Opposite Regulation of the Human Apolipoprotein M Gene by Hepatocyte Nuclear Factor 1 and Jun Transcription Factors*

HDL is a negative risk factor for atherosclerosis because of its multiple atheroprotective functions. Inflammation converts HDL particles from anti-atherogenic to pro-atherogenic, and this transformation is associated with changes in HDL structure and composition. Apolipoprotein M (apoM) has been recently shown to play a role in the maturation of HDL in plasma and to protect from atherosclerosis. ApoM gene is expressed primarily in the liver and kidney and is down-regulated by pro-inflammatory signals. We now show that the human apoM promoter harbors a dual specificity regulatory element in the proximal region that binds hepatocyte nuclear factor 1 (HNF-1) and members of the AP-1 family of pro-inflammatory transcription factors (c-Jun and JunB). Overexpression of c-Jun or JunB repressed both the basal and the HNF-1-mediated transactivation of the human apoM promoter. Treatment of HepG2 cells with potent inflammation-inducing phorbol esters or overexpression of PKCα was associated with a marked inhibition of apoM gene expression in a c-Jun/JunB-dependent manner. We provide evidence for a novel mechanism of inflammation-induced transcriptional repression that is based on the competition between HNF-1 and Jun proteins for binding to the same regulatory region. A similar mechanism accounts for the down-regulation of the liver-specific apolipoprotein A-II gene by Jun factors. Our studies provide novel insights on the mechanisms that control the expression of liver-specific apolipoprotein genes during inflammation and could affect the maturation and the functionality of HDL particles.

ApoM is a recently identified glycoprotein that belongs to the lipocalin protein superfamily (15)(16)(17)(18). ApoM is secreted primarily by the liver and associates with HDL particles through its retained N-terminal signal peptide (19,20). Lack of apoM expression in hepatocyte nuclear factor 1␣ (HNF-1␣) knockout mice or in mice in which the endogenous apoM gene was silenced via small interfering RNA technology was associated with the loss of pre-␤-HDL particles and the formation of larger HDL particles, suggesting that apoM may play a role in HDL remodeling (21). The role of apoM in atheroprotection is supported by studies showing that adenoviral overexpression of apoM in LDL receptor (Ϫ/Ϫ) mice or overexpression of apoM in the liver of transgenic mice reduced atherosclerosis development (21,22). Furthermore, apoM-containing HDL particles isolated from human plasma or apoM transgenic mice have been shown to be more resistant to oxidation and more efficient in protecting against LDL oxidation as well as stimulating cholesterol efflux from macrophages (22,23).
ApoM is expressed mainly in the liver and the kidney (24). The regulatory mechanisms that control human apoM gene transcription in the above tissues under healthy or disease states are not well understood. In the liver, apoM gene transcription has been shown to be controlled positively by HNF-1␣, liver receptor homolog 1 (LRH-1), and Forkhead box A2 transcription factors that bind to distinct sites on the proximal apoM promoter (25)(26)(27). We have shown recently that specific members of the hormone nuclear receptor superfamily occupy a proximal hormone response element and induce apoM gene transcription in response to retinoic acid, oxysterols, and fibrates, respectively (28). All of the above drugs have been shown to elevate HDL cholesterol levels in human or experimental animals.
ApoM, similar to other apolipoproteins such as apoA-I or apoC-III (29) is a negative acute response protein, and its levels decrease during infection and inflammation. Importantly, inflammation is associated with marked changes in HDL structure and functionality as demonstrated using HDL isolated from plasma of patients with chronic inflammatory diseases (30).
Key mediators of inflammatory responses are members of the AP-1 family of transcription factors. AP-1 proteins are dimeric leucine zipper transcription factors that bind as homoor heterodimers to AP-1-responsive elements (5Ј-TGA(G/ C)TCA-3Ј) on the promoters of many genes (31,32). Major inducers of AP-1 activity are pro-inflammatory cytokines, phorbol esters, and cAMP (31,32). Phorbol esters are potent activators of PKC (33), a family of at least 11 isozymes that are involved in many biological processes in different tissues and cell types (34).
