of the n-6 series) with those of palmitic and oleic acids. The uptake of unesterified fatty acids was linearly related to the free fatty acid/albumin molar ratio for 14 h and did not depend on the unbound free fatty acid level. Whatever the initial free fatty acid/albumin molar ratio, it dropped to 0.5 +/- 0.1 mM after 14 h, thus showing that hepatocytes have a high capacity for clearing free fatty acids from the medium at high free fatty acid/albumin molar ratios. The free fatty acid uptake become saturable when the free fatty acid and albumin concentrations were raised and the free fatty acid/albumin ratio remained constant. This strongly suggests that albumin-hepatocyte interaction mediates free fatty acid uptake. This uptake was identical whatever the fatty acid tested and did not depend on the relative amounts of fatty acids when they were added simultaneously. Triacylglycerol accumulation and synthesis, monitored by labelled fatty acids, were related to the free fatty acid/albumin molar ratio and exhibited no specificity for the series of fatty acids tested. Triacylglycerols were enriched in all the fatty acids tested by up to 60%, and fatty acid incorporation into diacylglycerols and triacylglycerols reflected the free fatty acid composition of the medium. By contrast, neither the level nor the synthesis of phospholipids varied with free fatty acid/albumin, but the rate of phospholipid turnover depended on the fatty acids tested. Accumulation of these acids was smaller in phospholipids than in triacylglycerols. When linoleic and arachidonic acids were added together, phospholipids (especially phosphatidylethanolamine and phosphatidylinositol) were more enriched in arachidonic acid than triacylglycerols. This might be due to the specificity for fatty acid of the enzymes involved in phospholipid metabolism.

An apolipoprotein-E (apo-E) cDNA probe, cloned by immunoscreening of a lambda GT11 rat liver cDNA library, was used to further characterize the expression of the apo-E gene in rat liver during development, in relation to plasma insulin and glucagon levels. The apo-E mRNA level was low in fetus liver, then abruptly increased at birth and rose further during the suckling period. It returned to the level at birth in 10-week-old adults. These variations were paralleled with dramatic changes in plasma glucagon, which rose at birth and remained high during suckling. At the same time, the insulin/glucagon molar ratio fell. Administration of N6,O2-dibutyryl cAMP to 5-day-old rats resulted in a significant induction of liver apo-E mRNA. Moreover, liver apo-E mRNA rose in 10-h-fasted suckling rats as compared to controls, while plasma glucagon increased and the insulin/glucagon ratio decreased. Conversely, glucose feeding of suckling rats did not induce any increase in liver apo-E mRNA, the insulin/glucagon ratio was 10-fold higher than in fasted animals. Our results are consistent with liver apo-E gene expression being under the control of plasma glucagon and of the glucagon/insulin balance.

Variations in apoprotein B hepatic expression were evaluated during lactation in dairy cows. Apoprotein B mRNA levels were significantly decreased in early lactation and this was accompanied by triglyceride accumulation in the liver.