Transcriptional regulation of the apolipoprotein A-I gene. Species-specific expression correlates with rates of gene transcription

Species-specific regulation of the alpha-2(I) procollagen gene by proximal promoter elements

Transcriptional regulation of the human alpha 2(I) procollagen proximal promoter involves the interaction of trans-acting factors at the inverted CCAAT box (G/CBE) located at position -80 and an adjacent GGAGGCCC-box at -70. Both these elements have previously been shown to be essential for activity of the human promoter. This study investigated nucleotide differences at three sites (-74, -72 and -71) between the human and mouse promoters that were sufficient to abolish trans-acting factor binding with the mouse sequence (GGAGACGT). Two distinct DNA-protein interactions were detected on the human -107/+54 promoter fragment while a single interaction was observed at the equivalent mouse promoter. One of these factors is the CCAAT-binding factor (CBF) and it's binding was observed on both the human and mouse promoters. Although the GGAGGCCC DNA-binding element was not detected on the mouse promoter, GGAGGCC-binding proteins were present in mouse nuclear extracts as observed by their interaction with the human promoter. Functional analysis of the human and mouse -343/+54 and -107/+54 promoter regions revealed significant differences between species; the human constructs having higher activity than the mouse. The differences in promoter activity between species may in part be a result of the nucleotide differences in the GGAGGCCC-box. Mutations in this region of the human -107/+54 promoter prevented DNA-protein interaction and lowered promoter activity. These results support the hypothesis that the GGAGGCCC-box in the human alpha 2(1) procollagen promoter has a regulatory function and that there exists a species-specific difference in transcription factor binding and regulation of the gene.

Significance of a polymorphism (G-->A transition) in the -75 position of the apolipoprotein A-I gene promoter on serum high density lipoprotein-cholesterol levels in Japanese hyperlipidemic subjects

High density lipoprotein-cholesterol (HDL-C) levels are inversely related to the incidence of coronary artery disease. We studied the influence of a G(-75)-->A transition in the promoter of the apolipoprotein (apo) A-I gene, a major protein component of HDL, on serum HDL-C levels in hyperlipidemic subjects. Seventy three hyperlipidemic subjects with serum levels of high HDL-C (HDL-C > or = 70 mg/dl, Group H) were compared with hyperlipidemic subjects with levels of HDL-C between 40 and 70 mg/dl (Group N) and those with HDL-C < 40 mg/dl (Group L). Group H showed a higher incidence (45.2%) of low plasma cholesteryl ester transfer protein (CETP) activity than Groups N (9.1%) and L (5.3%) (p < 0.001). Group H had a higher incidence of the G(-75)-->A transition (0.275) than Groups N (0.117, p < 0.05) and L (0.056, p < 0.01), among subjects with normal CETP activities. The HDL-C levels in subjects with the transition (84 +/- 16 mg/dl) were higher than those in subjects without the transition (56 +/- 12 mg/dl) (p < 0.05). These data suggest that a G(-75)-->A transition of the apo A-I gene promoter, in addition to the common mutation of CETP gene, contributes to high HDL-C levels among hyperlipidemic patients in Japan.

Transcriptional regulatory sequences within the first intron of the chicken apolipoproteinAI (apoAI) gene