We show here that treatment of HepG2 cells with AP-1inducing agents such as phorbol esters caused a potent downregulation of apoM gene expression via activation of PKC and Jun (c-Jun and JunB) proteins. ApoM gene repression by Jun requires a novel AP-1 element in the proximal apoM promoter that is localized inside a previously characterized binding site for HNF-1␣ (25) and is facilitated by the competition between HNF-1␣ and Jun proteins for binding to the same regulatory site. A similar mechanism accounts for the inhibition of the apolipoprotein A-II gene that also plays a role in HDL metabolism and the pathogenesis of atherosclerosis (35).

EXPERIMENTAL PROCEDURES
Plasmid Constructions-The human apoM promoter constructs (Ϫ950/ϩ42, Ϫ642/ϩ42, Ϫ402/ϩ42, Ϫ241/ϩ42, Ϫ105/ ϩ42, and Ϫ49/ϩ42)hapoM-luc were generated by PCR amplification of the corresponding fragments using human genomic DNA as template and subsequent cloning into the KpnI-HindIII sites of the pGL3basic vector (Promega). The apoM promoter construct (Ϫ20/ϩ8)hapoM-luc was generated by ligation of a double-stranded oligonucleotide corresponding to the Ϫ20/ϩ8 region of the human apoM promoter into the KpnI-HindIII sites of the pGL3basic vector. The (Ϫ950/ ϩ42)mut hapoM-luc, which bears mutations in the AP-1 site, was generated by site-directed mutagenesis using the QuikChange site-directed mutagenesis kit (Stratagene) according to the manufacturer's instructions. The mouse apoM promoter construct (Ϫ761/Ϫ7)mapoM-luc was generated by PCR amplification of the corresponding fragment using mouse genomic DNA as template and subsequent cloning into the KpnI-HindIII sites of the pGL3basic vector. The sh-c-Jun, sh-JunB, sh-HNF-1␣, and sh-control producing vectors were generated by ligation of double-stranded oligonucleotides that contained the siRNA-expressing sequences targeting c-Jun, JunB, HNF1␣, or a scrambled sequence (small interfering control), respectively, into the BglII-HindIII sites of the pSuper.GFP/ neo vector (Oligoengine, Seattle, WA). The sequences of all of the oligonucleotides are shown in supplemental Table S1. Expression vectors for human c-Jun and JunB have been described previously (36).
Cell Culture, Transient Transfections, and Reporter Assays-Human hepatoma HepG2 cells, mouse Hep6-16 cells, and human embryonic kidney cells (HEK293T) were cultured in DMEM supplemented with 10% fetal bovine serum, L-glutamine, and penicillin/streptomycin at 37°C in a 5% CO 2 atmosphere. For the treatment of cells with PMA, the cells were serum-starved for 3 h and then stimulated with 100 or 200 nM PMA for various times. The PKC inhibitor Gö 6976 (1 M), the JNK inhibitor SP600125 (10 M), or the histone deacetylase inhibitor trichostatin A (1 M) was added 1 h before the treatment with PMA. Transient transfections were performed using the Ca 3 (PO4) 2 co-precipitation method. Transient transfections of cells with the sh-RNA producing vectors were performed with the Lipofectamine TM 2000 reagent according to the manufacturer's instructions. Luciferase assays were performed using the luciferase assay kit from Promega according to the manufacturer's instructions. Normalization for transfection efficiency was performed by ␤-galactosidase assays.
ChIP Assays-Chromatin immunoprecipitation was performed as described previously (37) using chromatin from HepG2 cells and antibodies against JunB, c-Jun, or HNF-1␣. Immunoprecipitated chromatin was analyzed by PCR using primers shown in supplemental Table S3. PCR products were analyzed by agarose gel electrophoresis.
Statistical Analysis-The results are shown as the means ؎ S.D. Statistical significance was determined using the Student's t test. Differences with p Ͻ 0.05 were considered to be statistically significant.

Jun Proteins Inhibit ApoM Gene Transcription in Hepatic
Cells-To investigate the role of Jun proteins in apoM gene regulation, we overexpressed two members of the Jun family (c-Jun and JunB) in HepG2 cells and monitored the activity of the Ϫ950/ϩ42 human apoM promoter (28) by luciferase assays. This experiment showed that overexpression of c-Jun or JunB was associated with a potent and dose-dependent down-regulation of apoM promoter activity (Fig. 1). A similar observation was made using an additional member of the Jun family (JunD; data not shown).
