Transactivation of the Human Apolipoprotein CII Promoter by Orphan and Ligand-dependent Nuclear Receptors

The regulatory elements CIIC (−159/−116) and CIIB (−102/−81) of the apolipoprotein CII (apoCII) promoter have distinct specificities for orphan nuclear receptors (Vorgia, P., Zannis, V. I., and Kardassis, D. (1998) J. Biol. Chem. 273, 4188–4199). In this communication we investigated the contribution of ligand-dependent and orphan nuclear receptors on the transcriptional regulation of the humanapoCII gene. It was found that element CIIC in addition to ARP-1 and EAR-2 binds RXRα/T3Rβ heterodimers strongly, whereas element CIIB binds hepatic nuclear factor 4 (HNF-4) exclusively. Binding is abolished by mutations that alter the HRE binding motifs. Transient cotransfection experiments showed that in the presence of T3, RXRα/T3Rβ heterodimers transactivated the −205/+18 apoCII promoter 1.6- and 11-fold in HepG2 and COS-1 respectively. No transactivation was observed in the presence of 9-cis-retinoic acid. Transactivation requires the regulatory element CIIC, suggesting that this element contains a thyroid hormone response element. HNF-4 did not affect the apoCII promoter activity in HepG2 cells. However, mutations in the HNF-4 binding site on element CIIB and inhibition of HNF-4 synthesis in HepG2 cells by antisense HNF-4 constructs decreased the apoCII promoter activity to 25–40% of the control, indicating that HNF-4 is a positive regulator of the apoCII gene. ARP-1 repressed the −205/+18 but not the −104/+18 apoCII promoter activity in HepG2 cells, indicating that the repression depends on the regulatory element CIIC. In contrast, combination of ARP-1 and HNF-4 transactivated different apoCII promoter segments as well as a minimal adenovirus major late promoter driven by the regulatory element CIIB. Mutagenesis or deletion of elements CIIB or CIIC established that the observed transactivation requires DNA binding of one of the two factors and may result from HNF-4-ARP-1 interactions that elicit the transactivation functions of HNF-4. The combined data indicate that RXRα/T3Rβ in the presence of T3 and HNF-4 can upregulate the apoCII promoter activity by binding to the regulatory elements CIIC and CIIB, respectively. In addition, ARP-1 can either have inhibitory or stimulatory effects on the apoCII promoter activity via different mechanisms.

Orphan nuclear receptors as well as receptors for retinoids and thyroids are members of a nuclear receptor superfamily that controls diverse biological functions including growth, development, and homeostasis (11)(12)(13)(14)(15). They recognize specific hexameric AG(G/T)TCA motifs with variations in sequence, spacing, and orientation, designated hormone response elements (HREs) (16 -20). In the current study, we demonstrate that in the presence of T3, RXR␣/T3R␤ heterodimers bind to a thyroid hormone response element (TRE) present in element CIIC and transactivate the human apoCII promoter. Binding of ARP-1 to the same site repressed the promoter activity. Furthermore, antisense methodologies and promoter mutagenesis established that HNF-4 is a positive activator required for optimal activity of the apoCII promoter in HepG2 cells. Finally, combination of ARP-1 and HNF-4 superactivate the apoCII promoter via novel mechanisms that may require interaction of the two factors on the apoCII promoter.

EXPERIMENTAL PROCEDURES
Materials-The sources of materials utilized have been described (10).
Plasmid Constructions-The construction of the apoCII promoter plasmids Ϫ545/ϩ18 CII CAT, Ϫ388/ϩ18 CII CAT, Ϫ388/ϩ18 CII Bmut CAT, Ϫ388/ϩ18 CII Cmut CAT, Ϫ388/ϩ18 CII C/B mut CAT, Ϫ205/ ϩ18 CII CAT, Ϫ104/ϩ18 CII CAT as well as the pUCSHCAT vector have been described previously (10,21). Plasmids pMT2, pMT2-* This work was supported by Greek General Secretariat for Science and Technology Grants PENE 1663 and EPET 646 and Greek Ministry of Health Grant KESY E396. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
For the construction of the CIIC-AdML CAT plasmid, a doublestranded synthetic oligonucleotide corresponding to the apoCII regulatory element C (CIIC) flanked by HindIII restriction sites was phosphorylated by T4 polynucleotide kinase and cloned once into the HindIII site of vector AdML-44 CAT. Similarly, for the construction of the CIIB AdML CAT plasmid, a double-stranded synthetic oligonucleotide that contained the apoCII regulatory element CIIB flanked by HindIII and XbaI restriction sites at its 5Ј and 3Ј ends, respectively, was cloned into the corresponding sites of AdML-44 CAT vector. Positive clones were verified by dideoxy DNA sequencing using the sequenase kit from Amersham/U. S. Biochemical Corp.
