The Zinc Finger Protein 202 (ZNF202) Is a Transcriptional Repressor of ATP Binding Cassette Transporter A1 (ABCA1) and ABCG1 Gene Expression and a Modulator of Cellular Lipid Efflux*

The zinc finger gene 202 (ZNF202) located within a hypoalphalipoproteinemia susceptibility locus on chromosome 11q23 is a transcriptional repressor of various genes involved in lipid metabolism. To provide further evidence for a functional linkage between ZNF202 and hypoalphalipoproteinemia, we investigated the effect of ZNF202 expression on ATP binding cassette transporter A1 (ABCA1) and ABCG1. ABCA1 is a key regulator of the plasma high density lipoprotein pool size, whereas ABCG1 is another mediator of cellular cholesterol and phospholipid efflux in human macrophage. We demonstrate here that the full-length ZNF202m1 isoform binds to GnT repeats within the promotors of ABCA1 (−229/−210) and ABCG1 (−572/−552). ZNF202m1 expression in HepG2 cells dose-dependently repressed the promotor activities of ABCA1 and ABCG1. This transcriptional effect required the presence of the SCAN domain in ZNF202 and the functional integrity of a TATA box at position −24 of ABCA1, whereas the presence of GnT binding motifs was nonessential. The state of ZNF202 SCAN domain oligomerization affected the ability of the adjacent ZNF202 Krüppel-associated box domain to recruit the transcriptional corepressor KAP1. Overexpression of ZNF202m1 in RAW264.7 macrophages prevented the induction of ABCA1 gene expression by 20(S)OH-cholesterol and 9-cis-retinoic acid, further substantiating the interference of ZNF202 in critical elements of transcriptional activation. Finally, HDL and apoAImediated lipid efflux was significantly reduced in RAW264.7 cells stably expressing ZNF202m1. In conclusion, we have identified ABCA1 and ABCG1 as target genes for ZNF202-mediated repression and thus, provide evidence for a functional linkage between ZNF202 and hypoalphalipoproteinemia.

Molecular factors that determine plasma high density lipoprotein (HDL) 1 cholesterol levels include several known genes with defined roles in reverse cholesterol transport (1) and a variety of susceptibility loci with to date poorly characterized candidate genes. Mutations in the ATP binding cassette transporter A1 (ABCA1) gene have recently been causatively linked to familial HDL-deficiency syndromes (2)(3)(4). The transporter ABCA1 is regulated by cholesterol flux and facilitates the apoAIdependent cellular export of cholesterol and phosholipids, thereby acting as a key regulator of plasma HDL (2,3,5). The half-size transporter ABCG1 is another member of the group of cholesterol-responsive ABC transporters. ABCG1, like ABCA1, has been shown to regulate cellular cholesterol and phospholipid efflux (2,3,6).
In past years, evidence has accumulated to suggest that a number of transcription factors play critical roles in the coordinate transcriptional regulation of genes involved in lipid metabolism (7,8). Linkage analysis in large Utah pedigrees led to the identification of a low HDL-cholesterol locus on chromosome 11q23 that is distinct from the apoAI/C-III/AIV gene cluster (9). This novel familial susceptibility locus for hypoalphalipoproteinemia contains the zinc finger protein 202 (ZNF202) originally described as a testis-specific transcription factor (9,10). ZNF202 is expressed in two common splice variants. The m1 splice form encodes a full-length protein of 648 amino acids with an amino-terminal SCAN domain, a central KRAB repression domain, and 8 carboxyl-terminal Cys 2 -His 2 zinc finger motifs. The m3 splice form encodes a carboxylterminal-truncated protein of 133 amino acids that contains only the SCAN domain. The SCAN domain of ZNF202 has been shown to mediate selective protein oligomerization and the zinc finger motifs to bind to specific DNA elements (9,11). Intriguingly, the ZNF202 DNA binding elements are present in promotors of various genes involved in lipid metabolism including apolipoproteins and lipid-processing enyzmes. ZNF202 has been therefore proposed to function as a transcriptional regulator of lipid metabolism (9).