Previous studies demonstrated that the -82 to +87 nucleotides (nt) 5'-upstream region of the chicken apolipoprotein (apoAI) gene are necessary for maximum reporter chloramphenicol acetyl transferase (cat) gene activation in chicken hepatocarcinoma (LMH) cells [Bhattacharyya, N., Chattapadhyay, R., Oddoux, C., Banerjee, D., 1993. Characterisation of the chicken apolipoprotein A-I gene 5'-flanking region. DNA Cell Biol. 12, 597-604]. The -82 to +87nt contain the 5'-untranslated nt, part of the first intron, and the upstream regulatory sequences. In this study, we examined the role of the first intron in the transcriptional regulation of the chicken apoAI gene. Six different reporter cat gene constructs with or without part of the first intron were prepared and transfected into LMH, normal rat kidney (NRK) and human hepatocarcinoma (HepG2) cells. Cell extracts were prepared from each transfected cell line, and CAT activities were measured. All three cell-lines readily expressed CAT, indicating that transcriptional regulatory sequences are present within the first intron region of the chicken apoAI gene. In an enhancer assay, the first intron containing cat construct exhibited a 5.4-fold increase of reporter activity in NRK cells when compared to a SV 40 promoter containing cat plasmid, suggesting the presence of a moderate enhancer element within +29 to +87nt of the first intron. DNase I protection assays, electrophoretic mobility shift assays and binding experiments with nuclear proteins isolated from different chicken tissues and LMH cells showed interaction with +29 to +87nt. Nuclear proteins isolated from tissues like liver and intestine, that actively express apoAI gene, failed to interact with +29 to +87nt, whereas nuclear proteins isolated from tissues that are less active in apoAI gene expression readily interacted with this region. To show the binding of the LMH-specific trans-acting factors to the +50 to +68nt intron region, DNA-affinity chromatography step was performed by using 3H-labeled nuclear proteins. These studies demonstrate that the first intron region of the apoAI gene interacts with trans-acting proteins and plays an important role in transcriptional regulation of the apoAI gene.

The effect on transcription efficiency of the apolipoprotein AI gene of DNA variants at the 5' untranslated region

Elevated circulating levels of high-density lipoprotein and apolipoprotein AI are associated with reduced coronary artery disease risk. We have shown that a C to T substitution at +83 bp and a G to A substitution at -75 bp of the apolipoprotein AI gene are both related to increased high-density lipoprotein levels in a healthy population but not in a coronary population, among whom the same mutations are associated with increased disease severity. In the present study, we explored the effects of these base changes on transcriptional efficiency in vitro. We directionally cloned (using polymerase chain reaction) the 5' region of the apolipoprotein AI gene (-281 to +330 bp) with GC, GT, and AC haplotypes into a pGL3-luciferase reporter gene basic vector, and transfected the constructed vectors into HepG2 cells. The cells carrying the T allele at the +83 bp site (GT 112.3 +/- 12.4) had the same transcriptional efficiency as those bearing the C allele (GC 126.3 +/- 9.6). However, for cells with the A allele at -75 bp there was a twofold decrease in transcription (AC 63.1 +/- 9.3) accompanied by similar changes in Luc+ mRNA levels; this reduced transcription was only present if the apolipoprotein AI leader sequence was included in the insert. While the findings are inconsistent with the T or A allele being associated with higher high-density lipoprotein levels, they are consistent with the finding that the alleles are associated with an increased coronary artery disease risk, and demonstrate that the 5' leader region of the apolipoprotein AI gene participates in regulating apolipoprotein AI transcription. They also suggest that other regions of the apolipoprotein AI gene may have an active role in such regulation, and that environmental effects may influence allele-specific expression.

Characterization and modulation of LP(a) in human hepatoma HEPG2 cells

HepG2 cells have been widely used to study factors which affect the secretion of apoB100 lipoprotein particles. The objectives of this study were to determine if Lp(a) particles were present in conditioned medium from HepG2 cells and if so, was this accumulation affected by factors which alter apoB100 lipoprotein metabolism. The data demonstrate that Lp(a) accumulated in the medium in a time dependent manner over a 48 h incubation period. Ultracentrifugation fractionation and Western blot analysis demonstrated that lipoprotein particles containing apo(a) in complex with apoB100 were present at a density consistent with human plasma Lp(a). Incubation of the HepG2 cells with LDL or VLDL caused increases in Lp(a) accumulation in the medium (+33% +/- 14%, P NS and 56% +/- 21%, P < 0.05, respectively). In contrast, apo(a) mRNA decreased (-17% +/- 3%, P < 0.01 for both LDL and VLDL incubation). Increasing concentrations of amino acids in the medium resulted in progressively less Lp(a) and apoB100 in the medium, the effect being greater on apoB100. ApoB100 mRNA levels decreased with incubation of HepG2 cells with amino acids (-22% +/- 10%, P < 0.06) whereas apo(a) mRNA levels increased significantly (+47% +/- 14%, P < 0.005). Taken together, our data show that HepG2 cells express mRNA for apo(a), and accumulate Lp(a) in the medium. The close correlation of medium Lp(a) levels with medium apoB100 levels, and not with apo(a) mRNA levels, suggests that medium Lp(a) accumulation may be a function of lipoprotein synthesis and secretion and is consistent with extracellular assembly of Lp(a) lipoprotein particles.

