Distal Apolipoprotein C-III Regulatory Elements F to J Act as a General Modular Enhancer for Proximal Promoters That Contain Hormone Response Elements
Synergism Between Hepatic Nuclear Factor-4 Molecules Bound to the Proximal Promoter and Distal Enhancer Sites
Transient transfection assays have shown that the distal apoC-III promoter segments that contain the regulatory elements F to J enhance the strength of the tandemly linked proximal apoA-I promoter 5- to 13-fold in hepatic (HepG2) cells. Activation in intestinal (CaCo-2) cells to levels comparable to those obtained in HepG2 cells requires a larger apoA-I promoter sequence that extends to nucleotide −1500 as well as the presence of hepatic nuclear factor-4 (HNF-4). The distal apoC-III regulatory elements can also enhance 4- to 8-fold the strength of the heterologous apoB promoter in HepG2 and CaCo-2 cells. Finally, these elements in the presence of HNF-4 enhance 14.5- to 18.5-fold the strength of the minimal adenovirus major late promoter linked to two copies of the hormone response element (HRE) AID of apoA-I in both HepG2 and CaCo-2 cells. In vitro mutagenesis of the promoter/enhancer cluster established that the enhancer activity is lost by a mutation in the HRE present in the 3′ end of the regulatory element I (−736 to −714) and is reduced significantly by point mutations or deletions in one or more of the regulatory elements F to J of the apoC-III enhancer. The enhancer activity also requires the HREs of the proximal apoA-I promoter. The apoC-III enhancer can also restore the activity of the proximal apoA-I and apoB promoters that have been inactivated by mutations in CCAAT/enhancer binding protein binding sites, indicating that C/EBP may not participate in the synergistic activation of the promoter/enhancer cluster. The findings suggest that the regulatory elements F to J of the apoC-III promoter act as a general modular enhancer that can potentiate the strength of proximal promoters that contain HREs. Such potentiation in the HepG2 cells can be accounted for by synergistic interactions between HNF-4 or other nuclear hormone receptors bound to the proximal and distal HREs and SP1 or other factors bound to the apoC-III enhancer. Additional factors may be required for optimal activity in CaCo-2 cells as well as for the function of this region as an intestinal enhancer.
- apoC-III enhancer
- transcriptional synergism
- hepatic and intestinal expression
- apoA-I gene regulation
- hepatic nuclear factor-4 (HNF-4)
- Received December 19, 1995.
- Revision received May 3, 1996.
Recent studies have established the location of the regulatory elements and the nature of the factors that are important for the transcriptional regulation of the genes of the apoA-I/apoC-III/apoA-IV cluster.1 2 3 4 5 The apoA-I promoter consists of four regulatory elements within the proximal −220 to +17 region designated A to D.2 3 Elements B (−128 to −77) and D (−220 to −190) contain HREs and can bind a variety of orphan and ligand-dependent nuclear receptors.6 Element C binds heat-stable activities related to C/EBP, which act as positive regulators, as well as heat-labile activities, one of which acts as a negative regulator.2 The apoC-III promoter consists of four proximal (A to D) and six distal (E to J) regulatory elements between nucleotides −792 and −25.1 Element B (−87 to −72) contains an HRE and can bind a variety of hormone and ligand-dependent nuclear receptors. Elements C and D bind heat-stable activities related to C/EBP as well as heat-labile activities.7 The distal apoC-III promoter, which enhances the strength of the genes of the apoA-I/apoC-III/apoA-IV cluster, predominantly binds the ubiquitous factor SP1 in elements F, H, and I. Element G binds factors related to the orphan receptors ARP-1 and V-erb A–related-3, and element I binds HNF-4 between nucleotides −736 and −714. A new activity designated C-IIIJ1 binds to the regulatory element J, and a minor activity as well as HNF-4 binds to the regulatory element I.5 8 Finally, the proximal apoA-IV promoter consists of four regulatory elements located between nucleotides −22 and −274. Element C (−148 to −92) contains an HRE and can bind a variety of nuclear hormone receptors.4
Consistent with these in vitro observations, expression of segments of the apoA-I/C-III/A-IV gene cluster in transgenic mice showed that the proximal apoA-I or apoC-III promoters extending to nucleotides −255 and −200, respectively, are sufficient for hepatic expression of these two genes in vivo.8 9 10 However, the intestinal expression of the apoA-I/apoC-III/apo A-IV gene cluster requires other regulatory sequences localized in the 6-kb intergenic region separating the apoA-I and apoA-IV genes.11 In the case of the apoA-I gene, these sequences were recently shown to correspond to the regulatory elements F to J of the apoC-III promoter.8
In humans, apoA-I synthesis occurs predominantly in the liver, apoC-III synthesis occurs mainly in the liver and to a lesser extent in the intestine, and apoA-IV synthesis occurs mainly in the intestine and to a lesser extent in the liver.12 13 Epidemiological, genetic, and transgenic animal studies suggest that increased levels of apoA-I and apoA-IV are protective against atherosclerosis,14 15 16 whereas increased levels of apoC-III may contribute to hypertriglyceridemia.10 17 Thus, a better understanding of the basic mechanisms that control the expression of these genes is important.
In the present study we demonstrate that the distal apoC-III regulatory elements F to J can act as a general modular enhancer in hepatic (HepG2) cells and activate proximal promoters that contain HREs. It appears that enhancement of transcription requires synergistic interactions between HNF-4 or other nuclear hormone receptors that bind to the proximal promoter and the distal enhancer element I. Favorable interactions between nuclear hormone receptors bound to proximal and distal sites are brought about by SP1 and other factors that bind to the apoC-III enhancer. The apoC-III elements F to J also act as enhancers of the apoA-IV, apoC-III, and apoB promoters in intestinal (CaCo-2) cells. In the case of the apoA-I gene, high levels of expression in CaCo-2 cells require longer apoA-I promoter segments and the presence of HNF-4.