Based on this information, we tested the hypothesis of whether the ABC lipid transporters ABCA1 and ABCG1 are transcriptionally regulated by ZNF202. Our results provide evidence that ZNF202 acts as a transcriptional repressor of both ABCA1 and ABCG1 and, thus, establish a functional link between ZNF202 and hypoalphalipoproteinemia. seeded in 6-well plates overnight before transfection. In some transfection experiments HepG2 cells were washed 4 h after transfection and subsequently incubated in Dulbecco's modified Eagle's medium containing 5% lipoprotein-deficient serum, 10 M 20(S)-OH-cholesterol, and 10 M 9-cis-retinoic acid (Sigma). In all experiments cells were harvested after 24 h to measure luciferase activity or to prepare RNA and nuclear extracts.
Cloning of Expression Constructs-cDNAs encoding either the open reading frame of the ZNF202m1 isoform (GenBank™ accession number aF027219, nucleotide positions 8 -1960) (2,3,10) or a truncated product lacking the SCAN domain (nucleotide positions 920 -1960) were cloned into pcDNA3.1/V5/His-Topo plasmid (Invitrogen), and the sequence was confirmed by using an automated fluorescence DNA sequencer and the ALFexpress AutoRead sequencing kit (Amersham Pharmacia Biotech). The pcDNA3.1 ZNF202 (DV Ͼ AA) construct with a double amino acid substitution (DV Ͼ AA) in the ZNF202 KRAB domain was generated by using a site-directed mutagenesis kit (CLONTECH). GST fusions constructs with the ZNF202 SCAN or KRAB domain were generated as previously described (11). A cDNA encoding KAP1 was obtained from Edge Bio Systems and directionally subcloned into the pcDNA3.1/His expression vector (Invitrogen). Sequence fidelity was assessed by using an ABI PRISM 377 DNA sequencer (PerkinElmer Life Sciences).
In Vitro Protein Expression-KAP1 and ZNF202 cDNA templates were expressed in vitro in rabbit reticulocyte lysates in the presence of [ 35 S]methionine (Amersham Pharmacia Biotech) using the TNT T7 Quick coupled transcription/translation system (Promega) according to the manufacturer's procedure. The synthesized products were confirmed by SDS-PAGE. Lysate samples (10 l) were separated on readyto-use gels (Bio-Rad) and analyzed by autoradiography.
Electrophoretic Mobility Shift Assays-Radiolabeled double-stranded oligonucleotide probes (equivalent of 30,000 cpm) were added to ZNF202-expressing rabbit reticulocyte lysates (1-2 l) in a buffer containing 50 mM HEPES/HCl, pH 7.9, 6 mM MgCl 2 , 50 mM dithiothreitol, 100 g/ml bovine serum albumin, 0.01% Nonidet P-40, and 2 g of poly(dI-dC) (Amersham Pharmacia Biotech) and incubated for 20 min at room temperature. The respective promotor sequences encoded by the oligonucleotides are described in Fig. 1. Supershift experiments resulted from the addition of KRAB antibodies, N84 (1 l). In competition experiments, nuclear extracts were preincubated for 10 min with a 50-fold molar excess of unlabeled oligonucleotides before the addition of radiolabeled probe. DNA-protein complexes were finally resolved on a native 8% polyacrylamide gel and analyzed by autoradiography.