9-cis-retinoic acid increases apolipoprotein AI secretion and mRNA expression in HepG2 cells

HepG2 cells were studied as a model for regulation of hepatic apolipoprotein AI (apo AI) secretion and gene expression by 9-cis-retinoic acid. HepG2 cells cultured on plastic dishes were exposed to 9-cis-retinoic acid (9-cis-RA) for 48 h with a complete media change at 24 h. Apo AI mass in cultured media was determined by ELISA, by quantitative immunoblotting and by steady-state 35S-methionine labeling. Messenger RNA levels were determined by RNase protection using probes for apo AI and the housekeeping gene, glyceraldehyde 3-phosphate dehydrogenase (G3PDH). 9-cis-RA increased secretion of apo AI by 52% at doses of 10 and 1 microM (6.3 +/- 0.6 vs. 4.2 +/- 0.3; P < 0.005; 6.1 +/- 0.3 vs. 4.0 +/- 0.7 ng of apo AI/mg cell protein, P < 0.05) and by 35% at 0.1 microM (5.5 +/- 0.6 vs. 4.1 +/- 0.4 ng apo AI/mg protein, P < 0.05, n = 4). Immunoblotting results were consistent with results from ELISA (70% increase at 10 microM 9-cis-RA, P < 0.001; 34% increase at 1 microM, P < 0.005, n = 3). Metabolically labeled apoAI in the medium was increased by 39% following steady-state labeling in the presence of 10 microM 9-cis-RA (597 +/- 7 vs. 430 +/- 13 DPM/microliters media; P < 0.001; n = 4). 9-cis-RA (10 microM) also increased HepG2 cell apo AI mRNA expression by 76% (68 700 +/- 400 vs. 38 900 +/- 2700 DPM, P < 0.01, n = 4), whereas expression of G3PDH mRNA was slightly decreased (14%, P < 0.05). Thus, 9-cis-RA stimulates apo AI expression in HepG2 cells, suggesting a role for retinoids in activating endogenous apo AI gene expression.

Species-specific polymorphism in the promoter of the apolipoprotein A-I gene: restoration of human transcriptional efficiency by substitution at positions -189, -144 and -48 bp

Previous studies indicate that species-specific differences in apolipoprotein A-I (apo A-I) expression could be largely explained by cis-acting factors located within or near the 5' flanking region (-231 to +223 bp, where +1 is the start site of transcription). In the present studies, we have localized 7 sites within the (-231 to -15 bp) region of the African green monkey apo A-I gene that differ from the human apo A-I gene 5' flanking region. To identify which of the 7 polymorphic sites were essential for the species-specific differences in apo A-I gene expression, mutated promoter constructs were transfected into HepG2 cells and reporter gene expression was measured. Each of the 7 sites within a defined 5' flanking region of the human gene was individually mutated to the African green nucleotide sequence found at that position. Three of the sites (-189, -144 and -48) were found to raise the human apo A-I promoter activity to approx. 60-65% of the African green promoter. While double mutations (-144/-48 bp and -189/-144 bp), restored the human apo A-I promoter activity to 100% of that found with the African green monkey promoter. Additional studies revealed similar DNA: protein interactions with DNA probes from either human or African green monkey and HepG2 cell nuclear extract. In conclusion, these studies demonstrate that double and triple nucleotide substitutions within the human apo A-I promoter are sufficient to restore gene expression in HepG2 cells to levels seen with the African green monkey promoter. These data suggest that sites -189, -144 and -48 bp are involved in significantly altering the binding affinity of a nuclear factor determining the species-specific level of apo A-I gene transcription.