To activate endogenous Jun proteins in HepG2 cells, we used PMA, a potent inducer of AP-1 gene transcription (38). As shown in Fig. 2A, PMA caused a strong and dose-dependent reduction in apoM mRNA levels, and this reduction was counterbalanced by overexpressing shRNAs targeting both c-Jun and JunB but not by a control shRNA (Fig. 2B). In agreement with these findings, PMA inhibited the activity of the Ϫ950/ ϩ42 apoM promoter, and this inhibition was counterbalanced by overexpressing shRNAs targeting both c-Jun and JunB but not by a control shRNA (Fig. 2C). The efficiency of c-Jun and JunB gene silencing by the corresponding shRNAs is shown in supplemental Fig. S1.
We also monitored the protein levels of Jun and apoM in HepG2 cells treated with 200 nM PMA for 18 h. As shown in Fig.  2D, PMA induced of both c-Jun and JunB gene expression, and this was associated with a potent reduction in the protein levels of apoM. In conclusion, the data of Fig. 2 showed that phorbol esters down-regulate apoM gene expression in HepG2 cells in a Jun-dependent manner.
Protein Kinase C ␣ Inhibits apoM Gene Expression via Jun Proteins-Phorbol esters are potent activators of PKC (33,34). Treatment of HepG2 cells with PMA caused an early (10 min) activation of PKC␣, which was followed by an increase in total and phosphorylated forms of c-Jun at 1 h of treatment (supplemental Fig. 2, A and B). This activation of c-Jun by PMA was abolished by the JNK-specific inhibitor SP 600125 in agreement with previous studies showing that phorbol esters activate PKC and JNK, and this leads to an increase in c-Jun protein levels via a positive autoregulatory loop (39,40).
To investigate the potential role of PKC in apoM gene regulation, we treated HepG2 cells with the PKC␣/␤-specific inhibitor Gö 6976 (41). Gö 6976 abolished the inhibition of apoM gene expression by PMA both at the mRNA (  Following transfection, the cells were treated with 200 nM PMA for 18 h or left untreated. Total RNA was extracted, and apoM mRNA levels were analyzed by RT-PCR. The mRNA levels of the GAPDH gene were determined for normalization purposes. The relative mRNA levels of the apoM gene were quantified and are shown as a histograph. Each value represents the average from three independent experiments. C, HepG2 cells were transiently transfected with the (Ϫ950/ϩ42)human apoM-luc reporter plasmid (1.0 g) along with the sh-control or sh-c-Jun and sh-JunB producing vectors (2.0 g each), as indicated, and a ␤-galactosidase expression vector (1.0 g). Following transfection, the cells were treated with 100 nM PMA for 18 h or left untreated. Luciferase activity was normalized to ␤-galactosidase activity and presented with histographs. Each value represents the average (ϮS.D.) from at least three independent experiments performed in duplicate. D, HepG2 cells were treated with 200 nM PMA for 18 h or left untreated, and the protein levels of c-Jun, JunB, apoM, HNF-1␣, and actin (loading control) were determined by immunoblotting using the corresponding antibodies. The experiment was performed three times, and representative images are presented. **, p Ͻ 0.01; ***, p Ͻ 0.001. cells was associated with a 70% reduction in the mRNA levels of the apoM gene, and this reduction was counterbalanced by shRNAs targeting c-Jun and JunB but not by a control shRNA (Fig. 3C). In conclusion, the combined data of Figs. 2 and 3 establish a PMA/PKC/Jun pathway operating in HepG2 cells that inhibit apoM gene transcription.
The Ϫ53/Ϫ47 Region of the apoM Promoter Contains an AP-1-responsive Element-To map the region of the apoM promoter that is responsible for the inhibition by Jun proteins, we utilized a panel of apoM promoter deletion mutants shown schematically in Fig. 4A (28). Transient transfections of HepG2 cells with these reporter plasmids in the absence or in the presence of expression vectors for c-Jun (Fig. 4B) or JunB (Fig. 4C) established that the minimal region of the apoM promoter that is required for Jun-mediated inhibition is defined by nucleotides Ϫ105 and Ϫ49.