Plasmid pCDNA1-neo anti-HNF-4 ribozyme was constructed as follows. The human HNF-4 cDNA between position 402 and 1215 was amplified by polymerase chain reaction using as 5Ј and 3Ј primers the synthetic oligonucleotides HNF-4 RI and HNF-4 Xh respectively (see Table I). Primer HNF-4 Xh contains, in addition to HNF-4 sequence in the 3Ј to 5Ј orientation, part of the catalytic domain of hammerhead ribozymes (24). The amplified product was digested with EcoRI and XhoI and subcloned into the corresponding sites of the vector pBSRZ 12/1 (25). Vector pBSRZ 12/1 contains, in addition to the remaining catalytic domain of the hammerhead ribozymes, the trinucleotide ATC, which is required for the formation of helix I. The anti-HNF-4 ribozyme was subsequently cloned at the BamHI/HindIII sites of the eukaryotic expression vector pcDNA-1-neo (Invitrogen, Inc.). To construct plasmidneo pcDNA1 anti-HNF-4, the HNF-4 and half of the catalytic domain of the ribozyme was subcloned into the BamHI/Xho sites of vector pCDNA1-neo.
Transfections and CAT Assays-Maintenance and conditions of transfections of human hepatoma (HepG2) and monkey kidney COS-1 have been described (10,26). In the hormone induction experiments, the hormone (10 Ϫ7 M T3 or 10 Ϫ6 M 9-cis-retinoic acid) was added to the cells 18 h after transfections, and cells were cultivated for 22 h in Dulbecco's modified Eagle's medium supplemented with 5% charcoal stripped and delipidated fetal bovine serum. For the generation of permanent cell lines expressing antisense HNF-4, HepG2 cells were seeded at a density of 5 ϫ 10 5 cells/60-mm diameter dishes, the day before transfection. The cells were transfected with 17 g of pcDNA-neo anti-HNF-4 ribozyme, pCDNA-neo anti-HNF-4 or the empty pCDNA-1-neo vector by the Ca 3 (PO 4 ) 2 precipitation method. Eighteen h after transfection, cells were trypsinized, diluted 1/20 and 1/40 and replated in 100-mm diameter dishes in a Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 500 g/ml Geneticin (Life Technologies, Inc.). The concentration of Geneticin in the cell culture medium was increased to 1000 g/ml 2 weeks later to minimize the possibility of selection of false positive clones. Resistant clones were selected using a cloning cylinder were grown in Dulbecco's modified Eagle's medium ϩ 10% fetal bovine serum containing 1000 g/ml Geneticin, characterized, and stored in liquid nitrogen. Clones expressing the antisense HNF-4 were utilized in transfection experiments in parallel with control HepG2 cultures.
Gel Electrophoretic Mobility Shift Assays-Assays using rat liver nuclear extracts or COS-1 whole cell extracts were performed as described (21,27).

Heterodimers of RXR␣/T3R␤ Bind to a TRE on Element CIIC and Transactivate the Human apoCII Promoter in HepG2
and COS-1 Cells-DNA binding experiments were performed using oligonucleotide CIIC as a probe (Table I) and extracts from COS-1 cells expressing RXR␣, RAR␣, T3R␤, and PPAR␣. This analysis showed that element CIIC strongly binds RXR␣/ T3R␤, less efficiently binds T3R␤, and weakly binds RXR␣/ RAR␣ and RXR␣/PPAR␣ heterodimers ( Fig. 1A-D). Element CIIC contains two direct repeats 5Ј ACGTCC(CCCA)AGGTCA 3Ј in the noncoding strand between nucleotides Ϫ140 to Ϫ155 separated by four spacer nucleotides (included in parentheses) (Fig. 1A). Mutations in both half repeats of element CIIC abolished the binding of orphan and ligand-dependent nuclear receptors to this site ( Fig. 1A-D). Cotransfection experiments in HepG2 cells using the Ϫ205/ϩ18 apoCII promoter mutated in element CIIC showed that this mutation reduced the apoCII promoter activity to approximately 40% of the control value ( Fig. 2A).