Cloning of Reporter Gene Constructs-Reporter constructs for the ABCA1 (GenBank™ accession number AJ252201; position Ϫ919 to ϩ224), ABCG1 (GenBank™ accession number AJ289137; position Ϫ2912 to ϩ50), and apolipoprotein AIV (GenBank™ accession number X13368; position Ϫ718 to ϩ30) promotor sequences as well as reporter constructs for potential ZNF202 binding sites (GnT tandem repeats) were cloned into the BglII and NheI restriction sites of the pGL3-basic vector (Promega) and confirmed by sequence analysis (2,3,9,13). The functional impact of a TATA box within the proximal ABCA1 promotor region and of ZNF202 binding to GnT or GnC motifs was assessed with reporter gene constructs containing mutated binding sites or truncations at position Ϫ175, Ϫ79, or ϩ12. Constructs with mutated Ϫ229 GnT, Ϫ91 GnC, and Ϫ24 TATA motifs were generated by using a two-step cloning strategy. Two polymerase chain reaction fragments were generated that overlap at the DNA motif of interest, replacing it by either a HindIII (Ϫ229 GnT), SacI (Ϫ91 GnC), or an EcoRI (Ϫ24 TATA) restriction site. After restriction enzyme digestion, the overlapping fragments were ligated and finally cloned into pGL3 basic vectors. All vectors were sequenced on an automated fluorescence DNA sequencer using the ALFexpress AutoRead sequencing kit (Amersham Pharmacia Biotech).
Luciferase Reporter Assays-HepG2 cells were transiently transfected with Fugene ® reagent (Roche Molecular Biochemicals) as described by the manufacturer. Two g of the respective reporter gene constructs were co-transfected with 1 g of the pSV ␤-galactosidase vector (Promega) and increasing doses (up to 4 g) of ZNF202 expression vectors. The total amount of transfected DNA was set constant by supplementation with empty expression vector. A promotorless pGL3basic vector served as control. Cells were lysed in reporter lysis buffer (Promega), and after centrifugation, luciferase assay reagent with luciferyl-CoA (Promega) was added to the supernatant as recommended by the manufacturer. Luciferase activity was finally determined in a LUMAT LB9501 (Berthold). All data were normalized for protein concentrations (Bio-Rad protein assay) and ␤-galactosidase activity (Promega ␤-galactosidase enzyme assay). Each experiment was repeated three times, and measurements were performed in triplicates. Results are expressed either as relative inhibition of luciferase activity in cells transfected by ZNF202m1 in comparison to empty expression vector or as multiples in comparison to promotorless pGL3-basic vector activity.
Affinity Purification of KAP1-GST and GST fusion proteins of ZNF202 (1.0 g) were incubated overnight at 4°C with 2.0 l of the KAP1 cDNA template expressing rabbit reticulocyte lysate sample supplemented with HNTG buffer (20 mM HEPES, pH 7.4, 150 mM NaCl, 0.1% Triton X-100, and 10% glycerol) diluted to a final volume of 200 l. 15 l of glutathione-Sepharose beads (Amersham Pharmacia Biotech) were subsequently added for 40 min to collect the protein complexes. All samples were washed three times with ice-cold HNTG buffer, boiled in electrophoresis buffer, and analyzed by SDS-PAGE and autoradiography using an autoradiographic image enhancer (National Diagnostics).
Analysis of ZNF202m1 Binding to KAP1-For each binding reaction, 1 l of anti-KAP1 antibody was incubated with 500 g of total HeLa lysate proteins for 90 min at 4°C. After an additional incubation for 40 min at 4°C in the presence of 15 l of protein G-Sepharose beads (Amersham Pharmacia Biotech), the KAP1 immunocomplexes were washed four times with ice-cold HNTG buffer. The KAP1 immunocomplexes were thereupon incubated with 2 l of the in vitro translated proteins (ZNF202m1 and/or SDP1, mFPM315) in a final volume of 100 l for 90 min at 4°C. Bead-coupled KAP1 complexes were finally regenerated by centrifugation and washed four times with HNTG buffer. All binding reactions were analyzed by SDS-PAGE and fluorography.