Binding of endogenous c-Jun and JunB proteins to the apoM promoter in vivo was shown by chromatin immunoprecipitation assays in HepG2 cells. As shown in Fig. 5B, both c-Jun and JunB were recruited to the proximal (Ϫ214/Ϫ14) apoM promoter, although with different affinities (top row, lanes 3 and 4). In contrast, binding of both proteins to the distal region of the apoM promoter (Ϫ950/Ϫ616) was not significant (Fig. 5B, middle row, lanes 3 and 4). As a control, we used primers that amplify the Ϫ313/Ϫ185 region of the promoter of the human RhoB gene (42), and we found no recruitment of Jun proteins to this region (Fig. 5B, bottom row, lanes 3 and 4).
To identify more precisely the putative AP-1 element on the proximal apoM promoter, a series of DNA affinity precipitation (DNAP) assays was performed. As shown in Fig. 5C, both c-Jun and JunB proteins bound to a biotinylated Ϫ105/ϩ42 apoM promoter fragment. In contrast, no binding of c-Jun or JunB could be observed using a biotinylated Ϫ241/Ϫ81 apoM promoter fragment or a control biotinylated oligonucleotide corresponding to the Ϫ76/Ϫ43 region of the human RhoB promoter (Fig. 5C).
A close inspection of the DNA sequence of the proximal apoM promoter region between nucleotides Ϫ241 and ϩ42 revealed the presence of a putative AP-1 element at position Ϫ53/Ϫ47 (Fig. 5A). This element contains the sequence 5Ј-TTACTCA-3Ј and differs from the consensus AP-1 site (5Ј-TGACTCA-3Ј) (43) by only one nucleotide at the second position (T instead of a G). Interestingly, this putative AP-1 element resides within a previously characterized binding Initially, we confirmed the binding of c-Jun and JunB to this site by performing DNAP assays using a biotinylated oligonucleotide corresponding to the Ϫ61/Ϫ38 region of the apoM promoter (Fig. 5D). As anticipated, HNF-1␣ bound equally well to the same oligonucleotide as shown in Fig. 5E. To verify that the 5Ј-TTACTCA-3Ј element is a true AP-1-binding site, we introduced mutations at positions 5, 6, and 7 of this element (5Ј-TTACGGG-3Ј) (Fig. 5A). As shown in Fig. 5D, neither c-Jun nor JunB could bind to the mutated oligonucleotide (Ϫ61/ Ϫ38)mut. As expected, these mutations also abolished the binding of HNF-1␣ (Fig. 5E).
We introduced the same nucleotide substitutions into the corresponding positions of the Ϫ950/ϩ42 apoM promoter (Fig. 6A) and examined the activity of the mutant versus the wild type promoter in HepG2 cells by luciferase assays. At basal conditions, the activity of the mutant Ϫ950/ϩ42 apoM promoter was very low (8%) compared with the wild type promoter, confirming that this element is essential for apoM gene expression in hepatic cells (Fig. 6B). Importantly, mutations in the AP-1 element abolished the inhibition of apoM promoter activity by PMA (Fig. 6C) or by PKC␣ (Fig. 6D), strongly suggesting that this element, in addition to its importance for basal apoM expression, is essential for the regulation of the apoM gene by the PKC/Jun pathway.
We then showed that neither c-Jun nor JunB could repress the activity of the wild type Ϫ950/ϩ42 apoM promoter in HEK293T cells that lack endogenous HNF-1␣ protein (Fig. 7A). HNF-1␣ caused a strong (40-fold) transactivation of the wild type apoM promoter but only a minor (4.4-fold) transactivation of the mutant apoM promoter (Fig. 7A). This finding is in agreement with the data of Fig. 5E, which showed that the mutations abolished the binding of HNF-1␣ to the apoM promoter. More importantly, both c-Jun and JunB inhibited the HNF-1␣-mediated transactivation of the wild type but not of the mutated apoM promoter (Fig. 7A), suggesting negative cross-talk between Jun proteins and HNF-1␣.
