The SCAN Domain Mediates Selective Oligomerization*

The SCAN domain is described as a highly conserved, leucine-rich motif of approximately 60 amino acids found at the amino-terminal end of zinc finger transcription factors. Although no specific biological function has been attributed to the SCAN domain, its predicted amphipathic secondary structure led to the suggestion that this domain may mediate protein-pro-tein associations. A yeast two-hybrid screen identified members of two SCAN domain protein families that interact with the SCAN domain of the zinc finger protein ZNF202. The interacting ZNF191 protein represents the family of SCAN domain-containing zinc finger proteins, whereas the novel SDP1 protein establishes a new family of genes that encode an isolated SCAN domain. Isolated SCAN domain proteins may form asymmetric ho-modimers in solution. Biochemical binding studies confirmed the associations of ZNF191 and SDP1 with ZNF202 and established the SCAN domain as a selective hetero- and homotypic oligomerization domain. SCAN mediated protein associations might therefore represent a new regulatory mechanism of transcriptional activity. The SCAN or leucine-rich domain, originally identified by its homology with similar elements in several zinc finger transcription factors, consists of approximately 60 amino acids and is rich in leucine and glutamic acid residues (1). Most SCAN domain sequences are linked to Cys 2 -His 2 zinc finger motifs through their carboxyl-terminal end. Although the function of the SCAN domain has

The SCAN domain is described as a highly conserved, leucine-rich motif of approximately 60 amino acids found at the amino-terminal end of zinc finger transcription factors. Although no specific biological function has been attributed to the SCAN domain, its predicted amphipathic secondary structure led to the suggestion that this domain may mediate protein-protein associations. A yeast two-hybrid screen identified members of two SCAN domain protein families that interact with the SCAN domain of the zinc finger protein ZNF202. The interacting ZNF191 protein represents the family of SCAN domain-containing zinc finger proteins, whereas the novel SDP1 protein establishes a new family of genes that encode an isolated SCAN domain. Isolated SCAN domain proteins may form asymmetric homodimers in solution. Biochemical binding studies confirmed the associations of ZNF191 and SDP1 with ZNF202 and established the SCAN domain as a selective hetero-and homotypic oligomerization domain. SCAN mediated protein associations might therefore represent a new regulatory mechanism of transcriptional activity.
The SCAN or leucine-rich domain, originally identified by its homology with similar elements in several zinc finger transcription factors, consists of approximately 60 amino acids and is rich in leucine and glutamic acid residues (1). Most SCAN domain sequences are linked to Cys 2 -His 2 zinc finger motifs through their carboxyl-terminal end. Although the function of the SCAN domain has not yet been elucidated, the predicted amphipathic structure of the domain led to the suggestion that SCAN box elements have the capacity to interact with other proteins, in particular with components of the transcriptional machinery (1).
The zinc finger protein ZNF202 1 is expressed in two common splice variants, here referred to as m1 and m3 (2). Whereas the m1-splice form encodes a full-length protein of 648 amino acids with a SCAN box, a KRAB repression domain, and eight Cys 2 -His 2 zinc finger motifs, the 133 amino acid product of the m3-splice form encompasses only the SCAN domain. 2 These splice forms are conserved in the murine ZNF202 homolog, suggesting that the SCAN motif itself is an independent functional domain. 3 The existence of other genes that encode SCAN elements as an isolated structural feature is further demonstrated by the recent identification of the murine Leap1/PCG-2 gene (GenBank TM accession AF106473) (3).
In order to study the function of the SCAN domain, we performed an extensive yeast two-hybrid screen for the identification of SCAN binding proteins. The screen, supported by biochemical association studies, suggested that SCAN motifs have the ability to associate selectively with each other. Furthermore, the first human gene encoding an isolated SCAN domain was identified. The formation of SCAN domain-mediated protein complexes may therefore modulate the biological function of transcription factors.
SDP1 cDNA Analysis-A full-length cDNA for SDP1 was obtained by rapid amplification of cDNA ends experiments from Marathon cDNA libraries (CLONTECH). The cDNAs were sequenced on both strands by fluorescent-labeled dye primer sequencing on ABI 377 sequencers (Applied Biosystems, Inc.). The assembled sequence was confirmed by sequencing a representative EST cDNA (GenBank TM accession number N75095; IMAGE Consortium cDNA clone 284448 obtained from Genome Systems).