Northern Blot Analysis-HepG2 cells were transfected with increasing amounts of ZNF202 expression vectors using Fugene ® reagent (Roche Molecular Biochemicals). The transfection efficiency ranged between 50 and 70%. Total RNA was isolated 48 h after transfection for Northern blot analysis using the QIAamp RNA isolation kit (Qiagen). Aliquots of 10 g of total RNA were separated on denatured agarose gels and transferred to nylon membranes (Amersham Pharmacia Biotech). The membranes were hybridized with a random-prime radiolabeled ABCA1-specific probe that was derived from the 5Ј end of the coding sequence as described elsewhere (2,3,14). The membranes were rehybridized with a ubiquitin-specific probe to confirm equal RNA loading. The tissue-specific expression of ZNF202 was assessed by hybridization of a multiple tissue poly(A) ϩ RNA master blot featuring 75 distinct human tissues (CLONTECH) with a radiolabeled ZNF202specific probe. Individual signals on the autoradiogram were measured densitometrically and expressed as relative mRNA abundance in comparison with testis tissue (set to 100%).
Efflux Experiments-RAW264.7 cells were transfected with a ZNF202m1 encoding pcDNA3.1/V5/His-Topo vector or an empty control vector, and stable colonies were selected in Dulbecco's modified Eagle's medium containing 500 l/ml neomycin according to standard protocols. ZNF202 protein expression was confirmed by Western blotting. Efflux assays were performed as recently described with minor modifications (2,3,15). Briefly, stably transfected RAW264.7 cells were radiolabeled with 1.5 Ci/ml [ 14 C]cholesterol and 10 Ci/ml [ 3 H]choline and loaded with 40 g/ml enzymatically modified low density lipoprotein that was prepared as described elsewhere (16). Cells were incubated in 6-well plates for 24 h in Dulbecco's modified Eagle's medium supplemented with 5% lipoprotein-deficient serum and 10 M 20(S)-OH-cholesterol, 10 M 9-cis-retinoic acid (Sigma), or 0.02% v/v ethanol. The cells were then washed and chased for 17 h with either 100 g/ml HDL 3 protein, 10 g/ml purified apoAI (Sigma), or 0.2% bovine serum albumin in the described medium in the absence of radiolabeled lipids and enzymatically modified low density lipoprotein. Lipids were finally extracted as previously described (17). Radioactivity was determined by liquid scintillation counting. Lipid efflux is expressed as the ratio of counts in medium to total counts. Specific efflux rates were calculated by subtracting efflux rates in the presence of control bovine serum albumin from the efflux rates in the presence of the lipid acceptors apoAI or HDL.

RESULTS AND DISCUSSION
ZNF202 Binds Specific ABCA1 and ABCG1 Promotor Sequences-It has been proposed that ZNF202 functions as a transcriptional repressor of various genes related to lipid trans-port and lipoprotein metabolism (9). All these genes share in their promotor regions a common repetitive GnT motif that binds in vitro ZNF202. To identify additional targets of ZNF202, we analyzed the promotor sequences of the human ABCA1 and ABCG1 genes for the presence of GnT motifs. Both proteins act as transmembrane lipid transporters (6,15). In particular ABCA1 is a key regulator of the plasma HDL pool size and mediates the cellular efflux of cholesterol and phospholipids. Mutations of the ABCA1 gene, which lead to genetic HDL deficiency syndromes characterized by the almost complete absence of plasma HDL, further underscore the pivotal role of ABCA1 in reverse cholesterol transport and, ultimately, hypoalphalipoproteinemia (2-4).
As depicted in Fig. 1, we found GnT motifs within the promotor region of ABCA1 at positions Ϫ210/Ϫ229 and ABCG1 at positions Ϫ552/Ϫ572. To examine the ability of ZNF202 to bind these DNA sequences, gel shift assays were performed with rabbit reticulocyte lysates expressing in vitro full-length ZNF202m1 protein and radiolabeled oligonucleotides encoding specific promotor fragments (Fig. 1, A and B). The published consensus ZNF202 DNA binding sequence, as found in the apoAIV promotor at positions Ϫ265/Ϫ243 and represented by an oligonucleotide, led to a specific and dose-dependent band shift after incubation with the ZNF202m1 lysate (Fig. 1C, lane  1-3) (9). Oligonucleotides reflecting the ABCG1 and ABCA1 ZNF202 DNA binding motifs revealed similar migratory gel properties after incubation with the ZNF202m1 lysate. The distinct intensities of the detected DNA-protein complexes may reflect individual affinities of these promotor fragments for ZNF202m1. The addition of KRAB polyclonal antibody that recognizes epitopes between amino acid position 177 and 329 of ZNF202m1 to the incubation mixture affected the affinity of ZNF202 for the apoAIV, ABCG1, or ABCA1 promotor fragment (Fig. 1C, lane 4). The addition of unlabeled oligonucleotides in a 50-fold excess competitively abolished the band shift (Fig. 1,  lane 5), providing further evidence for the specific interaction of ZNF202m1 with its target sequences.