Restriction enzymes were purchased from Minotech. T4 DNA ligase, calf intestinal alkaline phosphatase, the Klenow fragment of DNA polymerase I, and polynucleotide kinase were purchased from New England Biolabs. The sequenase kit was purchased from United States Biochemicals. Peptone and yeast extract were purchased from Merck. o-Nitrophenyl-β-d-galactopyranoside and TLC silica gel plates for CAT assays (T-6770) were purchased from Sigma. [γ-32P]ATP, [α-32P]dCTP, [α-32P]dGTP, and [14C]chloramphenicol were purchased from ICN Biochemicals. Agarose (Seakem) was purchased from FMC, and cesium chloride (ultrapure) was purchased from BRL. Vent polymerase was purchased from New England Biolabs, and dNTPs (ultrapure 100 mmol/L) were purchased from Pharmacia LKB.
The prototype plasmid pBS(−255/−5)A-I.CAT was constructed as follows: The plasmid pEMBL(−255)A-I.CAT2 was digested with Sma I and EcoRI, and the excised fragment that contained the apoA-I gene promoter sequence between nucleotides −255 and −5 plus part of the CAT coding sequence was subcloned at the Sma I and EcoRI sites of pBluescript KS(+) vector (Stratagene). The remaining CAT gene plus SV40 splicing and polyadenylation signals were supplied from the pUCSH.CAT vector1 as an EcoRI fragment.
The reporter plasmids C-III(−1287/−197).A-I.CAT, A-I.CAT.C-III(−1287/−197), and A-I.CAT.C-III(−197/−1287) are all derivatives of the prototype plasmid and were constructed as follows: The apoC-III gene 5′ flanking region containing nucleotides −1287 to −197 was excised from plasmid −1411 apoC-III.CAT1 as a Pst I fragment and was inserted, in two steps, either upstream of the apoA-I promoter (at the Sma I site of plasmid pBS[−255/−5]A-I.CAT) or downstream of the CAT gene (at the EcoRV site of the same plasmid) in both orientations. The reporter plasmids containing deletions or nucleotide substitutions in the apoC-III enhancer designated −890, −755, −686, FM1, GM1, HM1, IM1, IM4, JM1 (Table 1⇓) were constructed similarly. In each case, the mutated apoC-III enhancer was excised from the corresponding apoC-III.CAT plasmid1 5 by digestion with Xba I and Pst I, blunted with Klenow DNA polymerase, and subcloned at the EcoRV site of plasmid pBS(−255/−5)A-I.CAT. Reporter plasmids designed to test the effect of the apoC-III enhancer on various apoA-I promoter mutations were constructed as follows: The mutagenized apoA-I promoters were excised from the corresponding apoA-I CAT plasmids A-ICM2CAT2 and A-IDmutCAT with Xba I and Xho I, blunted with Klenow, and inserted at the Sma I site of plasmid pBSCATC-III(−197/−1287). The latter plasmid contains a promoterless CAT gene and the apoC-III enhancer region −197/−1287 placed at the 3′ end of CAT. Plasmid A-ICM2CAT contains nucleotide substitutions in element C at positions −171/−166. Plasmid A-IDmutCAT contains nucleotide substitutions in element D at positions −204/−196 (Table 1⇓). To produce the −1500 apoA-I constructs, the apoA-I 5′ flanking region −1500 to −255 was excised from pEMBL.(−1500)A-I.CAT as an Sma I fragment and inserted at the Sma I site of the appropriate vectors.
The plasmids containing apoB promoter/apoC-III enhancer fusions (−267/+8 apoB CAT C-III [−197/−1287], −1800/+8 apoB CAT C-III [−197/−1287], apoBM2 CAT C-III [−197/−1287], and apoBM6 CAT C-III [−197/−1287]) were constructed by inserting the Pst I fragment containing the apoC-III enhancer (−197/−1287) into the corresponding apoB CAT plasmids at the unique Pst I site that is located at the 3′ end of the CAT gene.18 ApoBM2 and apoBM6 are mutagenized apoB promoters (−267/+8) that contain nucleotide substitutions at positions −77 to −73 and −61 to −60, respectively (Table 1⇑).
The plasmid AI(D2).AdML.CAT was constructed by inserting two copies of the apoA-I element D (−220/−190) of the apoA-I promoter at the Sal I site of AdML-CAT vector (AdML.CAT). This vector contains the AdML TATA box and the transcription start site upstream of the CAT gene. The plasmid C-III-AI(D2)-AdML-CAT was constructed by inserting two copies of the element D of the apoA-I promoter in the Sal I site of plasmid C-III.AdML.CAT, which contains, in addition to AdML, the apoC-III enhancer (−890/−590) upstream of the TATA box.5 The eukaryotic expression vector pMT2-HNF-4 consists of the cDNA of HNF-4 under the control of the AdML. The construction of this vector has been reported previously.19 The plasmid pRSV-βgal contains the β-galactosidase gene under the control of the Rous sarcoma virus LTR and was used for the normalization in the transfection efficiency.20
Cell Culture and Transfections
Monolayer cultures of HepG2 and CaCo-2 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% and 20% heat-inactivated fetal calf serum, respectively. Twenty-four hours before transfection, the cells were seeded at 50% to 60% confluence into 35-mm-diameter dishes. Reporter constructs (2 μg) along with 2 μg of either pMT2-HNF-4 or control pMT2 vector plus 3 μg pRSV-βGal plasmid were introduced to the cells by the calcium phosphate coprecipitation method.21 Forty-eight hours later, the cells were harvested and subjected to three consecutive freeze-thaw cycles. CAT activities were assayed by the use of constant amounts of protein as described previously.22 The β-galactosidase activity in the cellular extracts was determined as described,20 and the values were used to normalize the variations in transfection efficiencies. Incubation times were carefully selected by titration and kinetic experiments to ensure that the enzyme reactions were performed at the linear range. The values reported represent the average of four to six determinations obtained in two to three independent experiments.