In HepG2 cells (that express endogenous HNF-1␣), c-Jun and JunB inhibited the basal activity of the proximal Ϫ105/ϩ42 apoM promoter that contains the dual HNF-1␣/AP-1 element, and this inhibition was compromised by HNF-1 overexpression (Fig. 7B). In contrast, c-Jun and JunB could not suppress the activity of the Ϫ105/ϩ42 apoM promoter in HepG2 cells in which the endogenous HNF-1␣ gene had been silenced by a shRNA targeting HNF-1␣ ( Fig. 7C and supplemental Fig. S1A).
The Mouse apoM Promoter Does Not Contain a Functional AP-1 Site-The findings of Figs. 5-7 indicated that the Ϫ55/ Ϫ41 region of the human apoM promoter contains a dual HNF-1␣/AP-1-responsive element that mediates activation or repression of apoM gene expression by HNF-1␣ and Jun factors, respectively. Interestingly, the corresponding region in the mouse apoM promoter is highly conserved with the exception of a single nucleotide substitution at position 6 of the AP-1 element (5Ј-TTACTTA-3Ј) (supplemental Fig. S3A). Using DNAP assays, we initially showed that HNF-1␣ binds to both the mouse Ϫ109/Ϫ86 and the human Ϫ61/Ϫ38 elements with equal affinity (supplemental Fig. S3B). In contrast, c-Jun and JunB could not bind to the mouse element, possibly because of the nucleotide substitution (supplemental Fig. S3B). In agreement with the DNA binding data of supplemental Fig. S3B, c-Jun and JunB could not inhibit the basal activity of the mouse apoM promoter in HepG2 cells (supplemental Fig. S3C). Furthermore, PMA could not inhibit endogenous apoM gene  , input). The experiment was performed three times, and representative images are presented. C, DNA affinity precipitation using nuclear extracts from HEK293T cells transiently transfected with expression vectors for JunB or c-Jun and biotinylated PCR fragments corresponding to the Ϫ241/Ϫ81 or the Ϫ105/ϩ42 region of the human apoM promoter or no probe. A biotinylated oligonucleotide corresponding to the Ϫ76/Ϫ43 region of the human RhoB promoter was used as a negative control. Oligonucleotide-bound JunB or c-Jun was detected by Western blotting using an anti-JunB or anti-c-Jun antibody, respectively. D, DNA affinity precipitation using nuclear extracts from HEK293T cells transiently transfected with expression vectors for JunB or c-Jun and biotinylated oligonucleotides corresponding to the wild type Ϫ61/Ϫ38 region of the human apoM promoter, the Ϫ61/Ϫ38 region bearing mutations in the AP-1 site (mut), the Ϫ76/Ϫ43 region of the human RhoB promoter, or no oligonucleotide (no oligo). Oligonucleotide-bound JunB or c-Jun was detected by Western blotting using an anti-JunB or anti-c-Jun antibody, respectively. E, DNA affinity precipitation using nuclear extracts from HEK293T cells transiently transfected with an expression vector for HNF-1␣ and biotinylated oligonucleotides corresponding to the wild type Ϫ61/Ϫ38 region of the human apoM promoter, the Ϫ61/Ϫ38 region bearing mutations in the AP-1 site (mut), the Ϫ76/Ϫ43 region of the human RhoB promoter, or no oligonucleotide (no oligo). Oligonucleotide-bound ⌯⌵F-1␣ was detected by Western blotting using a polyclonal anti-⌯⌵F-1␣ antibody. All of the experiments in B-D were performed at least three times, and representative images are shown. expression in mouse hepatocytes (Hep6-16 cells; supplemental Fig. S3D).