Northern Blot Analysis-Labeled SDP1 probes were synthesized in vitro in the presence of [␣-32 P]dCTP (Amersham Pharmacia Biotech) from a cDNA template encompassing the open reading frame using random primers and Klenow enzyme (Promega) and were used to probe multiple human tissue Northern blots (CLONTECH). Hybridization analysis was carried out in Quickhyb solution (Stratagene) at 65°C and * 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.
The visualized by autoradiography. The blots were then stripped of radioactivity and reprobed with a 32 P random prime-labeled glycerol-3-phosphate dehydrogenase cDNA probe (CLONTECH) to confirm equal loading.
GST Fusion Proteins-The cDNA fragments encoding amino acids 1-199 (Z1) and 177-329 of the ZNF202-m1 open reading frame were subcloned directionally into the pGEX-4T bacterial expression vector (Amersham Pharmacia Biotech). A BamHI-EcoRI fragment of the ZNF191 cDNA which includes the SCAN domain was subcloned into pGEX1. The entire open reading frame of the SDP1 cDNA was subcloned directionally into pGEX-4T. All constructs were sequenced through the junctions to verify sequence fidelity and orientation. Expression and purification of the GST fusion proteins was done as described (7). Purity and integrity of the fusion proteins was assessed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and Coomassie Blue staining (8). 35 S-Labeling of GST-ZNF202[SCAN] was performed as detailed (9). Briefly, 200 ml of fresh LB medium with ampicillin (100 g/ml) is inoculated with an overnight bacterial culture and grown for 3 h at 37°C. The bacterial cell pellet was resuspended in 50 ml of prewarmed Dulbecco's modified Eagle's medium (Life Technologies, Inc.) with 2 mM glutamate and ampicillin (100 g/ml) but lacking cysteine and methionine and was induced with 0.2 mM isopropyl-␤-Dthiogalactopyranoside (Sigma) for 10 min. Tran 35 S-label (ICN) was then added (5 mCi), and the incubation was continued for 2 h. The purification and elution was according to the standard procedure, and the eluted proteins were dialyzed overnight against 10 mM Tris-HCl, pH 7.5. The GST fusion protein was thrombin-cleaved according to the procedure of Smith (10). Briefly, bead-coupled fusion protein was resuspended in cleavage buffer containing 50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 2.5 mM CaCl 2 and treated with 0.1% w/w thrombin (Sigma) at room temperature for 1 h. After centrifugation to extract the beads, the supernatant was analyzed by SDS-PAGE.
Antibodies-The ZNF202 antiserum was generated by immunizing rabbits with the GST-ZNF202[KRAB] protein fragment that encompasses amino acids 177-329 of the ZNF202-m1 open reading frame. Antigen-specific antibodies were isolated from the polyclonal antiserum by immunoaffinity purification using an UltraLink immobilization column (Pierce). Anti-Xpress epitope mouse monoclonal antibodies were obtained from Invitrogen.
Transient Cell Transfections and Immunoprecipitations-A polymerase chain reaction amplicon of the ZNF202-m1 open reading frame was subcloned into the expression vector pBIND (Promega) to afford pGal4-ZNF202-m1. The SDP1 open reading frame was directionally subcloned into the pcDNA3.1/His vector (Invitrogen) which encodes an aminoterminal Xpress epitope. HEK 293 cells (2.5 million in a 100-mm dish) were transfected with expression plasmids using LipofectAMINE Plus (Life Technologies, Inc.) per manufacturer's recommendations. Control plates were transfected with green fluorescent protein and were inspected by fluorescent microscopy after 24 h to confirm Ͼ40% transfection efficiency. After 48 h, cells were removed from the plates, were pelleted by centrifugation, and then were lysed in 0.25 ml of lysis buffer that contained 1.0% Triton X-100 in Tris-HCl-buffered saline in the presence of sodium azide and a protease inhibitor mixture (Roche Molecular Biochemicals).
For each precipitation, 50 l of cell lysate was mixed with 1 g of monoclonal antibody against the Xpress tag (Invitrogen) or 2 l of affinity purified anti-ZNF202 KRAB domain antibody in a total volume of 0.25 ml of binding buffer (0.1% Triton X-100 in Tris-buffered saline with sodium azide). Immune complexes were permitted to form for 2 h at 4°C on an orbital mixer. Protein-A Sepharose slurry (40 l, Amersham Pharmacia Biotech) was added and rocked for an additional hour. The beads were pelleted by centrifugation; the supernatant was removed, and the pelleted beads were washed three times with 1 ml of binding buffer. The beads were finally mixed with 40 l of reducing loading buffer and heated prior to SDS-PAGE on 4 -12% gradient gels. Reference samples (10 l) of the lysates were loaded on each gel to confirm expression of the constructs and to permit an estimation of precipitation efficiency. Immunoblotting was performed under standard conditions using enhanced chemiluminescence detection (Amersham Pharmacia Biotech).