Expression of HNF-4 cDNA in Bacteria and Footprinting Analysis of ApoC-III Promoter
HNF-4 cDNA was cloned in the bacterial pET15b vector under the control of the T7 promoter and was expressed in Escherichia coli BL21 strain. The bacterial extracts obtained were used for footprinting analysis of the −283 to +24 and −890 to −686 apoC-III promoter, as described previously.2 For this analysis, the −890 to −686 and −283 to +24 apoC-III promoter fragments were labeled at either the −283 or −890 nucleotide with [γ-32P]ATP and T4 polynucleotide kinase. Footprinting was performed as described.1
Distal ApoC-III Promoter Region Enhances Strength of Tandemly Linked Proximal ApoA-I Promoter −255 to −5 in HepG2 Cells; Transcriptional Activation in CaCo-2 Cells Requires Additional Upstream Elements and Presence of HNF-4
The apoC-III regulatory region −1287 to −197, which contains the regulatory elements F to J, was cloned 5′ or 3′ in both orientations relative to the apoA-I gene promoter fragment −255/−5 that was linked to the CAT reporter gene. Transient transfections in HepG2 and CaCo-2 cells showed that this apoC-III promoter region acts as an enhancer that increases the strength of the proximal apoA-I promoter. The enhancement in HepG2 cells is approximately 13-fold when the apoC-III promoter is cloned 5′ of the apoA-I promoter and 5-fold when it is cloned 3′ of the CAT gene in either orientation (Fig 1A and 1C⇓⇓). The proximal −255 to −5 and the −1500 to −5 of the apoA-I promoter had low levels of activity in CaCo-2 cells in both the presence and the absence of the enhancer. This activity could be increased to levels comparable to that of HepG2 cells in the presence of HNF-4 (Fig 1B and 1D⇓⇓). In subsequent experiments we have used the −1500 apoA-I promoter/apoC-III enhancer cluster to evaluate its properties in CaCo-2 cells.
Contribution of Distal ApoC-III Regulatory Elements and Proximal HREs to ApoA-I Promoter Strength in HepG2 Cells
The contribution of the apoC-III regulatory elements to the strength of the proximal apoA-I promoter in HepG2 cells was evaluated by transient transfection experiments with the use of promoter constructs containing 5′ deletions. This analysis showed that deletion of the 5′ apoC-III promoter region extending to nucleotides −890 increased by 30% the activity of the apoA-I promoter/apoC-III enhancer cluster. The −197 to −899 apoC-III promoter region contains the regulatory elements E to J.1 The promoter/enhancer activity was nearly abolished by deletion of the regulatory elements J, I, and H (Fig. 2A⇓). Recent studies have shown that the regulatory element H and I predominantly bind SP1 and that the regulatory element J binds a new activity designated C-IIIJ1.5 DNA binding assays have shown that the 3′ end of element I can bind HNF-4, ARP-1, and RXRα-T3Rβ heterodimers (References 8 and 23 and S. Lavrentiadou and V. Zannis, unpublished data, 1996). The contribution of the distal apoC-III regulatory elements to the enhancer activity was also evaluated by point mutations that abolish the binding of specific factors to their cognate sites. This analysis showed that the enhancer activity was abolished by point mutations in the HRE of element I, which bind HNF-4. The locations and boundaries of the HNF-4 binding sites on the apoC-III promoter were determined by footprinting analysis of the −890 to −686 and −283 to +24 apoC-III promoter regions with the use of bacterially expressed HNF-4 (Fig 2C⇓). This analysis showed that HNF-4 has two distinct binding sites on the apoC-III promoter. One that was recognized previously is between nucleotides −87 and −72, and the other is between nucleotides −736 and −714 (Fig 2C⇓). The mutation IM4 shown in Table 1⇑ abolished the binding of HNF-4 and other nuclear hormone receptors to the HRE of element I. Furthermore, the activity of the apoC-III enhancer/apoA-I promoter cluster was reduced to 40% to 45% of its value by mutations in elements H and G and to 55% to 70% of its value by mutations in elements I, J, or F. Element G binds activities related to orphan receptors ARP-1 and EAR-3, and element F binds SP1 as a major and C-IIIJ1 as a minor activity.5 The findings indicate that the enhancer activity depends on HNF-4, which binds to the regulatory element I. In addition, all the other factors that bind to the upstream apoC-III promoter region are required to enable it to activate the closely linked apoA-I promoter optimally. Similar mutagenesis analysis showed that the ability of the apoC-III enhancer to activate transcription driven by the proximal apoA-I promoter is greatly affected by mutations in the regulatory element D of apoA-I (Fig 2B⇓). This mutation reduced the strength of the promoter/enhancer complex to 6% of its original value. The regulatory element AID contains an HRE and binds HNF-4 as well as a variety of other orphan and ligand-dependent nuclear hormone receptors.6 In contrast, mutations in the regulatory element C of apoA-I, which abolished the binding to this region of heat-stable activities related to C/EBP,2 reduced the strength of the promoter/enhancer complex to only 65% of its original value.
