Much of what is known about the organization of eukaryotic enhancer regions comes from systematic studies of the SV40 enhancer, where it was observed that individual cis-acting elements often do not have intrinsic enhancer function but rather function in conjunction with one or more adjacent cis-acting sequences (1, 2) . The individual cis-acting elements have been referred to as enhansons, and they constitute the binding sites for specific trans-acting proteins. Different types of enhansons have been described for the SV40 enhancer including: 1) enhansons that have intrinsic enhancer activity, 2) enhansons that function when dimerized or multimerized, and 3) enhansons that function in combination with other enhanson elements. We previously described a nuclear protein, termed EF-C, that binds to important functional sites in the polyomavirus (Py) ()and hepatitis B virus (HBV) enhancer regions (Fig. 1, A and B) (3, 4, 5) . EF-C binds to an inverted repeat sequence (5`-GTTGCYNGGCAAC-3`, Fig. 1C). While the EF-C binding site is among the key cis-acting elements essential for efficient enhancer activity(4, 5, 6) , an individual EF-C site does not have intrinsic enhancer activity, nor does it function when arranged in tandem copies ( (7) and data not shown). Therefore, the EF-C binding site appears to correspond to the third class of enhansons, and we have found that EF-C functions in conjunction with adjacent enhancer elements(5, 6) .

Figure 1: A and B, schematic views of the polyomavirus (Py, A) and hepatitis B virus (HBV, B) enhancer regions. The Py enhancer from nt 5104-5177 is displayed with the binding sites for nuclear factors PEA-3, EF-1A, PEA-1, PEA-2, and EF-C indicated; see (5) for specific references for these binding activities. The HBV enhancer from nt 1117-1247 is shown with the binding site for the nuclear factors HBLF, HNF-4, RXR, COUP-TF, EF-C, NF-1, C/EBP, and ATF indicated. See (6) for specific references for these binding activities. C, a consensus EF-C binding site is shown with the EF-C inverted repeat indicated by arrows above the sequence. The point mutations introduced at specific nucleotides are shown below the sequence.

We and others have shown that EF-C corresponds to two seemingly unrelated activities that have been described. The first is MDBP, a protein that binds to certain DNA sites only when methylated at CpG dinucleotide base pairs(8, 9, 10) . EF-C and MDBP share indistinguishable binding properties to a number of different binding sites, including the binding to certain sites only when methylated at CpG base pairs (11) . Recently, the identification that EF-C and MDBP correspond to an additional activity, RFX-1, was described(12, 13) . RFX-1 is a member of a family of related transcription factors that form homo- and heterodimers, and that bind to the conserved and functionally important X box in MHC class II antigen promoter regions(14) . Experiments using RFX-1 antisense RNA have shown that RFX-1 is a functional regulator of MHC class II gene expression in vivo(15) . However, it is clear from the analysis of cell lines from patients with combined immunodeficiency syndrome that other RFX-1-related activities, as well as other activities that are distinguishable from RFX-1, are important for MHC class II gene expression(14, 16, 17) . Thus a complicated picture of different transcriptional regulators that bind to the MHC class II promoter X box has emerged. How these activators regulate basal and induced expression of MHC class II gene expression has not been determined.

In this report, we have investigated the binding of EF-C purified from HeLa cell extracts and RFX-1 produced in vitro to binding sites in the HBV and Py enhancer regions that carry mutations that either disrupt one arm of the EF-C inverted repeat, or alter the spacing between the repeats. These mutations result in the generation of EF-C half-sites which mimic the RFX-1 binding site in the MHC class II promoter. Our results show that the interaction of EF-C with an intact inverted repeat is required for functional activity of viral enhancer regions, and that the interaction of EF-C and RFX-1 with the DRA MHC class II promoter truly represents half-site interaction. The binding of EF-C and RFX-1 to an EF-C half-site is unstable, while binding to the inverted repeats present in viral enhancer regions is stable. These results suggest that an additional activity may be required to stabilize RFX-1 interaction with the MHC class II promoter, and that viral enhancer regions have evolved high affinity binding sites for dimeric EF-C/RFX-1.