ZNF202m1 Represses ABCA1 and ABCG1 Promotors-To investigate the effect of ZNF202 on the transcriptional activity of both ABC transporter promotors, luciferase reporter gene assays were performed. As shown in Fig. 2, transfection of a reporter gene construct containing the promotor region of either the apoAIV, ABCA1, or ABCG1 gene together with various amounts of ZNF202m1 expression vector led to a dose-dependent inhibition by up to 80% of all promotor activities (Fig.  2, black bars). Interestingly, a truncated ZNF202 construct lacking the SCAN domain lost the ability to repress the promotor activities (Fig. 2, crossed bars). The SCAN domain of ZNF202 has been shown to mediate selective oligomerization with itself or other SCAN domain-containing proteins such as ZNF191 or SCAN domain protein 1 (SDP1) (11). The state of ZNF202 oligomerization may therefore impact the transcriptional activity of the ZNF202m1 protein.
To resolve the discrepancy between the relatively low binding affinity of ZNF202 for the ABCA1 promotor sequence (Fig.  1) and the high capacity to repress the promotor activity (Fig.  2), we further investigated the interaction of ZNF202 with the ABCA1 promotor. Besides its affinity for GnT motifs, ZNF202 shares with other zinc finger proteins the ability to bind to GC boxes (9). The ABCA1 promotor harbors two GC boxes at positions Ϫ91 and Ϫ157 that bind the zinc finger proteins Sp1 and Sp3 (M. Porsch-Ö zcü rü mez, data not shown). We generated reporter gene constructs with truncated ABCA1 promotor inserts and mutated putative ZNF202 binding sites (Fig. 3). Surprisingly, we did not observe any loss in ZNF202-mediated repression with a reporter construct containing only the Ϫ79/ϩ 224 ABCA1 region and, thus, lacking all potential ZNF202 binding sites. ZNF202 may therefore interact with other elements of the transcriptional complex.
KRAB domain-mediated transcriptional repression has been reported to result from interference with the TATA box-dependent basal transcription machinery (18,19). The Ϫ79/ϩ 224 ABCA1 reporter gene construct with a mutated TATA box at

FIG. 2. ZNF202 inhibits ABCA1 and ABCG1 promotor activity.
Luciferase assays performed with ABCA1 (Ϫ919/ϩ224), ABCG1 (Ϫ2912/ϩ50), apolipoprotein AIV (Ϫ718 to ϩ30), and promotorless pGL3 reporter gene constructs. HepG2 cells were co-transfected with 2 g of reporter gene construct, 1 g of pSV ␤-galactosidase vector, and ZNF202m1 expression vector in various amounts as indicated (black bars) or SCAN domain-truncated ZNF202 vector (crossed bars). Cells were analyzed for promotor activity 24 h post-transfection. Results are presented as inhibition of luciferase activity (mean Ϯ S.D. of triplicate measurements of three independent experiments) in comparison to cells transfected with empty control vector. Luciferase activity was normalized for ␤-galactosidase activity and protein concentrations. Absolute luciferase levels were 119-fold for ABCA1, 31-fold for apolipoprotein AIV, and 6-fold for ABCG1 above control vector level. position Ϫ24 markedly decreased promotor activity but also abolished ZNF202-mediated repression (Fig. 3). Identical results were obtained with the ϩ12/ϩ224 construct containing only exon 1 of ABCA1. These data indicate that (i) the TATA box is of functional importance for ABCA1 expression and (ii) transcriptional repression by ZNF202 is likely mediated by a similar mechanism as reported for other Krü ppel-type zinc finger proteins.