Contribution of ApoC-III Regulatory Elements and Proximal HRE to HNF-4–Mediated Transactivation of the Promoter/Enhancer Complex in CaCo-2 Cells
The −1500 apoA-I promoter/apoC-III enhancer CAT constructs were used in cotransfection experiments with plasmids expressing HNF-4 to assess the effects of mutations in the proximal HREs, as well as the distal apoC-III promoter elements (Table 1⇑) on the HNF-4–mediated transactivation in CaCo-2 cells. As shown in Fig 1B⇑, these constructs could be transactivated optimally with HNF-4 compared with shorter apoA-I promoter constructs. In general, the mutations affected the HNF-4–mediated transactivation of the −1500 apoA-I promoter/apoC-III enhancer cluster in CaCo-2 cells to the same extent that they affected the strength of the −255 to −5 apoA-I promoter/apoC-III enhancer cluster in HepG2 cells. The wild-type −1500 apoA-I promoter/apoC-III enhancer cluster was transactivated 7-fold by HNF-4. The transactivation was reduced to 40% and 45% of its original value by deletion of the regulatory element J and by point mutations in element H, respectively, and was nearly abolished by deletions of elements J, I, and H. Point mutations in elements J or F and I reduced the transactivation of the promoter/enhancer cluster to 65% and 90% of its original value, respectively (Fig 3A⇓). Mutations in the regulatory element AID (HRE) of the proximal apoA-I promoter reduced the transactivation of the promoter/enhancer cluster to 7% of its original value, whereas mutations in the regulatory element AIC of the proximal apoA-I promoter did not affect the HNF-4–mediated transactivation of the promoter/enhancer cluster in CaCo-2 cells (Fig 3B⇓). The findings suggest that the HNF-4–mediated transactivation of the apoA-I promoter/apoC-III enhancer cluster in CaCo-2 cells is promoted by interactions between HNF-4, which binds to the HRE of the proximal apoA-I promoter, and the distal enhancer. These interactions are optimized by SP1 and the other factors that bind to the regulatory elements F to J of the apoC-III enhancer.
Distal ApoC-III Regulatory Elements Act as General Modular Enhancer and Potentiate Strength of Heterologous Promoters That Contain HREs
The concept that the apoC-III enhancer operates through synergistic interactions between upstream activators bound to the enhancer and nuclear hormone receptors bound to proximal sites was further reinforced by cotransfection experiments involving heterologous promoters. The apoB promoter was shown previously to contain both an HRE as well as a C/EBP binding site on element A.18 Cotransfection experiments using apoB promoter/apoC-III enhancer CAT constructs showed a 4- to 8-fold activation in HepG2 cells and a 7- to 10-fold activation in CaCo-2 cells, depending on the length of the apoB promoter construct. HNF-4 transactivated further the apoB promoter/apoC-III enhancer cluster in CaCo-2 cells (Fig 4A⇓) but not in HepG2 cells (data not shown). The enhancement of transcription of the apoB promoter/enhancer cluster achieved in CaCo-2 cells in the presence of HNF-4 was 13- to 19-fold. The −1800 to +8 apoB promoter constructs provided stronger activation than the −267 to +8 apoB promoter constructs in both HepG2 and CaCo-2 cells (Fig 4A⇓). More importantly, mutation in the HRE of element A abolished the activity of the promoter/enhancer cluster, whereas mutation in the adjacent C/EBP binding site reduced the activity of the promoter/enhancer cluster to 33% and 50% of its original value in CaCo-2 and HepG2 cells, respectively. The activity of the mutated promoter/enhancer cluster increased to 52% of the control value in CaCo-2 cells in the presence of HNF-4 (Fig 4B⇓).
The preceding analysis of the proximal apoA-I and apoB promoter mutations indicated that the apoC-III enhancer has the ability to bypass partially or totally the effect of mutations in C/EBP sites that inactivate the proximal promoters, but it loses its function by mutations in the proximal HREs. This is clearly illustrated in the cotransfection experiments of Fig 5A and 5B⇓⇓. This analysis shows that mutations in the C/EBP binding site of the apoA-I and apoB promoter reduced the proximal promoter strength to 8% and 13%, respectively. Fusion of the mutated constructs with the apoC-III enhancer increased the strength of the mutated apoA-I and apoB promoter/enhancer cluster to 38- and 16-fold, respectively, in HepG2 cells. The levels of transcription achieved by the mutated promoter/enhancer cluster is 2- to 3-fold higher than that achieved with the wild-type promoters, which lack the apoC-III enhancer. The findings show that despite the inability of C/EBP to bind to its cognate site in either of the two mutated promoters, the synergism between the remainder of the factors that bind to the proximal promoters and the distal enhancer persists.