Adenine for guanine substitution -78 base pairs 5' to the apolipoprotein (APO) A-I gene: relation with high density lipoprotein cholesterol and APO A-I concentrations

A common mutation, adenine (A) for guanidine (G) substitution (G/A) has been located -78 bp 5' to the apo A-I gene. This region has been shown to be involved in the transcriptional regulation of the apo A-I gene. Previous studies have shown that this mutation is associated with altered high density lipoprotein cholesterol (HDL-C) levels, although these findings have not been consistent. We have studied the frequency of this mutation in 125 subjects (60 males and 65 females) selected because they had HDL-C levels below the 25th (low HDL) or above the 75th (high HDL) percentile of the population distribution. The presence of the mutation was detected by Msp I digestion of a 259 bp fragment of PCR amplified DNA. The allele frequency was similar in both groups (0.20 for the lowest HDL group and 0.28 for the highest HDL group, p > 0.05), although a non-significant trend was observed in a higher frequency of the A/A genotype in the highest HDL females (17.5%) vs only 6.7% in the lowest HDL female group. In conclusion, in this population the G/A mutation was not significantly associated with HDL-C or apo A-I plasma levels.

Increased production of apolipoprotein A-I associated with elevated plasma levels of high-density lipoproteins, apolipoprotein A-I, and lipoprotein A-I in a patient with familial hyperalphalipoproteinemia

Familial hyperalphalipoproteinemia (FHA) is a heritable trait associated with elevated plasma concentrations of high-density lipoprotein (HDL) cholesterol and possibly with longevity and protection against coronary heart disease (CHD). The metabolic basis and molecular etiology of FHA have not been established in most kindreds. The proband of a kindred with FHA and possible longevity was found to have elevated plasma levels of HDL cholesterol, apolipoprotein (apo) A-I, and lipoproteins containing apo A-I without apo A-II (Lp A-I), but normal levels of apo A-II and lipoproteins containing apo A-I with apo A-II (Lp A-I:A-II). The in vivo kinetics of apo A-I and apo A-II were studied in the FHA proband and in control subjects using both exogenous radiotracer (125I-apo A-I and 131I-apo A-II) and endogenous stable isotope (primed constant infusion of 13C6-phenylalanine) labeling techniques. The production rate (PR) of apo A-I was markedly increased in the FHA subject (28.9 mg/kg.d) compared with the control subjects (12.0 +/- 2.1 mg/kg.d), whereas the apo A-II PR was not substantially increased. The primary sequence of the proband's apo A-I gene, including 1.2 kb of the 5'-flanking sequence, was normal. We conclude that a selective upregulation of apo A-I production is one metabolic cause of FHA, and results in high plasma concentrations of HDL cholesterol, apo A-I, and Lp A-I and possibly in protection from atherosclerotic CHD.

Characterization of the chicken apolipoprotein A-I gene 5'-flanking region

Apolipoprotein A-I (apoA-I) is a major protein component of plasma high-density lipoprotein in all species studied, and plays an important role in cholesterol homeostasis. In an earlier study, we cloned and structurally characterized the chicken apoA-I gene. In this study, the 5'-flanking region of the chicken apoA-I gene was sequenced and functionally characterized. Sequence analysis of the 510-nucleotide 5' upstream region revealed the presence of TATA and CCAAT boxes. In addition, we identified binding sites for several transcription factors such as Sp1, AP1, and NFI.2. When the 5' fragment was ligated into a promoterless CAT vector and transfected into a chicken hepatocarcinoma cell line (LMH), the bacterial chloramphenicol acetyl transferase (CAT) gene was expressed, suggesting transcriptional regulation is associated with this region. Transfection studies with other 5' deletion constructs revealed that the sequence spanning the region -82 to +87 contained the major transcriptional activity. DNase I footprinting, gel retardation, and Southwestern blot analyses showed that the fragment interacts with nuclear proteins.