Binding of HNF-1␣ and AP-1 Factors to the Proximal apoM Promoter Is Mutually Exclusive-Using chromatin immunoprecipitation assays, we sought to examine the recruitment of HNF-1␣ and AP-1 factors on the apoM promoter under basal conditions as well as under conditions of AP-1 activation by PMA. As shown in Fig. 8 (A and B), under basal conditions, HNF-1␣ bound strongly to the apoM promoter, whereas binding of either c-Jun or JunB was barely detectable. Upon a 2-h treatment with PMA, binding of HNF-1␣ to the apoM promoter was markedly reduced (Fig. 8, A and B, top row, lane 7). In contrast, binding of both c-Jun (Fig. 8A, top row, compare lanes  4 and 8) and JunB (Fig. 8B, top row, compare lanes 4 and 8) to  the proximal apoM promoter was enhanced upon PMA treatment. No binding of HNF-1␣ or Jun proteins was observed on the distal (Ϫ950/Ϫ616) apoM promoter or to the unrelated RhoB promoter (Fig. 8, A and B, middle and bottom row, respectively). By immunoblotting analysis, we showed that the 2-h PMA treatment had no effect on the expression of the endogenous HNF-1␣ gene in HepG2 cells (Fig. 8C).
The findings of Fig. 8 indicated that the binding of HNF-1␣ and Jun proteins to the dual specificity responsive element on the proximal apoM promoter could be mutually exclusive. To further investigate this mechanism, we performed a series of in vitro DNAP assays. These experiments showed that binding of HNF-1␣ to the Ϫ61/Ϫ38 apoM oligonucleotide probe could be competed out by increasing concentrations of c-Jun (Fig. 9A,  top row, compare lane 3 with lanes 6 and 9) or JunB (Fig. 9B, top  row, compare lane 3 with lanes 6 and 9). Similarly, binding of c-Jun (Fig. 9C) or JunB (Fig. 9D) to the Ϫ61/Ϫ38 apoM oligonucleotide could be competed out by increasing concentrations of HNF-1␣ (bottom row, compare lane 3 with lanes 6 and 9).
These findings provided additional evidence that binding of HNF-1␣ and Jun proteins to the Ϫ61/Ϫ38 region of the apoM promoter is mutually exclusive.

Antagonistic Interactions between HNF-1␣ and AP-1 Factors Control the Transcription of the Human Apolipoprotein A-II
Gene-In a search for additional genes that could be regulated by a similar mechanism, we identified a putative AP-1 site within a previously identified HNF-1␣-binding site in the promoter of the human apoA-II gene (element H, Ϫ573/Ϫ554) (supplemental Fig. S4A) (44,45). This element is identical in the mouse apoA-II gene (supplemental Fig. S4H).
Initially, we showed that apoA-II mRNA levels in HepG2 cells were decreased following a 18-h PMA treatment, and this decrease was not observed upon overexpression of shRNA targeting c-Jun gene expression or in the presence of the PKC␣/␤ inhibitor Gö 6976 (supplemental Fig. S4, B and C, respectively). Using chromatin immunoprecipitation assays, we established that c-Jun, but not JunB, was recruited to the Ϫ649/Ϫ343 region of the apoA-II promoter that contains the putative AP-1 element under basal conditions (supplemental Fig. S4D). Overexpression of c-Jun in HepG2 cells was associated with a 50% reduction in the activity of the Ϫ911/ϩ29 apoA-II promoter (supplemental Fig. S4E). DNAP assays established that the region of the apoA-II promoter between nucleotides Ϫ573 and Ϫ554 is a dual HNF-1/AP-1 site that binds both c-Jun and HNF-1␣ (supplemental Fig. S4F). Finally, using chromatin immunoprecipitations, we showed that there is competition between c-Jun and HNF-1␣ for binding to the apoA-II promoter under conditions of c-Jun activation by PMA (supplemental Fig. S4G).

DISCUSSION
HDL is a heterogeneous population of particles in the plasma comprising larger spherical particles (such as HDL 2 and HDL 3 ) as well as smaller (pre␤-HDL) discoidal particles. HDL particles in plasma are in a dynamic equilibrium that is governed by a continuous remodeling mediated by plasma enzymes, membrane transporter proteins, and receptors (46). HDL undergoes significant changes in structure, protein/lipid composition, antioxidative, and anti-inflammatory activities during the acute phase response (47). Apolipoprotein M is a recently described apolipoprotein that is associated with HDL and has been shown to play an important role in HDL maturation and remodeling (21). The atheroprotective role of apoM was established by studies showing that adenoviral apoM overexpression in LDL receptor-deficient mice (animal model for premature atherosclerosis) or hepatic overexpression of apoM in apoM transgenic mice led to a reduction in atherosclerosis development (21). Thus, it could be anticipated that reduction in apoM synthesis in the liver during inflammation could contribute to the development of atherosclerosis and coronary artery disease.