Gel Filtration Chromatography, Sedimentation Equilibrium Analysis, and Chemical Cross-linking of the ZNF202-m3 Protein-A cDNA amplicon encoding the ZNF202-m3 splice form was subcloned into the NdeI-cloning site of the pET11a expression vector (Novagen). The expressed SCAN protein of 133 amino acids was purified over a cation exchange column using a POROS HS20 column (Perkin-Elmer) and analyzed by fast protein liquid chromatography. The analytical size exclusion chromatography was performed on a Superdex-75 HR 10/30 column (Amersham Pharmacia Biotech) equilibrated with a buffer containing 50 mM HEPES, pH 7.5, 100 mM KCl, and 2 mM dithiothreitol. A purified protein sample of approximately 1.0 mg in 100 l of column equilibration buffer was loaded on the column and was eluted at a flow rate of 0.5 ml/min at 4°C. Fractions of 0.5 ml were analyzed by absorbance at 280 nm and by the Bradford method for protein detection (Bio-Rad). The column was calibrated with bovine serum albumin (67 kDa), ovalbumin (43 kDa), chymotrypsinogen (25 kDa), and ribonuclease A (13.7 kDa) as protein standards (Amersham Pharmacia Biotech).
For the postcentrifugation measurement of sedimentation equilibrium, 1.0 mg of purified protein sample was layered on top of a 5-20% linear sucrose gradient established in equilibration buffer. As molecular weight standards, aldolase (154 kDa) and ovalbumin (43 kDa) were analyzed onto a second gradient, whereas bovine serum albumin (67 kDa) was eluted onto a third gradient. Centrifugation was carried out in a Beckman SW40 rotor at 39,000 rpm for 23 h at 4°C. Fractions of 0.4 ml were collected from the bottom to the top of the gradient, and the presence of protein was analyzed by the Bradford method.
For chemical cross-linking experiments, a purified ZNF202-m3 sample of 1.6 mg/ml (0.1 mM protein) concentration was incubated with 0.4 or 2 mM bis(sulfosuccinimidyl)suberate (Pierce) for 30 min at room temperature. The cross-linking reaction was subsequently quenched with 100 mM Tris-HCl, pH 7.5, and analyzed by denaturing SDS-PAGE followed by Coomassie Blue staining. A recombinant HtrA protease (a dodecamer in solution) and bovine serum albumin (a monomer) were used as cross-linking control samples in parallel reactions (11).
In Vitro Expression of SCAN Domain Proteins-DNAs encoding the m1-and m3-splice form of ZNF202, SDP1, ZNF191 along with two SCAN domain encoding cDNA clones (mLD5-1, the murine homologue of LD5-1 or ZNF192: GenBank TM accession number NM006298; Zfp110, the murine neurothrophin receptor interacting factor: Gen-Bank TM accession number MMU242914; mFPM315, the murine homologue of FMP315: GenBank TM accession number O14978) were directionally subcloned into the mammalian expression vector pcDNA3.1(Ϫ) (Invitrogen) or pBluescript II SK(ϩ) (Stratagene). The cDNA templates were subsequently expressed in vitro in rabbit reticulocyte lysates in the presence of [ 35 S]methionine using the TNT T7 Quick Transcription/Translation System (Promega) according to the manufacturer's procedure.
Affinity Purification of SCAN Domain Proteins-GST fusion proteins (1.0 g) were incubated overnight at 4°C with 2.0 l of rabbit reticulocyte lysate sample expressing a SCAN domain encoding cDNA template supplemented with HNTG buffer to a final volume of 200 l. HNTG buffer contained 20 mM HEPES, pH 7.4, 150 mM NaCl, 0.1% Triton X-100, and 10% glycerol. 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).