Similar cotransfection experiments in HepG2 cells showed that the activity of the -205/ϩ18 apoCII promoter was repressed by approximately 50% by RXR␣/T3R␤ heterodimers in the absence of ligand ( Fig. 2A). The promoter activity increased 3-fold by the addition of 10 Ϫ7 M T3 and was 1.6-fold higher than the activity observed in the absence of both RXR␣/T3R␤ heterodimers and T3 ( Fig. 2A). T3R␤ alone in the presence or absence of T3 or RXR␣/T3R␤ heterodimers in the presence of 9-cis-retinoic acid did not alter the apoCII promoter activity ( Fig. 2A). Mutations in the apoCII promoter, which prevented the binding of hormone nuclear receptors to this site, abolished the T3-dependent transactivation of the mutant promoter ( Fig.  2A). In the presence of T3, RXR␣/T3R␤ heterodimers also transactivated a synthetic promoter containing a single copy of the element CIIC fused to the minimal AdML promoter 3-fold (Fig. 2B). A synthetic promoter under the control of a mutated version of the TRE found in the growth hormone promoter was transactivated 24.5-fold (Fig. 2B). This promoter is directed by an ideal DR4 direct repeat AGGTCA(GATC)AGGTCA motif and could be transactivated 21-fold by RXR␣/T3R␤ heterodimers in the presence of T3 (28).
The Ϫ205/ϩ18 apoCII promoter was also transactivated 4-fold by RXR␣/T3R␤ heterodimers in the presence of T3 in COS-1 cells (Fig. 2C). The promoter activity was unaffected by RXR␣/T3R␤ heterodimers in the absence of T3 and by T3R␤ in the presence of T3. T3R␤ in the absence of T3 or RXR␣/T3R␤ heterodimers in the presence of 9-cis-retinoic acid reduced the apoCII promoter activity in COS-1 cells by approximately 50%. The combined data of Figs. 1A-D and 2A-C indicate that the element CIIC is a functional TRE that confers T3-dependent transactivation of the apoCII promoter by RXR␣/T3R␤ heterodimers.
HNF-4 Binds Exclusively to the Regulatory Element CIIB and Is Required for Optimal Activity of the apoCII Promoter in Cells of Hepatic Origin-A previous study showed that the regulatory element CIIB binds HNF-4 but not ARP-1 or EAR-2 (10). This element contains a direct AAGTCCTGGCCA repeat between nucleotides Ϫ87 to Ϫ98 of the noncoding strand without spacer nucleotides between the two half repeats (DR0). Mutations within the two half repeats (Fig. 1A) abolish the binding of HNF-4 to this site (Fig. 3A). DNA binding experiments using the oligonucleotide CIIB as probe (Table I) and extracts from COS-1 cells expressing different hormone nuclear receptors showed that homodimers of RXR␣ and heterodimers of RXR␣ with RAR, T3R␤, and PPAR␣ do not bind to this element, indicating that the HRE of element CIIB is an exclusive HNF-4 binding site (Fig. 3A).
We have shown recently that HNF-4 does not increase the apoCII promoter strength in HepG2 cells (10). On the other hand, mutations in element CIIB that abolish the binding of HNF-4 to this site reduced the apoCII promoter activity by 60% of the control (Fig. 3B). To assess the role and the importance of HNF-4 for the function of the apoCII promoter, we generated  (7)(8)(9). Double arrows show the intergenic distances in kilobases. The lower part of the panel shows the position of the footprinted regions CIIA to CIIE as well as the mutations within the HREs of elements CIIB and CIIC, which affect the promoter strength and the binding of nuclear receptors. Panels B-D. DNA binding gel electrophoretic mobility shift assays using wild-type and mutated oligonucleotides CIIC (Ϫ151/Ϫ116) as probes and rat liver nuclear extracts or extracts of COS-1 cells expressing ARP-1, EAR-2, RXR␣, RAR␣, T3R␤, and PPAR␣. A double-stranded oligonucleotide corresponding to the wild-type and mutated apoCII footprint CIIC shown in Table I  cell lines expressing two antisense HNF-4 constructs or the empty vector. One of the constructs expresses the 402 to 1215 HNF-4 antisense sequence, whereas the other expresses the same sequence fused to the catalytic domain of the hammerhead ribozyme (29). Cotransfection experiments using the parental HepG2 and the stable HepG2 cell lines expressing the antisense constructs showed that the activity of the Ϫ545/ϩ18 promoter decreased by approximately 60 -75% of the control in HepG2 cells expressing the two antisense HNF-4 sequences. The activity of the Rous sarcoma virus promoter, which does not contain the HNF-4 binding site, was not affected. The activity of the promoter in HepG2 cell lines expressing the antisense constructs could be restored to the same level as the control by cotransfection with an HNF-4 expression vector (data not shown). The findings indicate that HNF-4 is an important regulator of the apoCII promoter activity and that its concentration is limited in the cell lines expressing the antisense HNF-4 constructs.