MATERIALS AND METHODS

Plasmid DNAs, Probes, and in Vivo Enhancer Assays

EF-C Purification, RFX-1 Synthesis, and Electrophoretic Mobility Shift Assays

Figure 3: Analysis of purified EF-C. A, the proteins present in pooled fractions containing EF-C activity obtained during purification were fractionated by SDS-polyacrylamide electrophoresis and the gel was silver-stained. The arrow indicates the presumptive EF-C product that was highly enriched in the final EF-C containing fraction compared to fractions that lacked EF-C activity (see ``Materials and Methods''). Proteins present in the final affinity-purified fraction were concentrated 25-fold by acetone precipitation before analysis by SDS-polyacrylamide gel electrophoresis. B, polyclonal rabbit antibodies directed against recombinant RFX-1 were added to in vitro binding reactions containing affinity-purified EF-C or RFX-1 produced by in vitro translation. - Ab indicates no addition of antibody to the binding reactions. The adjacent two lanes indicate the addition of preimmune or RFX-1 antibodies to the binding reaction. C, polyclonal antibodies directed against recombinant RFX-1 were added to in vitro binding reactions containing nuclear extract from HeLa, NIH 3T3 (MP8), or HepG2 cells, as described in B.

Electrophoretic mobility shift assays were performed in a final volume of 12.5 µl and contained 0.5-1.0 µl of EF-C B1 fraction or 3-4 µl of in vitro translated RFX-1 in 20 mM HEPES, pH 7.4, 100 mM KCl, 1.0 mM MgCl, 0.5 mM DTT, 0.1 mM EDTA, 0.1% CHAPS, 100 µg/ml single-stranded calf thymus DNA, and 40,000 cpm of probe (2.0 fmol). The DNA binding reactions were performed at room temperature for 30-60 min, and the resulting DNA-protein complexes were resolved by electrophoresis on 4% (30:1) polyacrylamide gels run in 0.5 TBE (25 mM Tris, pH 8.3, 25 mM borate, 0.5 mM EDTA) at 200 V at 4 °C. In mobility shift assays in which antiserum was used, the RFX-1 polyclonal antiserum was diluted and added to the DNA binding reactions 30 min into the incubation, and the incubation continued for an additional 30 min. For off-rate analyses, binding reactions were performed for 30 min, at which time a 500-fold molar excess of cold competitor DNA was added to the binding reaction. Aliquots of the reactions were withdrawn at various times points and analyzed by gel mobility shift assay. OPCU footprinting and DEPC interference assays were performed as described previously(20) . Oligonucleotide binding site competitors have been described previously(11) .

RESULTS

We first compared the binding of EF-C purified from HeLa cells and RFX-1 produced by in vitro translation to EF-C and RFX-1 binding sites using a gel mobility shift assay (Fig. 2). The binding sites corresponded to the Py EF-C inverted repeat binding site and the DRA MHC class II promoter X box RFX-1 binding site; the nucleotide sequence of the binding sites in these probes is shown in Fig. 7. Affinity-purified HeLa cell EF-C formed a complex with the Py EF-C probe that represents dimeric protein binding to both arms of the inverted repeat (upper complex, lane -, affinity-purified EF-C + EF-C probe; see below). The faster migrating complexes in this lane appears to reflect EF-C protein breakdown during purification since the appearance of this complex is largely blocked by a mixture of protease inhibitors and these complexes generally are not evident in the starting material or in certain EF-C preparations (see Fig. 8and 11, for example). A complex with similar mobility was observed using in vitro translated RFX-1, as well as a complex with faster mobility (lane -, in vitro translated RFX-1 + EF-C probe). The slowest migrating complex with RFX-1 corresponds to occupancy of both arms of the EF-C inverted repeat (see below) and has slightly reduced mobility in comparison to purified EF-C due to the high protein content in the reticulocyte lysate present in the binding reaction as confirmed by mixing experiments (data not shown). In contrast, the vast majority of the RFX-1 complex on the DRA X box RFX probe displayed the more rapid mobility (lane -, in vitro translated RFX-1 + RFX probe), which corresponds to monomeric RFX binding to the probe DNA(15) . Purified EF-C and in vitro translated RFX-1 displayed identical patterns of binding on the EF-C probe when a series of specific (Py-EF-C-WT, HBV-EF-C, CMV-1, RF-X) and nonspecific (Py-EF-C-X, pBR) competitor DNAs were added to the binding reaction (100-fold molar excess to the probe DNA), including the specific competition of both binding proteins by a fragment from pBR322 DNA only when methylated at CpG dinucleotide base pairs (pBRversus pBR). The pattern of competition of monomeric RFX-1 to the RFX site was comparable except that the methylated pBR site was a weak competitor. The conclusion that EF-C and RFX-1 are identical or highly related activities was confirmed using a polyclonal rabbit antibody developed against purified RFX-1. This antibody recognized (supershifted) affinity-purified EF-C and in vitro translated RFX-1 (Fig. 3B) and recognized EF-C-specific complexes previously described (4, 6) using different cell sources including HeLa, NIH 3T3 (MOP8), and HepG2 cells (Fig. 3C), all of which contain functional EF-C activity ( (11) and (12) and see below). Preimmune serum did not alter the mobility of these EF-C complexes. While this analysis does not address the possibility that RFX-1 may be present as mixed heterodimers in the different cellular extracts or that the antibody may cross-react with other members of the RFX family, these results indicate that EF-C and RFX-1 from a variety of cell sources are highly related, if not identical, activities. We conclude that EF-C purified from HeLa cell displays comparable binding properties to RFX-1, as previously found using crude nuclear extracts (12, 13) .