RESULTS
A Yeast Two-hybrid Screen Identifies Interacting SCAN Domain Proteins-A yeast two-hybrid approach was used to search for proteins that interact with the SCAN domain in vivo. The amino-terminal end of ZNF202 (amino acids 1-199) was fused in frame to the DNA binding domain of Gal4 and was used as bait (Z1) to screen independently brain, kidney, liver and B-cell Gal4 activation domain libraries (Fig. 1A). The screen resulted in the isolation of 48 interacting cDNA clones, 42 of which encoded a SCAN homology domain (Fig. 1B). Five cDNA clones corresponded to ZNF191, 28 to SDP1 (SCAN domain protein 1), 4 to SDP2, and 5 to SDP3. Whereas ZNF191 encodes a SCAN domain-containing zinc finger protein, the novel SDP1 cDNA of 0.9 kilobases was found to encode an isolated SCAN box without adjacent zinc finger motifs. The presence or absence of zinc finger motifs in SDP2 and SDP3 is unknown since the respective cDNAs are incomplete.
An additional yeast two-hybrid screen was performed with a SDP1 bait and a kidney library that resulted in the isolation of five SCAN domain encoding cDNA clones (Fig. 1B). One cDNA clone corresponded to SDP1, indicating self-association, three clones represented SDP2, and one cDNA clone was identical to GenBank TM EST N29000 encoding SDP4. All individual interactions were subsequently confirmed by domain swap experiments (data not shown).
In order to further define the ZNF202 binding domain, additional ZNF202 Gal4 activation domain fusion constructs were directly probed with SCAN domains derived from ZNF202 (Z1), ZNF191, or SDP1 (Fig. 1A). The directed yeast mating reactions demonstrated that amino acids 1-171 of ZNF202 (Z7), in contrast to 1-44 (Z6), were sufficient for the interaction (Fig.  1C). The directed mating reactions indicated also the ability of ZNF202 to self-associate. In comparison, two KRAB domainencompassing baits of ZNF202 (Z8 and Z10) did not interact confirming a sequence-specific association between these SCAN domain proteins.
The yeast two-hybrid screen suggested associations between SCAN domain proteins with two distinct motif features. ZNF191 or ZNF202 represent SCAN domain-containing zinc finger proteins, whereas SDP1 is characterized, similar to the m3-splice form of ZNF202, by an isolated SCAN domain. Tissue mRNA analysis showed that the SDP1 is ubiquitously highly expressed ( Fig. 2A). The homology between the SCAN domains in SDP1, ZNF191, and ZNF202 is high with 47% identical residues over an alignment of 60 amino acids (Fig. 2B). Furthermore, the SCAN domains align with the conserved sequence residues published by Castillo et al. (3).
Hetero-and Homotypic Binding Ability of ZNF202-In order to confirm the binding interactions of ZNF202 with ZNF191 and SDP1, we generated GST fusion proteins encompassing the ZNF202 KRAB domain or the respective SCAN boxes of these proteins. In an affinity purification study, we tested the ability of GST-ZNF202 fusion fragments to extract in vitro radiosynthesized SDP1 or ZNF191 protein from reticulocyte lysate preparations (Fig. 3, A and B). The SDS-PAGE showed that the GST-ZNF202[SCAN] protein construct, in contrast to the GST-ZNF202[KRAB] or the GST tag itself, precipitated the radiolabeled SDP1 or ZNF191 protein. The ability of ZNF202 to self-associate as detected in the yeast two-hybrid screen was further validated by affinity purification studies and far Western blotting (Fig. 3, C and D). The affinity purification study confirmed the ability of the GST-ZNF202[SCAN] protein construct to precipitate radiolabeled ZNF202-m3 protein from cell lysates (Fig. 3C). In a far Western blot approach, GST-linked ZNF202 protein fragments were probed with an [ 35 S]methionine/cysteine metabolically labeled GST-ZNF202[SCAN] protein (Fig. 3D). The radiolabeled probe associated specifically with its unlabeled counterpart. The GST protein tag was not responsible for the observed protein complex formation. Likewise, far Western membranes of SCAN domain-encompassing GST fusion proteins probed with a radiolabeled GST-ZNF202[SCAN] fusion protein demonstrated binding of ZNF202 to SDP1, ZNF191, or ZNF202 itself (Fig. 3E). In order to test the possibility that the observed associations were facilitated by the GST protein tag, a GST-cleaved radiolabeled ZNF202[SCAN] protein was shown to maintain the association with the SDP1 fusion protein (Fig. 3F).