ARP-1 Can Repress the apoCII Promoter Activity by Binding to the Regulatory Element CIIC. Combination of ARP-1 and HNF-4 Superactivates the apoCII Promoter-Cotransfection of
HepG2 cells with the Ϫ205/ϩ18 apoCII promoter CAT construct along with an ARP-1 expression vector repressed the apoCII promoter activity by 50%. Repression could be reversed by cotransfection with HNF-4 (Fig. 4A). This result indicated that ARP-1 plays a negative role in apoCII gene regulation. ARP-1 was unable to repress the activity of the -104/ϩ18 apoCII promoter, which lacks element CIIC in HepG2 cells, indicating that repression depends on the presence of the regulatory element CIIC (Fig. 4B). Unexpectedly, cotransfection of HepG2 cells with ARP-1 and HNF-4 transactivated the Ϫ205/ ϩ18 or the Ϫ104/ϩ18 apoCII promoter 4.7-and 2.2-fold respectively, despite the fact that the activity of these promoters in HepG2 cells is not affected by HNF-4 (10) (Fig. 4, A and B). In addition, the combination of ARP-1 and HNF-4 transactivated the synthetic CIIB-AdML promoter, which contains a single copy of element CIIB, whereas HNF-4 had no effect, and ARP-1 increased 2-fold the activity of this promoter (Fig. 4C). Transactivation occurs despite the fact that ARP-1 cannot bind to either the Ϫ104/ϩ18 apoCII promoter or the CIIB AdML promoter. The findings suggest that the observed transactivation may be the result of direct protein-protein interactions, which increase the transactivation potential of one or both of these factors. The possibility that ARP-1 squelches negative regulators is less likely since element CIIB is an exclusive HNF-4 binding site. In addition, interactions of negative regulators such as ARP-1 with components of the basal transcription complex are expected to exert negative rather than positive effects on transcription (30). The positive effect of ARP-1 on the apoCII promoter activity was further demonstrated by cotransfection experiments in COS-1 cells that lack or contain very low amounts of endogenous HNF-4 and ARP-1. This analysis showed that HNF-4 transactivated the -104/ϩ18 apoCII promoter 4-fold. The same promoter, in the presence of increasing concentrations of ARP-1, was transactivated up to 15-fold (Fig. 4D). Transactivation of the Ϫ205/ϩ18 apoCII promoter by the combination of ARP-1 and HNF-4 also occurred when the regulatory element CIIB, which is the binding site of HNF-4, was mutated (Fig. 4E). The combined data of Fig. 4, D and E indicate putative HNF-4-ARP-1 interactions on the apoCII promoter when either one of the two factors is bound to the DNA. A schematic representation of the putative mechanisms of activation or repression of the apoCII promoter activity by HNF-4, ARP-1, and RXR␣/T3R␤ heterodimers is shown in Fig. 5.

DISCUSSION
The Regulatory Element CIIC Is a TRE-Thyroid hormone receptors recognize specific hexameric half repeat motifs AG(G/ T)TCA, preferably separated by four spacer nucleotides (16 -20). The regulatory element CIIC, which is recognized by ARP-1 and EAR-2, contains two direct repeats with DR4 spacing on the noncoding strand between nucleotides Ϫ140 to Ϫ155. Previous studies have suggested that this motif is the preferred binding site of RXR␣/T3R␤ heterodimers (28). Indeed, the DNA binding data confirmed that T3R␤/RXR␣ heterodimers bind very strongly to this element. T3R␤ binds less efficiently, and homodimers of RXR␣ or heterodimers of RXR␣ with RAR␣ and PPAR bind weakly, thus establishing that element CIIC contains a TRE.