Figure 2: Binding of purified EF-C and in vitro translated RFX-1 to the Py EF-C and DRA RFX binding sites. EF-C was purified from HeLa cell extracts as described under ``Materials and Methods,'' and RFX-1 was produced by in vitro transcription/translation. EF-C and RF-X (indicated at the top) were incubated with P-labeled probes of the Py EF-C site and DRA RFX-1 site (see Fig. 7). Binding reactions either lacked specific competitor DNA(-) or contained the Py-EF-C-WT, Py-EF-C-X, HBV EF-C, CMV-1, RF-X (DRA X box) and nonmethylated or methylated pBR322 MDBP competitor DNAs in a 100-fold molar excess to the probe DNA. Reactions contained the Py EF-C or DRA RFX binding site probes are indicated at the bottom. The products of the binding reactions were analyzed in a gel mobility shift assay. The positions of monomeric (M) and dimeric (D) RF-X DNA-protein complexes are indicated on the right; this designation is derived from the experiments of Reith et al.(15) .

Figure 7: Schematic view of the EF-C and DRA RFX binding sites. The opencircles represent sites of DEPC interference. The darkbars represent regions of protection from OPCU cleavage. The twoasterisks above the T residues in the DRA RFX site correspond to conserved positions in the EF-C consensus binding site discussed in the text.

Figure 8: Binding of in vitro translated RFX-1 and purified EF-C to wild type and mutant Py enhancer binding sites. Proteins prepared as described for Fig. 2were incubated with the wild type Py enhancer probe (nt 5104-5177) or mutant sites described in the text and shown in Fig. 1C. The products of the binding reactions were analyzed in a gel mobility shift assay. The positions of monomeric (M) and dimeric (D) RFX-1 are indicated on the right.

To analyze the physical interaction of EF-C and RFX-1 with the Py EF-C binding site and the MHC class II promoter X box RFX-1 binding site, we performed DEPC interference and OPCU chemical footprinting assays using purified EF-C bound to the Py EF-C and DRA RFX binding sites, and DEPC interference assays with RFX-1 bound to these sites. The reticulocyte lysate precluded the use of in vitro translated RFX-1 in OPCU footprinting reactions (data not shown). The data are shown in Fig. 4Fig. 5Fig. 6, and are summarized schematically in Fig. 7. Purified EF-C and RFX-1 displayed identical patterns of modification interference on both strands of each binding site (circles above and below sequences in Fig. 7). A more extensive pattern of interference sites were evident with the dimeric EF-C binding site probe in comparison to the DRA RFX half-site. Since this assay is limited by the purine residues that are are available for modification and whose modification have an impact on protein-DNA binding, OPCU chemical footprinting was used to analyze the extent of interaction of these proteins with the binding sites. These results (darkbars in Fig. 7) show that a broader segment of DNA is protected with the binding of EF-C to the Py inverted repeat in comparison to the DRA X box RFX-1 binding site. This supports the idea that only a monomeric subunit of the dimeric EF-C protein complex interacts with the X box half-site. Additionally, the region of protection and pattern of DEPC-interference of EF-C and RFX-1 on the DRA RFX binding site indicate that a monomer of RFX-1 or EF-C makes significant contacts within this binding site. Only two nucleotide contact sites were different when RFX-1 monomers and dimers were compared in DEPC interference assays (the leftmost two Gs on the upper strand of the Py EF-C site in Fig. 7; data not shown).

Figure 4: DEPC interference assay of RFX-1 and purified EF-C at the Py EF-C binding site. Purified EF-C and in vitro translated RFX-1 were bound to DEPC-modified probe DNA as the products separated in a preparative mobility shift gel. The complexed (B, bound) and noncomplexed (U, unbound) DNAs were eluted, treated with piperidine, and the products analyzed in a denaturing polyacrylamide-urea gel. GA and CT are sequencing ladders of the homologous probes. The darkdots adjacent to the autoradiograms represent sites of modification interference.