The synergistic or antagonistic interactions between nuclear hormone receptors bound to the HREs of element AID of the proximal apoA-I promoter and C-III-I of the apoC-III enhancer can also be visualized by the minimal AdML promoter linked to two copies of the regulatory elements AID (HRE) of apoA-I, the distal apoC-III regulatory elements F to J, or both (Fig 6A and 6B⇓⇓). Cotransfection experiments using these constructs and plasmids expressing HNF-4 showed that the strength of the minimal AdML promoter linked to either two copies of the regulatory elements AID (HRE) of apoA-I or the distal apoC-III elements F to J increases moderately up to 3-fold by HNF-4, depending on the cell line. However, fusion of both sets of the regulatory elements to the 5′ region of the minimal AdML promoter increased 18.5- and 14.5-fold the strength of the minimal promoter in HepG2 and CaCo-2 cells, respectively (Fig 6A and 6B⇓⇓). The change in transactivation achieved by both elements is 600% in HepG2 and CaCo-2 cells compared with the sum of the transactivation achieved by the HREs and the apoC-III enhancer alone (Fig 6A and 6B⇓⇓). The findings suggest that HNF-4, which can bind to the proximal and distal HREs and SP1 and the other factors that bind to the enhancer, can transactivate synergistically the minimal AdML promoter. A putative mechanism of transactivation of the apoA-I promoter/apoC-III enhancer cluster, which is consistent with the currently available data, is shown in Fig 6C⇓. This figure shows potential direct protein-protein interactions among the factors bound to the promoter/enhancer cluster as well as interactions with the proteins of the basal transcription complex. Similar mechanisms of transcriptional activation may apply for the other two genes of the cluster (Fig 7⇓). Interestingly, in the absence of HNF-4, the two apoA-I HREs were unable to activate the AdML promoter in HepG2 or CaCo-2 cells. The apoC-III enhancer alone containing the regulatory elements F to J activated 6- and 8-fold, respectively, the AdML promoter HRE cluster in HepG2 and CaCo-2 cells. However, the combination of both the apoA-I HREs and the apoC-III enhancer decreased the strength of the AdML promoter/apoC-III enhancer cluster. Thus, the activity of the AdML promoter linked to the apoA-I HRE and the apoC-III enhancer is 50% to 70% compared with the sum of the activities of the AdML linked to the individual elements. This finding suggests that in both cell types, the proteins that bind to the enhancer and the mixture of nuclear receptors that occupy the HRE in the two cell types act antagonistically, leading to a small reduction in transcription. This finding also indicates that the types of the nuclear hormone receptors that can occupy the apoA-I HRE may determine the synergistic or antagonistic interactions between the other factors that bind to the enhancer and the nuclear hormone receptors that occupy the proximal and distal sites.
Distal ApoC-III Regulatory Elements F to J Act as General Modular Enhancer for Genes That Contain Proximal HREs
The present study demonstrates that the distal apoC-III promoter region, when it is linked upstream or downstream of the apoA-I promoter in tandem or reverse orientation, increases the strength of the apoA-I promoter 5- to 13-fold in HepG2 cells. This property classifies the distal apoC-III promoter region as a transcriptional enhancer of the apoA-I gene. Previous studies have shown that this enhancer activity also operates on the other two genes of the cluster.1 4 5
The enhancer activity requires all the regulatory elements F to J of the distal apoC-III promoter and depends on the HRE that is present on the 3′ end of the element CIII I. This element contains two direct repeats −736AGTGGG(TCCAG)AGGGCA-720 on the coding strand separated by five spacer nucleotides (shown in parentheses). This sequence has homology to the consensus half-site AGG/TTCA motif, which has sequence homology to a hormone nuclear receptor binding site.26 27 Footprinting analysis in the present study, in which bacterially expressed HNF-4 as well as DNA binding assays were used, has shown that this element binds HNF-4, ARP-1, and RXRα-T3Rβ heterodimers (References 8 and 23 and S. Lavrentiadou and V. Zannis, unpublished data, 1996). In addition to the HNF-4 site, optimal enhancer activity requires the presence of intact SP1 sites as well as sites for other factors that bind to the enhancer. The study establishes that the enhancer acts as a module and that the integrity of the entire unit is essential for its functions. Segments of the enhancer are not sufficient to confer full enhancer activity in HepG2 or CaCo-2 cells in vitro. The apoC-III enhancer is also able to increase the strength of the proximal apoB promoter. A common feature of the human apoA-I, apoC-III, and apoA-IV promoters is that they contain one or more HREs in their proximal promoter regions that are essential for their function.1 4 18 19 Recent studies have also shown that apoC-III promoter constructs containing the entire or portions of the apoC-III/apoA-IV intergenic sequences24 could activate the proximal apoA-I promoter in HepG2 or CaCo-2 cells,8 25 but these studies did not assess the contribution of the individual elements of the apoC-III enhancer to the strength of the promoter/enhancer cluster. One of the studies showed that additional elements extending between nucleotides −255 and −595 contribute to the intestinal expression of the apoA-I gene in vitro.25 The present study demonstrates that promoter/enhancer constructs containing upstream 5′ regulatory sequences extending to nucleotide −1500 show much greater transactivation by HNF-4 in CaCo-2 cells than shorter constructs extending to nucleotide −255. These observations suggest that factors bound to upstream apoA-I regulatory regions may be required for optimal expression of the apoC-III gene in intestinal cells. They also indicate that HNF-4 concentrations in CaCo-2 cells, under the experimental conditions used, may be limited. In this regard, in vivo footprinting analysis of the proximal apoA-I promoter in CaCo-2 cells showed gradual protection of the apoA-I HREs as a function of the age of the culture. Optimal protection occurs after a 3-week culture period (D.K. and V.Z., unpublished data, 1996). It was shown previously that the nuclear activities that bind to the HRE of the apoA-I are induced and are maintained at high levels for a period of 1 to 3 weeks in postconfluent CaCo-2 cultures.28
Distal ApoC-III Regulatory Elements F to J Act as General Modular Enhancer
The apoC-III enhancer activates genes through interactions of HNF-4 or other nuclear hormone receptors bound to proximal and distal sites; these interactions are optimized by SP1 and other factors bound to the apoC-III enhancer.