Because we could demonstrate association between SCAN domain-containing proteins in several experimental systems in vitro, we next sought to determine whether SCAN domaincontaining proteins interact in a cellular milieu. Various expression constructs of ZNF202 and SDP1 with epitope tags were employed in co-immunoprecipitation/immunoblot experiments in order to demonstrate intracellular associations. Fig.  4A shows an anti-ZNF202[KRAB] immunoblot of lysates from cells transfected with pGal4-ZNF202-m1 alone or with pGal4-ZNF202-m1 plus pXpress-SDP1. Lysates from these cells contain an appropriately sized (ϳ110 kDa) immunoreactive band (3rd and 4th lanes). The anti-Xpress immunoprecipitation of the pGal4-ZNF202-m1/pXpress-SDP1 cotransfection shows a robust band (2nd lane), indicative of intracellular association between ZNF202 and SDP1, whereas a control antibody yielded no band (1st lane). Fig. 4B represents the reciprocal of Fig. 4A. Instead of coprecipitating ZNF202 with antibodies to tagged forms of SDP1, Xpress-SDP1 was precipitated with anti-ZNF202[KRAB] antibody (2nd lane). Thus, in lysates from appropriately transfected cells, ZNF202 precipitates SDP1 and SDP1 precipitates ZNF202 precisely as was seen in affinity purification studies and in far Western experiments.
The ZNF202-m3 Protein Forms an Asymmetric Homodimer-In order to determine the stoichiometry of the ZNF202 SCAN protein complex in solution, a bacterially ex- FIG. 1. The yeast two- pressed ZNF202-m3 protein was purified over a cation exchange column and subjected to size exclusion chromatography. A fast protein liquid chromatography experiment using a prepacked Superdex-75 HR column eluted the SCAN protein in a 52-kDa molecular size fraction as calculated by plotting the logarithm of the molecular mass of the protein standards against the elution volume (Fig. 5A). In contrast, an SDS-PAGE analysis of this protein separated the denatured and thus monomeric form at 15-kDa molecular size as detected by Coomassie Blue gel staining (Fig. 5A, inset). In addition, the postcentrifugation sedimentation equilibrium of the ZNF202-m3 protein through a 5-20% sucrose gradient was analyzed. As shown in Fig. 5B, the sedimentation profile of the m3 protein showed a well defined peak at approximately 48 kDa in the presence of 100 mM KCl. Thus, the migratory properties of the ZNF202-m3 protein subjected to size exclusion chromatogra- phy and sedimentation analysis suggest that the SCAN domain of ZNF202 forms either a trimeric complex or a modestly asymmetric dimer in a native environment. Chemical cross-linking experiments, however, did not reveal a trimeric complex but rather a dimeric species (Fig. 5C). Thus, the combination of gel filtration, sedimentation equilibrium, and cross-linking experiments of the ZNF202-m3 protein support the presence of an asymmetric dimeric protein.
Selective SCAN Domain Association Abilities of ZNF202, SDP1, and ZNF191-In a preliminary attempt to evaluate the selectivity of the SCAN domain-mediated protein associations, we transcribed and translated in vitro cDNA templates for ZNF202-m1, ZNF191, SDP1, and three unrelated SCAN domain-containing zinc finger proteins indicated as mLD5-1, Zfp110, and mFPM315. The cell lysates were incubated with GST-fused SCAN box sequences derived from ZNF202, SDP1, or ZNF191, and the recaptured fusion proteins were analyzed for binding in vitro produced radiolabeled protein (Fig. 6). All three GST fusion proteins associated strongly with radiolabeled SDP1. Conversely, the GST-SDP1 construct readily precipitated ZNF202, SDP1 itself, ZNF191, mLD5-1, and Zfp110. Therefore, the SCAN domain encoded by SDP1 revealed the highest association affinity that was also observed in the far Western experiment (Fig. 3C). Furthermore, SDP1 showed the ability to self-associate. which confirmed the interaction detected in the yeast two-hybrid screen. The fusion proteins of ZNF202 and ZNF191 revealed similar affinities for the radiolabeled proteins yet differed in their affinity for mLD5-1 and Zfp110. Whereas the ZNF202 fusion protein interacted weakly with mLD5-1, the ZNF191 fusion protein precipitated Zfp110. In comparison, the GST-SDP1 construct bound both mLD5-1 and Zfp110. None of the fusion proteins interacted with mFPM315. Therefore, the individual affinity of the fusion proteins to the unrelated SCAN proteins suggests the ability of SCAN domains to associate selectively with each other.