Hormone nuclear receptors play an important role in the regulation of HDL genes during inflammation. We have shown recently that the proximal apoM promoter harbors a multifunctional hormone response element that serves as a binding site for a plethora of orphan and ligand-inducible nuclear receptors including HNF-4, homodimers of retinoid X receptor ␣, and heterodimers of retinoid X receptor ␣ with liver X receptor ␣, retinoic acid receptor ␣, peroxisome proliferator-activated receptor ␣, and thyroid hormone receptor ␤ (28). This site was shown to mediate apoM gene induction by ligands for the above nuclear receptors, whereas mutagenesis of this site severely reduced both the basal and the inducible activity of the apoM promoter in hepatic cells (28). Furthermore, preliminary evidence for a cooperation between HNF-1 and the above hormone nuclear receptors was obtained (28). These findings, combined with our previous studies showing that TNF-␣ inhibits the transcriptional activity of nuclear receptor HNF-4␣ in hepatic cells via NF-〉 activation (48), suggested that regulation of apoM gene transcription during inflammation may be complex and mediated by multiple signaling pathways and molecular cross-talks.
The findings of the present study established PKC as a negative modulator of apoM gene transcription. PKC comprises a family of serine/threonine kinases that play roles in many diseases including atherosclerosis, fibrosis, and cardiac hypertrophy (34). In the case of atherosclerosis, it was shown that inhibition of PKC blocks the production of superoxide anion by activated monocytes and the subsequent oxidation of LDL (49). Both ␣ and ␤ PKCs were shown to be up-regulated upon stimulation of monocytes and translocate to the particulate fraction (50). These findings suggest that specific PKC isozymes may increase the pro-oxidant environment of the atherosclerotic plaque and may lead to plaque instability. Furthermore, by down-regulating apoM gene expression in the liver, activated PKC is expected to influence the maturation of HDL particles in FIGURE 9. Mutually exclusive binding of c-Jun/JunB and HNF-1␣ to the ؊61/؊38 region of the apoM promoter. A, DNA affinity precipitation using nuclear extracts from HEK293T cells transiently transfected with an expression vector for HNF-1␣ and increasing amounts of c-Jun and biotinylated oligonucleotides corresponding to the Ϫ61/Ϫ38 region of the human apoM promoter or no oligonucleotide (no oligo). Oligonucleotide-bound HNF-1␣ and c-Jun were detected by Western blotting using an anti-HNF-1␣ or anti-c-Jun antibody, respectively. B, DNA affinity precipitation using nuclear extracts from HEK293T cells transiently transfected with an expression vector for HNF-1␣ and increasing amounts of JunB and biotinylated oligonucleotides corresponding to the Ϫ61/Ϫ38 region of the human apoM promoter or no oligonucleotide (no oligo). Oligonucleotide-bound HNF-1␣ and JunB was detected by Western blotting using an anti-HNF-1␣ or anti-JunB antibody, respectively. C, DNA affinity precipitation using nuclear extracts from HEK293T cells transiently transfected with an expression vector for c-Jun and increasing amounts of HNF-1␣ and biotinylated oligonucleotides corresponding to the Ϫ61/Ϫ38 region of the human apoM promoter or no oligonucleotide (no oligo). Oligonucleotide-bound HNF-1␣ and c-Jun was detected by Western blotting using an anti-HNF-1␣ or anti-c-Jun antibody, respectively. D, DNA affinity precipitation using nuclear extracts from HEK293T cells transiently transfected with an expression vector for JunB and increasing amounts of HNF-1␣ and biotinylated oligonucleotides corresponding to the Ϫ61/Ϫ38 region of the human apoM promoter or no oligonucleotide (no oligo). Oligonucleotide-bound HNF-1␣ and JunB was detected by Western blotting using an anti-HNF-1␣ or anti-JunB antibody, respectively. All of the experiments were performed at least three times, and representative images are shown. the plasma, and as a consequence, HDL structure and functions including atheroprotective functions on the arterial wall and the reverse cholesterol transport. Thus, selective regulators of PKC isozymes may be promising candidates for the prevention or even the regression of atherosclerosis via HDL particles (34).