Additional insights on the mechanism of the activation of target genes by the apoC-III enhancer were obtained by the use of the minimal heterologous AdML system as well as by analysis of promoter/enhancer constructs that carry mutations in non-HRE sites on the proximal target promoters. The first set of experiments showed that the minimal AdML promoter linked either to the apoA-I HRE or the apoC-III enhancer was moderately transactivated by HNF-4. However, linkage of both elements to the AdML promoter increased the strength of the minimal promoter 14.5- and 18.5-fold in CaCo-2 and HepG2 cells, respectively, in the presence of HNF-4. The increase in the promoter strength conferred by the two elements is not additive but synergistic. This implies synergistic interactions between HNF-4, which binds to the proximal promoter and distal enhancer HREs. It should be noted that the heterologous promoter construct containing the two apoA-I HREs and the apoC-III enhancer is transactivated by HNF-4 in HepG2 cells, whereas the proximal apoA-I promoter apoC-III enhancer cluster is not. This finding is compatible with previous studies which have shown that the apoA-I HRE in the context of a minimal promoter is transactivated severalfold by ligand-dependent nuclear receptors,29 30 31 whereas it may be either activated moderately or repressed by the same receptors in the context of the entire promoter.6 Interestingly, in the absence of HNF-4 the combination of the two apoA-I HREs and the apoC-III enhancer had 50% to 70% of the activity compared with the sum of the activities of the AdML linked to the individual elements. This finding indicates that the interactions of the nuclear hormone receptors that bind to the proximal promoter and the distal enhancer sites can be either synergistic or antagonistic. In the present case, synergism is observed when the HRE is occupied by HNF-4, and antagonism is observed when the HRE is occupied by the mixture of nuclear hormone receptors that are present in the nucleus of the HepG2 and CaCo-2 cells.
The second set of experiments showed that the apoC-III enhancer compensates for a mutation in a C/EBP binding site that inactivates the proximal apoA-I or apoB promoters. We have shown previously that a mutation in the regulatory element C of apoA-I or the regulatory element A of apoB that prevents the binding to these sites of C/EBP and other heat-stable activities drastically decreased the promoter strength.18 Linkage of these mutated promoters with the apoC-III enhancer increased the strength of the mutated apoA-I and apoB promoter/enhancer constructs 38- and 16-fold, respectively. In addition, the mutated apoA-I and apoB promoter/apoC-III enhancer constructs were transactivated by HNF-4 in CaCo-2 cells 50% to 100% compared with the wild-type promoter. The findings reinforce the concept that although C/EBP may contribute to the optimal strength of the promoter/enhancer cluster, it is not required for the transcriptional synergism between the factors that bind to the enhancer and the different types of nuclear hormone receptor that can occupy the proximal promoter.
The overall picture that emerges from these and previous studies1 2 4 5 8 9 25 is that the distal apoC-III regulatory elements act as a general modular enhancer that can potentiate the strength of proximal promoters that contain one or more HREs. In the case of the apoA-I promoter, this potentiation involves synergistic interactions between HNF-4 or other nuclear hormone receptors that bind to proximal and distal sites and is facilitated by additional factors that bind to promoter and enhancer regions. Similar to other systems, it is assumed that HNF-4 or other nuclear hormone receptors that bind to the proximal and distal HREs and the SP1 and the other factors that bind to the apoC-III enhancer and the proximal promoter elements form a stereospecific DNA-protein complex.32 33 This complex may interact directly or indirectly through TAFs with the factors of basal transcription complex, thus leading to the transcriptional activation of the target gene (Fig 6C⇑). Recent studies have established that multiple interactions of transcription factors with different TAFs and the proteins of the TFIID complex are responsible for the synergistic activation of the target genes.34 35 Mutations in the enhancer or the proximal promoter that prevent the binding of one or more of the participating factors may affect the configuration of this complex and thus explain the reduction in the strength of the promoter/enhancer cluster. The AdML promoter data also indicate that the types of the nuclear hormone receptors that can occupy the HRE of the proximal apoA-I promoter and the distal apoC-III enhancer may lead to either transcriptional synergism or antagonism.
It is possible that SP1, which binds to elements F, H, and I, may act as an architectural component and serve to facilitate interactions among molecules of nuclear hormone receptors that bind to the proximal and distal sites. These interactions may be favorable or unfavorable, resulting in transcriptional synergism or transcriptional repression. Combinatorial interactions among factors have been described in other enhancers, including the T-cell receptor α-gene enhancer36 37 38 and the virus-induced human interferon beta enhancer.33 39 40 In the case of the T-cell receptor α enhancer, binding of lymphoid enhancer factor-1 promotes interactions among the other factors that bind to the enhancer. In the case of the interferon beta enhancer, binding of high-mobility group 1Y [HMG1(Y)] protein increases the binding affinity as well as the interactions among the factors nuclear factor-κB, activating transcription factor-2, and interferon regulatory factor-1, which also bind to the enhancer.33 39 40
Role of Distal ApoC-III Regulatory Elements as Intestinal Enhancer