The functionality of this TRE was established with cotransfection experiments involving homologous and heterologous promoters. Ligand-dependent transactivation by RXR␣/T3R␤ heterodimers was achieved with the wild-type -205/ϩ18 apo-CII promoter as well as with a synthetic AdML promoter under the control of the wild-type CIIC TRE. Mutations in the TRE that prevented the binding of the nuclear receptors to this site abolished the transactivation. The transactivation achieved by the TRE in the context of the Ϫ205/ϩ18 apoCII promoter and the minimal AdML promoter in HepG2 cells was 1.7-and 3-fold respectively. This 3-fold transactivation of the synthetic promoter by the CIIC TRE is comparable with the 3.6-fold transactivation by RXR␣/T3R␤ heterodimers ϩ T3 of a synthetic promoter under the control of the growth hormone TRE. Similar to the CIIC element, the growth hormone TRE contains a direct AGGTAA(GATC)AGGGAC repeat with DR4 spacing (28). Much greater promoter transactivations can be achieved by an ideal version of this element, which contains (AGGT(A/ C)A) 2 DR4 repeats (28). In contrast, in the absence of T3, the apoCII promoter activity was repressed. The 60% repression in apoCII promoter activity by the RXR␣/T3R␤ heterodimer in the absence of T3 is in agreement with the recently proposed model for their mode of action. According to the proposed model, in the absence of ligand, the RXR/T3R heterodimers bind transcriptional co-repressors such as nuclear repressor corepressor Sin3 and N-CoR, which results in histone deacetylation and condensation of chromatin (32,33). This leads to a repression of transcription. Binding of ligand to the heterodimer results in the displacement of the repressors by activators containing acetylase activity such as p300/CREB binding protein and P/CREB binding protein-associated factor. This leads to acetylation of histones, activation of chromatin, and gene transcrip- tion (32,33). HREs with different spacing between the half repeats (DR0, DR1, DR2) are also found in the proximal regions of other apolipoprotein promoters (23,34). In contrast to apo-CII, the promoter of apoA-I is repressed by RXR␣/T3R␤ heterodimers in the presence of T3 (23,34). Limited in vivo studies showed that hyperthyroidism increased the ratio of apoCII to apoCIII in the VLDL fraction as well as the activity of postheparin lipoprotein lipase activity, which is activated by apoCII (35). Clearly, more rigorous in vivo studies are needed to establish how the thyroid status affects the apoCII gene expression.
HNF-4 Is an Important Activator of the apoCII Promoter-The DNA binding data of this and a previous study (10) established that the HRE present in element CIIB recognizes HNF-4 exclusively but does not recognize other orphan or ligand-dependent nuclear receptors that bind to the regulatory element CIIC (Fig. 3A). The importance of HNF-4 for the function of the apoCII promoter could not be assessed by cotransfection experiments, possibly due to the saturating amounts of HNF-4 in HepG2 cells (10). In the current study, utilization of antisense methodologies very convincingly established that HNF-4 is an important activator of the apoCII promoter. It is interesting that both the mutagenesis of the DNA recognition motif of HNF-4 as well as the antisense HNF-4 constructs reduced the apoCII promoter activity by approximately 60 to 75% of the control. The finding suggests that HNF-4 contributes to optimal promoter strength, but other factors, such as CCAAT enhancer binding protein C/EBP or related activities that bind to element CIID, may account for the remaining 25 to 40% of the promoter activity in the absence of HNF-4. The fact that HNF-4 is an activator of the apoCII promoter strength is also supported by cotransfection experiments in COS-1 cells, where the promoter is transactivated 9-fold in the presence of HNF-4 (10). The preservation of partial promoter activity in the absence of HNF-4 also differentiates the apoCII promoter from the promoters of the apoA-I, CIII, and AIV gene cluster, where intact HREs are essential for promoter activity (23,34).
ARP-1 May either Repress or Transactivate the apoCII Promoter Activity. Repression Requires Binding of ARP-1 to the TRE of Element CIIC. Transactivation Requires the Presence of HNF-4 and Is Independent of DNA Binding-The regulatory element CIIC that contains the TRE is required for optimal promoter activity, since mutagenesis of the TRE that abolishes binding of hormone nuclear receptors to this site reduces the apoCII promoter activity to approximately 40% of the control ( Fig. 2A). On the other hand, this element is the binding site of orphan nuclear receptors ARP-1 and EAR-2. Previous studies have shown that ARP-1 and EAR-2 can usually but not always (36,37) repress the promoter activity of other genes by competing for binding to the same HRE (38 -43). In the case of RXR␣ and T3R␤, transrepression may also be involved (30). The DNA binding data, in combination with the transactivation data of this study, established that the transcriptional repression is caused by the binding of ARP-1 to the regulatory element CIIC, which in addition to orphan nuclear receptors, strongly binds RXR␣/T3R␤ heterodimers. In the absence of this element, the activity of the apoCII promoter is not affected by ARP-1.