ZNF202 Binding to the Transcriptional Corepressor KAP1-ZNF202 shares a number of similar amino acids with the proposed consensus sequence for the KRAB homology domain, which was derived from the KOX1 gene (12). Within the KRAB domain, the amino acid identity between ZNF202 and KOX1 is 42%, and the similarity is 64%, as aligned with the BLAST algorithm (20). KOX1 has been shown to associate via its KRAB domain with KAP1, a 97-kDa nuclear phosphoprotein with all the hallmarks of a universal corepressor. The RING finger, B boxes (␤1 and ␤2) and a coiled coil region at the amino terminus collectively constitute the KRAB interaction, or RBCC domain (12). We therefore analyzed the ability of ZNF202 to bind KAP1. Indeed, in vitro radiosynthesized KAP1 protein associated with the ZNF202 KRAB domain, as shown by affinity purification studies with GST-tagged ZNF202 fusion proteins (Fig. 4A). To evaluate the specificity of the ZNF202-KAP1 interaction, a double amino acid substitution (DV Ͼ AA) that has been previously shown to functionally inactivate the KOX1 protein was reproduced in the ZNF202 KRAB domain (12,21). The wild type and mutated ZNF202m1 cDNA sequences were radiosynthesized in vitro and analyzed for their ability to bind KAP1. Resin-coupled KAP1 immunocomplexes were isolated from HeLa cells and, after incubation with ZNF202-expressing rabbit reticulocyte lysates, was analyzed by SDS-PAGE and fluorography (Fig. 4B). KAP1 associated readily with wild type ZNF202, whereas the DV Ͼ AA amino acid substitution within the KRAB domain abolished, as anticipated, the association of ZNF202 with KAP1.
The apolipoprotein AIV and ABCA1 genes both contain a TATA box that may be targeted by KRAB domain-mediated signaling (22). The underlying mechanism of ZNF202-induced transcriptional repression of the ABCG1 gene, which lacks a TATA box, needs to be further elucidated. The relatively low constitutive activity of the ABCG1 promotor may explain the possibility that minor ZNF202 alterations affect the pyrimidine-rich initiator element of ABCG1 (13). It also remains unclear how specificity of ZNF202 repression is defined. Pengue and Lania (19) provide evidence that KRAB domainmediated transcriptional repression is not caused by a general FIG. 3. Repression of ABCA1 promotor activity by ZNF202 does not depend on the specific binding to the (؊229/؊210) GnT motif or GC boxes. 1 ϫ 10 6 HepG2 cells were transfected with 1 g of pSV␤-galactosidase and 2 g of reporter gene vectors for the ABCA1 promotor encompassing the full-length (Ϫ919/ϩ224) wild-type or the fragmented form. The GnT motif, two GC boxes, and the TATA box were mutated as indicated in the vector scheme. Transfected cells were cultured for 24 h before being assayed for luciferase activity. A representative out of three independent experiments is shown. Luciferase activity was normalized for ␤-galactosidase activity and protein concentrations. Results are expressed as multiples of promotorless control vector and indicated as the mean Ϯ S.D. of triplicate measurements. FIG. 4. KAP1 binding to the KRAB domain of ZNF202 is modulated by the oligomerization state of the adjacent SCAN domain. A, affinity purification of KAP1 with ZNF202. Rabbit reticulocyte lysates expressing in vitro produced radiolabeled KAP1 were extracted with GST or GST fusion proteins of ZNF202 containing either the KRAB or the SCAN domain of ZNF202. The precipitates were subsequently analyzed by SDS-PAGE and autoradiography. Molecular mass markers are indicated at the right side in kilodalton units (kDa). B, affinity purification of ZNF202 with KAP1. Resine-coupled KAP1 immunocomplexes were incubated with rabbit reticulocyte lysates equally expressing wild-type or mutated ZNF202 cDNA templates as shown below the gel. A KAP1 pre-bleed immunocomplex (PB) was used as control. The purified immunocomplexes were subsequently analyzed by SDS-PAGE and autoradiography. C, effect of SCAN domain oligomerization on KAP1 binding. KAP1 immunocomplexes were isolated from HeLa cell lysates and tested for their ability to purify in vitro radiosynthesized ZNF202m1 from rabbit reticulocyte lysates in the presence or absence of added SDP1 or mFPM315 protein.