Figure 5: DEPC interference assay of RFX-1 and purified EF-C at the DRA RFX binding site. Purified EF-C and in vitro translated RFX-1 were bound to DEPC-modified probe DNA as the products separated in a preparative mobility shift gel. The complexed (B, bound) and noncomplexed (U, unbound) DNAs were eluted, treated with piperidine, and the products analyzed in a denaturing polyacrylamide-urea gel. GA and CT are sequencing ladders of the homologous probes. The darkdots adjacent to the autoradiograms represent sites in modification interference.

Figure 6: OPCU footprint analysis of EF-C bound to the Py EF-C and DRA RFX binding sites. Purified EF-C was incubated with the Py EF-C or DRA RFX probes, and DNA protein complexes were resolved by gel mobility shift assay. In situ OPCU cleavage reactions were performed, DNA present in protein bound (B) and unbound (U) complexes was eluted from the gel, and analyzed in a denaturing polyacrylamide-urea gel. GA and CT are sequencing ladders of the homologous probes. The darkbars to the left and right indicate regions of protection.

To address the binding specificity of monomeric versus dimeric EF-C and RFX-1 at the inverted repeat binding site and to provide a direct correlation between binding studies performed in vitro and enhancer activity assays performed in vivo, we generated a series of single nucleotide point mutations at conserved residues in the Py EF-C binding site (Fig. 1C). The EF-C consensus binding site represented in Fig. 1C represents a compilation of high affinity binding sites including the sites in the Py and HBV enhancer regions as well as two sites in the cytomegalovirus enhancer. Among these sites, 6 nucleotides are invariant. We mutated each of these sites using nonconservative nucleotide changes. These mutations were introduced into the Py enhancer region in a vector that carries an adjacent Py origin of DNA replication. With polyomavirus, there is a stringent requirement for the presence of a functional enhancer region adjacent to the origin of replication for efficient DNA replication in vivo. The effects of these mutations was first examined using in vitro binding analyses with purified EF-C and RFX-1 and the individual mutant sites as probes in binding reactions (Fig. 8). As predicted, each mutation reduced the binding of purified EF-C dramatically with the exception of mutation at nt position 8. Weak binding also was evident with a mutation at nt positions 2 and 12. Similar results were observed when the binding of in vitro translated RFX-1 to these sites was examined, with the exception that strong monomeric RFX-1 binding was evident with PM8 and PM12. The other mutant sites were dramatically reduced for monomeric and dimeric RFX-1 binding in vitro, even though the mutations are present on opposite arms of the inverted repeat (see Fig. 1C, PM2, PM3, and PM5 versus PM9). We conclude that the interaction of purified EF-C and monomeric or dimeric RFX-1 with nucleotides in both arms of the inverted repeat are required for stable DNA-protein interaction. It is clear that monomeric RFX-1 interacts with nucleotides outside the context of a single EF-C half-site.

We tested these mutant sites for enhancer activity using two assays. The first was an in vivo DNA replication assay where the enhancement of Py DNA replication by the Py enhancer region was assayed in transfected murine NIH3T3 cells (MOP8) that express the Py large T antigen (Fig. 9A). The second was an in vivo transient expression assay where the enhancement of expression from the SV40 early promoter (enhancer-less vector) was assayed in transfected human HeLa cells (Fig. 9B). Quantitatively similar results were obtained with the two assays comparing the enhancer activity of the wild type Py enhancer to the enhancer regions containing EF-C site mutations. The effects of the mutations on EF-C and RFX-1 binding properties in vitro were virtually identical to the effects of the mutations on enhancer activity in vivo. With the replication assay (Fig. 9A), no replication was evident with the enhancer-less replication vector (ENH-). A significant effect of the intact enhancer (Py 5104-5177) was observed, while the deletion of the region containing the EF-C site (Py 5104-5159) strongly reduced this stimulation. Point mutation at positions 3, 5, and 9 reduced enhancer activity to the level observed when the EF-C site was deleted. Mutations at positions 2 and 12 also had significant, but less dramatic, effects. Mutation at position 8 had no effect on enhancer activity. With the expression assay (Fig. 9B), weak promoter activity was evident with the plasmid lacking an enhancer, while the wild type Py enhancer stimulated expression 7-fold. Mutations at positions 3, 5, and 9 reduced enhancer activity to background levels, while mutations at positions 2 and 12 had lesser effects and mutation at position 8 had only a modest effect. These results directly correlate the binding of EF-C to this site in vitro with enhancer function in vivo. Additionally, the comparison of results in vivo and in vitro with PM8 and PM12 show that while monomeric RFX-1 bound to both mutant sites in vitro, the activity of enhancer regions with mutations in these sites was quite different in vivo. Enhancer activity in vivo directly correlated with dimeric RFX-1 binding in vitro. This interpretation is supported by experiments with the HBV enhancer described below.