An interesting question that requires further clarification is whether the distal apoC-III regulatory elements F to J, in addition to increasing the strength of the promoters of the apoA-I/apoC-III /apoA-IV gene cluster, as well as of other heterologous promoters, also suffice to confer tissue-specific expression of one or more of these genes in vivo. To this point transgenic experiments suggest that intestinal expression of the apoA-IV gene requires extension of the apoC-III promoter region to nucleotide 1400.11 This region contains the apoC-III enhancer.1 4 5 Similarly, the proximal apoC-III promoter extending to nucleotide −200 is sufficient to direct hepatic expression of apoC-III gene in vivo.10 Intestinal expression requires a 6-kb intergenic sequence between apoC-III and apoA-IV genes. A recent study has shown that the proximal apoA-I promoter extending to nucleotide −300 linked to the regulatory elements F to J of apoC-III could direct intestinal expression in villus-associated enterocytes along the duodenal to ileal axis but also produced inappropriate expression in cryptoendothelial cells as well as in subpopulations of enteroendocrine cells.8 It has been pointed out that other sequences are required for the correct expression of the human apoA-I gene in the intestine.8 Previous studies have also shown that correct tissue-specific expression of the apoE and apoC-I genes is altered when specific promoter sequences are removed.41 It is possible that the sequences required for correct intestinal expression may be located in the regulatory region extending upstream of nucleotides −255. In this regard, footprinting analysis identified nuclear activities present only in CaCo-2 cells that bind to the −523 to −492 and −488 to −467 apoA-I promoter regions.25
Potential Mechanisms of Transcriptional Activation of the ApoA-I/ApoC-III/ApoA-IV Gene Cluster
Currently, tissue culture experiments (including the present study as well as transgenic mouse experiments from different laboratories) have provided the following information pertinent to the expression of the apoA-I/apoC-III/apoA-IV gene cluster illustrated in Fig 7⇑.1 2 4 5 8 9 25
The hepatic and intestinal transcription of the apoA-I gene is controlled by synergistic interactions between nuclear hormone receptors that bind to elements D and B of the proximal promoter and element I of the apoC-III enhancer.2 6 8 9 Optimal transcription requires SP1, which binds to the apoC-III enhancer, and other factors that bind to the proximal promoter and enhancer region. Other factors bound to distal apoA-I regulatory elements may contribute to correct intestinal expression and may further increase the promoter strength. Availability of HNF-4, which can bind to the regulatory elements AIB, AID, and C-III-I and possibly other upstream sites, may control the transcription of the apoA-I gene in the intestine.25
The hepatic transcription of the apoC-III gene is controlled by synergistic interactions between nuclear hormone receptors that bind to apoC-III regulatory elements B and I, orphan nuclear receptors or related activities that bind to element G and SP1, and other factors that bind the apoC-III enhancer.1 5 10 The intestinal expression is controlled by HNF-4 or other nuclear hormone receptors that bind to elements B and I, SP1, other factors that bind to the enhancer, and possibly additional factors that bind to distal regulatory sites.1 5 10
Finally, the intestinal and hepatic expression of the apoA-IV gene is controlled by synergistic interactions between HNF-4 or other nuclear hormone receptors that bind to element C of the proximal promoter, element I of the apoC-III enhancer, SP1, and other factors that bind to the apoC-III enhancer4 11 or possibly to other distal regulatory sites.4 11
The present and previous in vivo and in vitro data clearly indicate that the distal apoC-III regulatory elements F to J are sufficient to enhance the strength of homologous and heterologous promoters and can direct expression in villus-associated enterocytes as well as other intestinal cell populations.8 Further in vivo studies are required to assess the role of this region as a specific intestinal element.
Selected Abbreviations and Acronyms
|AdML||=||adenovirus major late promoter|
|ARP-1||=||apolipoprotein A-I regulatory protein-1|
|ATF-2||=||activating transcription factor-2|
|C/EBP||=||CCAAT/enhancer binding protein|
|HNF-4||=||hepatic nuclear factor-4|
|HRE||=||hormone response element|
|IRF-1||=||interferon regulatory factor-1|
|RXRα||=||retinoic X receptor-α|
|T3Rβ||=||thyroid hormone receptor-β|
|TAFs||=||TATA box binding protein associate factors|
This study was supported by European Community (EC) grant BIOT-CT93-0473, by a grant by the Greek Ministry of Health, by National Institutes of Health grant HL-33952, and AHA Grant-in-Aid 93012420. We thank Dina Barda, Maria Laccotripe, and Sophia Lavrentiadou for excellent technical assistance, Dr Savvas Makrides for helpful comments on the manuscript, and Anne Plunkett for typing the manuscript.
Ogami K, Hadzopoulou-Cladaras M, Cladaras C, Zannis VI. Promoter elements required for hepatic and intestinal transcription of the human apoCIII gene. J Biol Chem. 1990;265:9808-9815.
Papazafiri P, Ogami K, Ramji DP, Nicosia A, Monaci P, Cladaras C, Zannis VI. Promoter elements and factors involved in hepatic transcription of the human apoA-I gene: positive and negative regulators bind to overlapping sites. J Biol Chem. 1991;266:5790-5797.
Widom RL, Ladias JAA, Kouidou S, Karathanasis SK. Synergistic interactions between transcription factors control expression of the apolipoprotein A-I gene in liver cells. Mol Cell Biol. 1991;11:677-687.
Ktistaki E, Lacorte JM, Katrakili N, Zannis VI, Talianidis I. Transcriptional regulation of the apolipoprotein A-IV gene involves synergism between a proximal hormone response element and a distant enhancer located in the upstream promoter region of the apolipoprotein CIII gene. Nucl Acids Res. 1994;22:4689-4696.
Talianidis I, Tambakaki A, Toursounova J, Zannis VI. Complex interactions between SP1 bound to multiple distal regulatory sites and HNF-4 bound to the proximal promoter leads to the transcriptional activation of the liver-specific human apoCIII gene. Biochemistry. 1995;34:10298-10309.
Zannis VI, Tzameli I. Binding specificity and modulation of the apoA-I promoter activity by homo- and heterodimers of nuclear receptors. J Biol Chem. 1996;271:8402-8415.
Cladaras C, Ogami K, Kardassis D, Hadzopoulou-Cladaras M, Zannis VI. Recognition of proximal human apoCIII promoter by NFY* type, C/EBP and a family of heat stable activities. Circulation. 1991;84(suppl II):II-109. Abstract.