and unspecific inhibition of the RNA polymerase II machinery but strongly depends on the specific arrangement of basal promotor elements.
The amino-terminal SCAN domain of ZNF202 has been shown to readily oligomerize with SDP1 (11). We investigated therefore the possibility that SCAN domain-mediated protein oligomerization may affect the ability of the juxtaposed KRAB domain to interact with KAP1. ZNF202m1, SDP1, and mFPM315, a SCAN domain-encoding protein without detectable affinity for ZNF202 and, therefore, used as negative control (11), were expressed in vitro in rabbit reticulocyte lysates in the presence of a sulfur radiolabel. A HeLa cell-isolated KAP1 immunocomplex readily purified ZNF202 in the presence or absence of mFPM315 (Fig. 4C). However, in the presence of SDP1, ZNF202 lost its ability to bind the KAP1 immunocomplex. The state of SCAN domain oligomerization may therefore modulate the ability of ZNF202 to recruit the corepressor KAP1 and to mediate transcriptional repression. Although zinc finger proteins of the Cys 2 -His 2 class are reported to bind DNA in monomeric form, our results suggest that SCAN domain-containing zinc finger proteins require specific homo-or heterodimerization for DNA binding and transcriptional modulation. The existence of SDP1 as a gene encoding an isolated SCAN domain as well as the truncated ZNF202m3 splice variant, which also contains only a SCAN domain, further highlights the modulating impact of SCAN domain interactions (11,23).

ZNF202 Affects Induction of ABCA1 and Cellular Lipid
Efflux-To further confirm the role of ZNF202 as a transcriptional repressor of ABCA1, we determined the ABCA1 mRNA abundance by Northern blot analysis using RNA preparations from HepG2 human hepatoma cells transfected with increasing amounts of expression vector encoding either ZNF202m1 (Fig.  5A, upper panel) or a ZNF202 fragment lacking the SCAN domain (Fig. 5A, lower panel). In accordance with the data from the promotor assays, the endogenous expression of ABCA1 mRNA was significantly decreased in ZNF202m1-transfected HepG2 cells, whereas mRNA levels were unaffected upon transfection with a ZNF202-truncated protein.
Because KRAB domain proteins were shown to repress both basal and activated promotor activity, we investigated the effect of ZNF202 under conditions that induce ABCA1 expression (Fig. 5B). Oxysterols such as 20(S)-OH-cholesterol in combination with 9-cis-retinoic acid have been shown to strongly induce ABCA1 expression by interaction with LXR/RXR heterodimers via a DR4 element (5); the physiological relevance of these compounds still remains unclear. Incubation with 10 M 20(S)-OH-cholesterol and 10 M 9-cis-retinoic acid led to a 3.5-fold induction of ABCA1 promotor activity in RAW264.7 rat macrophages (Fig. 5B). Transfection of these cells with 4 g of ZNF202m1 expression vector almost completely abolished oxysteroldependent induction of ABCA1. These data underscore the strong inhibitory capacity of ZNF202, which is likely driven FIG. 5. Effects of ZNF202 on cellular ABCA1 induction and lipid efflux. A, ABCA1 mRNA expression is repressed by ZNF202. HepG2 cells transfected with increasing amounts of expression vector for ZNF202m1 (upper panel) or for ZNF202 with a truncated SCAN domain (lower panel) were subjected after 48 h to Northern blot analysis. The blots were probed for ABCA1 and ubiquitin expression. Lane 1, 2 g of empty control vector; lanes 2, 3, and 4, 0.5, 1, or 2 g of ZNF202m1 and ZNF202⌬SCAN vectors. B, ZNF202 abolishes oxysterol-induced ABCA1 promotor activity. HepG2 cells were co-transfected with 2 g (Ϫ919/ϩ224) ABCA1 reporter gene vector, 1 pSV ␤-galactosidase vector and either 