DISCUSSION
The SCAN domain is a sequence motif of unknown function that is common to zinc finger proteins. We provide evidence that the SCAN domain mediates selective protein oligomerization. A yeast two-hybrid screen identified homotypic and heterotypic interactions between SCAN domain-containing proteins. Biochemical binding experiments confirmed the interactions in vitro as well as in vivo and revealed selectivity in the SCAN domain-mediated protein associations as individual SCAN domains exhibited distinct differences in their ability to bind to other SCAN domain proteins. The structural components within a SCAN domain protein that contribute to binding selectivity and affinity remain to be elucidated. Sequence alignments of SCAN domains show a high degree of amino acid conservation that suggests that relatively few amino acids contribute to binding specificity (3). A hydrodynamic analysis of the ZNF202-m3 protein by gel filtration chromatography and postcentrifugation measurement of sedimentation equilibrium suggests the formation of a trimeric or asymmetric dimeric protein complex in solution. Dynamic light scattering investigations of the ZNF202-m3 protein in solution could not distinguish between these two oligomerization states. 4 In vitro chemical cross-linking experiments, however, indicated a dimeric protein in identical buffer conditions. Therefore, we presume the ZNF202-m3 protein to form an asymmetric dimer in solution.
The state of oligomerization in SCAN domain-mediated homo-or heterotypic protein associations may share similarities with the BTB/POZ domain-containing proteins (12,13). The GAGA protein has been shown to form BTB/POZ domain-mediated dimers, tetramers, and oligomers in vitro, whereas the BTB/POZ domain from the promyelocytic leukemia zinc finger oncoprotein crystallized in a tightly intertwined dimer (14,15). Furthermore, the BTB/POZ domain formed dimer-dimer interactions in the crystals that suggested a mode of higher order protein oligomerization (16).
Transcription factors are composed of functional domains that contribute to DNA binding, ligand binding, and transcriptional activation or repression (17,18). One important regulatory mechanism affecting both DNA binding affinity as well as transcriptional activation is oligomerization (19,20). Whereas zinc finger transcription factors of the Cys 2 -His 2 class are reported to bind DNA in monomeric form, SCAN domain-mediated association with other zinc finger proteins may modulate their biological function (21-23). The juxtaposition of zinc fin-ger motifs through protein oligomerization may lead to altered DNA binding activities. For example, the self-oligomerization of the amino-terminal BTB/POZ domain of the GAGA zinc finger protein resulted in increased DNA binding affinity and transcriptional activity (14,24). Recently, BTB/POZ domainmediated oligomerization of transcription factors has been suggested to serve as combinatorial codes for gene expression (25).
The identification of the SDP1 protein establishes a new family of proteins characterized by an isolated SCAN domain. The amino acid sequence conservation between SDP1 and PGC-2 is high, and consequently SDP1 may represent the human ortholog of the murine PGC-2 protein (3). A third protein that contains only a SCAN domain is encoded by the m3-splice form of ZNF202.
Single domain proteins that function as intermolecular regulators of transcription have been described previously. The PGC-2 protein was shown to interact with the nuclear receptor peroxosome proliferator-actived receptor ␥. Upon association, PGC-2 increases the transcriptional activity of peroxosome proliferator-actived receptor ␥ in the absence of any intrinsic transcriptional activity of PGC-2 (3). The biological effect of peroxosome proliferator-actived receptor ␥ may be modulated through competitive binding of PGC-2 and other transcriptional cofactors. Conversely, the DNA binding activities of some basic helix-loop-helix motif-containing transcription factors were shown to be inhibited by Id proteins. The Id family of helixloop-helix proteins, which lacks a basic DNA-binding domain, functions as a negative regulator of basic helix-loop-helix proteins through the formation of inactive heterodimers with intact basic helix-loop-helix transcription factors (26,27). Similarly, isolated SCAN domain proteins such as SDP1 or ZNF202-m3 may modulate the formation of functional SCAN domain zinc finger transcription factors.
The elucidation of the biological consequences arising from associations between SCAN domain proteins may offer new insights into the regulation of zinc finger transcription factors. Transient transfection experiments with appropriate reporter gene constructs may address the impact of SCAN domainmediated oligomerization on gene expression. Particularly, the association of two different splice forms, such as the m1 and m3 forms of ZNF202, may represent a novel biological mechanism to regulate transcriptional activity.