We report here the identification of a dual specificity regulatory element in the proximal apoM promoter that mediates activation or repression of apoM promoter activity by HNF-1 and Jun proteins, respectively, in hepatic cells. The experimental evidence that supports the dual functionality of this element of the apoM promoter is as follows: (a) Jun proteins repress apoM promoter activity only in cells expressing HNF-1␣. Junmediated repression was abolished in HepG2 cells in which endogenous HNF-1␣ expression had been silenced by specific shRNAs (Fig. 7). (b) Mutations in the AP-1 element abolished the binding of HNF-1␣ and the HNF-1␣-mediated transactivation of the apoM promoter (Figs. 5-7). (c) The mouse apoM promoter does not contain a functional AP-1 site because of single nucleotide difference, but it retains a functional HNF-1␣-responsive element. As a consequence, Jun proteins cannot inhibit the mouse apoM promoter activity, and PMA cannot inhibit the expression of the endogenous apoM gene in mouse hepatocytes (supplemental Fig. S3). (d) Competition experiments showed that binding of Jun proteins and HNF-1␣ to the apoM promoter is mutually exclusive (Fig. 9), and HNF-1 overexpression in hepatic cells could overcome the inhibitory effect of Jun proteins on the apoM promoter (Fig. 7B). (e) Chromatin immunoprecipitation assays established that AP-1 activation leads to the recruitment of c-Jun and Jun B proteins to the proximal apoM promoter with the simultaneous displacement of HNF-1 (Fig. 8). (f) A similar mechanism of Jun-mediated repression of promoter activity via dual specificity AP-1/HNF-1-responsive element operates on the promoter of the human apolipoprotein A-II gene (supplemental Fig. S4), the expression of which is also under negative regulation during the acute phase response (51). Importantly, this dual specificity element is highly conserved in the mouse apoA-II gene (supplemental Fig. 4H).
A previous study had shown that administration of LPS to mice (model of bacterial infection) was associated with a marked reduction in the expression levels of HNF-1␣ in the liver (52). The same group also showed that both apoM and HNF-1␣ genes were down-regulated in the kidney of LPStreated mice (53). We show here that the expression of HNF-1␣ in HepG2 cells was not affected by phorbol ester treatment (Fig.  8C), suggesting that a mechanism involving a direct effect of Jun on HNF-1␣ gene expression in hepatic cells could be excluded.
The question that remains is why recruitment of Jun proteins to the apoM and apoA-II promoters and the displacement of HNF-1␣ is associated with transcriptional repression. We could hypothesize that Jun proteins bind to the AP-1 elements of the apoM and apoA-II promoter without activating transcription. Active repression by Jun could be excluded in light of the findings that both c-Jun and JunB failed to repress the activity of the apoM promoter in HEK293T cells that lack hepatocyte nuclear factors (Fig. 7A), as well as by our finding that a general inhibitor of histone deacetylases (trichostatin A) failed to block apoM gene down-regulation by PMA (supplemental Fig. S5). The inability of Jun proteins to activate the apoM promoter could be either due to the absence of cooperating factors in the vicinity of the AP-1 element or due to the deviation of the apoM AP-1 element from the consensus sequence (Fig. 5A). Transcriptional repression by Jun and other AP-1 proteins has been reported in many cases including the genes encoding for c-Fos, osteocalcin, adipocyte P2, creatinine kinase, myoD, major histocompatibility class I, chorionic gonadotropin ␣ and ␤, apoC-III, and insulin (36, 54 -61). However, in the majority of the above cases, the molecular mechanisms by which AP-1 proteins inhibit transcription are not understood.
In summary, our model described here provides a mechanistic explanation for the inhibitory effect of pro-inflammatory cytokines or other factors on the expression of acute phase HDL genes such as apoM and apoA-II. HDL is a lipoprotein particle with multiple but still not fully characterized atheroprotective properties. Understanding how the HDL genes are controlled during health and disease will be essential for the development of novel therapeutic and diagnostic tools for diseases such as coronary artery disease and diabetes that affect a large proportion of the population in Western countries.