4 g of ZNF202m1 expression vector or an empty control vector (Mock). Four hours after transfection, cells were stimulated with 10 M 20-OH(S)-cholesterol and 10 M 9-cis -retinoic acid (9CRA) dissolved in 0.2% (v/v) ethanol. Cells were harvested 24 h after transfection and assayed for luciferase activity. Results are given as ϫ-fold of the promotorless control vector. C, ZNF202 reduces apoAI and HDL 3 -mediated phospholipid and cholesterol efflux. Stably transfected RAW264.7 cells were radiolabeled and loaded for 24 h with 40 g/ml enzymatically modified low density lipoprotein in the presence of 10 M 20-OHcholesterol and 10 M 9-cis-retinoic acid or 0.02% (v/v) ethanol. Subsequently, cells were washed and chased for 17 h with either 100 g/ml HDL 3 protein, 10 g/ml purified apoAI, or 0.2% bovine serum albumin. Results are illustrated as the mean Ϯ S.D. from measurements of four wells. *, p Ͻ 0.05, comparing the specific efflux of ZNF202m1 and mock-transfected cells in independent samples by t test.
by a mechanism that targets crucial elements of transcriptional activation.
Finally, to demonstrate the functional relevance of ZNF202 on cellular phospholipid and cholesterol efflux, RAW264.7 cells that stably overexpress ZNF202m1 were generated. Because ABCA1 and ABCG1 expression are strongly induced by oxysterols (24,25), we compared efflux rates under basal conditions and during incubation with 20(S)-OH-cholesterol and 9-cis -retinoic acid. As shown in Fig. 5C , apoAI and HDL 3 -mediated lipid efflux markedly depend on the stimulation with oxysterols. ZNF202-overexpressing cells revealed significantly reduced specific phospholipid efflux (upper panel) and cholesterol efflux (lower panel) rates as compared with mock-transfected cells. Interestingly, the strongest suppression was observed in HDL 3 -mediated cholesterol efflux. Since recent studies demonstrated that apoAI and not HDL 3 is the lipid acceptor for the ABCA1 transporter, one may conclude that ZNF202 modulates additional genes with selective cholesterol transport activity independent of the ABCA1 efflux pathway (26).

ZNF202 Expression Profiling in Various Human Tissues-
The original description of ZNF202 as a testis-specific zinc finger protein (10) was based on expression studies with limited tissue samples. Using an RNA master blot derived from 75 distinct human tissues, we observed that ZNF202 is highly expressed not only in testis tissue but also in various other tissues such as uterus, brain, intestine, fetal tissues, bone marrow, and leukocytes, which support a potential role for ZNF202 in the transcriptional regulation of ABCA1 in vivo (Table I). Since ABCA1 and ABCG1 are induced in human monocytes during phagocytic differentiation and subsequent lipid loading using modified low density lipoprotein (6,14), we tested under these conditions the expression of ZNF202 in human monocytes. Preliminary results indicate that ZNF202 expression is down-regulated during monocyte differentiation and repressed by cholesterol loading, thus demonstrating inverse regulation of ABCA1/ABCG1 and ZNF202 (C. Schumacher; data not shown).
The identification of additional ZNF202-interacting proteins and the elucidation of the transcriptional regulation of ZNF202 will help to understand the role of ZNF202 in modulating the expression of ABCA1 and ABCG1 in vivo. It will be of special interest to identify genotypic sequence variations in regulatory motifs and functional domains of the ZNF202 gene. Abnormalities in ZNF202-oligomerizing proteins may also contribute to genetically based hypoalphalipoproteinemia and, thereby, further confirm the role of chromosome 11q23 as a HDL susceptibility locus.