It has been proposed that transcription factors bound to distal regulatory elements form a stereospecific DNA protein complex (44). This complex may interact directly or indirectly through the TATA box-binding protein-associated factors or transcriptional mediators/intermediary factors (45)(46)(47)(48) with the factors of the basal transcription complex, thus leading to the transcriptional activation or repression of the target gene (Figs. 4 and 5). In the case of ARP-1, it has been suggested that its interactions with the basal transcription factor TFIIB may freeze the pre-initiation complex in an inactive configuration and may account for the repressor activity of this factor (30). The repression of the Ϫ205/ϩ18 apoCII promoter activity is reversed in the presence of excess HNF-4. Since ARP-1 and HNF-4 bind on distinct HREs on elements CIIC and CIIB, respectively ( Figs. 1 and 3), reversion of the repression could be the result of favorable protein-protein interactions involving the two factors. Additional indirect evidence of this type of interaction was provided by the observation that the Ϫ104/ϩ18 apoCII promoter, which lacks element CIIC, or a synthetic promoter, which contains a single copy of element CIIB, is greatly transactivated by combination of ARP-1 and HNF-4. Since these promoters cannot bind ARP-1 and ARP-1 cannot form heterodimers with HNF-4 (49), the observed potentiation of the transcriptional activity of HNF-4 could result from transient interactions of HNF-4 with ARP-1 on the apoCII promoter. Evidence in support of this type of interaction was obtained in recent studies involving the HNF-1 promoter. This promoter, which contains an exclusive HNF-4 binding site, could likewise be transactivated by combination of ARP-1 and HNF-4 (49). Direct interactions between chicken ovalbumin upstream promoter transcription factors I and II (EAR-3 and ARP-1) were demonstrated by in vitro protein-protein interaction experiments involving glutathione S-transferase fusion proteins (49). In addition, cell culture studies showed that truncated HNF-4 forms that lack the nuclear localization domain can be transported into the nucleus of COS-1 cells by cotransfection with ARP-1 or EAR-3 (49). The direct proteinprotein interactions between these two factors involve residues 227 to 271 of HNF-4, and the formation of a functionally active complex between these two factors requires an intact activation domain (residues 130 -368) of HNF-4 (49). Deletion of residues 354 -368 of HNF-4 located within a region homologous to the activation function-2 domain found in other hormone receptors (50 -52) abolished the synergistic activation of the target promoter by HNF-4 and ARP-1 or EAR-3 (49). Earlier studies established that the transcriptional activity of hormone nuclear receptors is modulated by direct protein-protein interactions involving transcription intermediary factors, which may act either as activators or repressors (45,(51)(52)(53)(54)(55)(56). It has been proposed that association of ARP-1 or EAR-3 with HNF-4 bound to its cognate site on the promoters alters the conformation or the activation function-2 of HNF-4 or it may facilitate the formation of the pre-initiation complex (49).
The present study as well as previous studies have shown convincingly that HNF-4 is a positive transcriptional activator of a number of liver specific genes (34,38,43). HNF-4 can also synergize with a variety of other transcription factors such as C/enhancer binding protein (57), c-AMP response elementbinding protein (58), and HNF-1 (39) bound to their target sites. The current study shows that in promoters such as the apoCII, which contain exclusive ARP-1 binding sites, HNF-4 may also play the role of a transcriptional mediator. This HNF-4 function is the topic of ongoing research and may likewise involve interactions with ARP-1 dimers bound to DNA. Such interactions could allow utilization of the activation domain of HNF-4 bound to ARP-1 to drive transcription. Overall, the current study demonstrates that orphan as well as liganddependent nuclear receptors can modulate the apoCII gene transcription positively or negatively via different mechanisms. Thus different combinations of hormone nuclear receptors in hepatic cells and the availability of the ligands may affect the overall apoCII synthesis and plasma levels of apoCII and thus affect the catabolism of triglyceride-rich lipoproteins.