Bisaha JG, Simon TC, Gordon JL, Breslow JL. Characterization of an enhancer element in the human apolipoprotein CIII gene that regulates human apolipoprotein A-I gene expression in the intestinal epithelium. J Biol Chem. 1995;270:19979-19988.
Walsh A, Azrolan N, Wang K, Marcigliano A, O'Connell A, Breslow JL. Intestinal expression of the human apoA-I gene is controlled by a DNA region 3′ to the gene in the promoter of the adjacent convergently transcribed apoCIII gene. J Lipid Res. 1993;34:617-623.
de Silva HV, Lauer SJ, Wang J, Simonet WS, Weisgraber KH, Mahley RW, Taylor JM. Overexpression of human apolipoprotein CIII in transgenic mice results in an accumulation of apolipoprotein B48 remnants that is corrected by excess apolipoprotein E. J Biol Chem. 1994;269:2324-2335.
Lauer SJ, Simonet WS, Bucay N, de Silva HV, Taylor JM. Tissue-specific expression of the human apolipoprotein A-IV gene in transgenic mice. Circulation. 1991;84:1390.
Castelli WP, Doyle JT, Gordon T, Haames CG, Hjortland MC, Hulley SB, Kagan A, Zukel WJ. HDL cholesterol and other lipids in coronary artery disease: the Cooperative Lipoprotein Phenotyping Study. Circulation. 1977;55:767-772.
Heiss G, Tyroler HA. Proceedings of a Workshop on Apolipoprotein Quantification. Bethesda, Md: National Institutes of Health; 1982:7-24. US Dept of Health and Human Services publication NIH 83-1266.
Ito Y, Azrolan N, O'Connell A, Walsh A, Breslow JL. Hypertriglyceridemia as a result of human apoCIII gene expression in transgenic mice. Science. 1990;249:790-793.
Kardassis D, Zannis VI, Cladaras C. Organization of the regulatory elements and nuclear activities participating in the transcriptional activation of the human apoB gene. J Biol Chem. 1991;267:2622-2632.
Ladias JAA, Hadzopoulou-Cladaras M, Kardassis D, Cardot P, Cheng J, Zannis VI, Cladaras C. Transcriptional regulation of human apolipoprotein genes apoB, apoCIII and apoA-II by members of the steroid hormone receptor superfamily HNF-4, ARP-1, EAR-2 and EAR-3. J Biol Chem. 1992;267:15849-15860.
Edlund T, Walker MD, Barr PJ, Rutter WJ. Cell-specific expression of the rat insulin gene: evidence for role of two distinct 5′ flanking elements. Science. 1985;230:912-916.
Gorman CM, Moffat LF, Howard BH. Recombinant genomes which express chloramphenicol acetyltransferase in mammalian cells. Mol Cell Biol. 1982;2:1044-1051.
Lavrentiadou S, Hadzopoulou-Cladaras M, Ladias J, Cladaras C, Zannis VI. Modulation of the apoCIII promoter activity by heterodimers of ligand dependent nuclear receptors RXRα-RARα, RXRα-T3Rβ and RXRα-PPARα. Circulation. 1995;92:291.
Karathanasis SK. Apolipoprotein multigene family: tandem organization of human apolipoprotein A-I, CIII and A-IV genes. Proc Natl Acad Sci U S A. 1985;82:6374-6378.
Ginsburg GS, Ozer J, Karathanasis SK. Intestinal apolipoprotein A-I gene transcription is regulated by multiple distinct DNA elements and is synergistically activated by the orphan nuclear receptor, hepatocyte nuclear factor 4. J Clin Invest. 1995;96:528-538.
Mangelsdorf DJ, Umesono K, Evans RM. The retinoid receptors. In: The Retinoids: Biology, Chemistry and Medicine. 2nd ed. New York, NY: Raven Press; 1994;8:319-349.
Reisher SR, Hughes TE, Ordovas JM, Schaefer EJ, Feinstein SI. Increased expressing of apolipoprotein genes accompanies differentiation in the intestinal cell line CaCo-2. Proc Natl Acad Sci U S A. 1993;90:5757-5761.
Widom RL, Rhee M, Karathanasis SK. Repression by ARP-1 sensitizes apolipoprotein A-I gene responsiveness to RXRα and retinoic acid. Mol Cell Biol.. 1992;12:3380-3389.
Zhang X-K, Salbert G, Lee M-O, Pfahl M. Mutations that alter ligand-induced switches and dimerization activities of the retinoic X receptor. Cell Biol. 1994;14:4311-4323.
Leng X, Blanco J, Tsai SY, Ozato K, O'Malley BW, Tsai M-J. Mechanisms for synergistic activation of thyroid hormone receptor and retinoic X receptor on different response elements. J Biol Chem. 1994;269:31436-31442.
Sauer F, Hansen SK, Tjian R. DNA template and activator-coactivator requirements for transcriptional synergism by Drosophila bicoid. Science. 1995;270:1825-1828.
Sayer F, Hansen SK, Tjian T. Multiple TAFIIs directing synergistic activation of transcription. Science. 1995;270:1783-1788.
Carlsson P, Waterman ML, Jones KA. The hLEF/TCF-1α HMG protein contains a context-dependent transcriptional activation domain that induces the TCRα enhancer in T cells. Genes Dev. 1993;7:2418-2430.
Simonet WS, Bucay N, Pitas RE, Lauer SJ, Taylor JM. Multiple tissue-specific elements control the apolipoprotein E/C-I gene locus in transgenic mice. J Biol Chem. 1991;266:8651-8654.