Figure 9: Enhancer activity of wild type and mutant EF-C containing enhancer regions in vivo.A, a transient replication assay was used to measure Py enhancer activity. The wild type and mutant Py enhancer regions, linked adjacent to the Py origin of DNA replication, were transfected into MOP8 cells which constitutively express the Py large T antigen. Twenty-four h after transfection, low molecular weight DNA was isolated, digested with DpnI + EcoRI, and the products analyzed by Southern blot analysis. The DpnI-resistant DNA represents newly replicated DNA, while the DpnI-sensitive DNA (non-replicated) represents input, transfected plasmid DNA. B, a transient expression assay was used to measure Py enhancer activity. The wild type and mutant Py enhancer regions, linked upstream of the SV40 early promoter region fused to the CAT gene, were transfected into HeLa cells. Twenty-four hours after transfection, CAT activity in cellular extracts was measured. The level of promoter activity with the enhancer-less vector was set at 1.0, and the enhancer activity of the different Py enhancer regions is given as -fold enhancement relative to this level.

We previously tested the functional properties of HBV enhancer regions carrying mutations that altered the spacing between the arms of the EF-C inverted repeat and found that the deletion of one base pair or the insertion of more than two base pairs strongly reduced transcriptional enhancer activity in transfected HeG2 hepatoblastoma cells(4) . The stable interaction of RFX-1 with certain EF-C half-sites in a mobility shift assay prompted us to test if monomeric RFX-1 interacts with the HBV EF-C site and to determine the effect of spacer mutations within this site on EF-C and RFX-1 binding. The analysis of monomeric RFX-1 binding was accomplished by in vitro translation of a truncated version of RFX-1 that lacks the C-terminal dimerization domain but retains the more N-terminal DNA binding domain and that binds as a monomer to the DRA X box binding site(15) . Interestingly, no interaction of monomeric RFX-1 was observed with the HBV EF-C site even though this protein bound efficiently to the Py EF-C site and DRA X box RFX-1 binding site (data not shown). Binding analyses using full-length RFX-1 also showed only dimeric RFX-1 binding to the HBV EF-C site and not the binding of monomeric RFX-1 (data not shown). These results are consistent with the poor binding of purified EF-C that was observed to HBV sites carrying mutations that disrupt enhancer function (Fig. 10, 1, IS3, IS9, and IS10). These results indicate that while the HBV EF-C binding site displays dimeric EF-C and RFX-1 binding, either half-site of this binding element is intrinsically weak for monomeric RFX-1 interaction.

Figure 10: Binding of purified EF-C to wild type and mutant HBV enhancer binding sites. Purified EF-C was incubated with the wild type HBV enhancer probe (nt 1117-1247) or mutant sites described in the text. The products of the binding reactions were analyzed in a gel mobility shift assay.

We compared the binding of affinity-purified EF-C to the Py EF-C and DRA RFX binding sites using a competition binding assay (Fig. 11A). Binding reactions contained either the Py EF-C or DRA RFX binding site probe and increasing molar concentrations of homologous cold competitor DNAs added simultaneously with the probe DNA. These results demonstrated that the Py EF-C binding site was a better competitor for EF-C binding than the DRA RFX site. To test the possibility that the interaction of EF-C with a Py EF-C site inverted repeat may be intrinsically more stable than binding to the RFX DRA half-site, we performed off-rate analyses using these sites with purified EF-C as shown in Fig. 11B. Binding reactions were established where comparable levels of EF-C-DNA complexes were observed with the individual probes, and following a 30-min binding reaction, a 1000-fold molar excess of the same binding site was added as a cold competitor DNA to measure DNA-protein complex decay. With each probe, t - competitor shows the level of EF-C binding to each probe in the absence of specific competitor DNA. t + competitor represents the addition of the competitor simultaneously with the probe DNA. The remaining time points represent competitor DNA added after a 30-min EF-C binding reaction. These results clearly show that EF-C binding to the inverted repeat of the Py enhancer is stable, while EF-C binding to the RFX X box is unstable. Thus, even though EF-C makes extensive contacts with nucleotides within the RFX X box binding site (Fig. 7), the intrinsic binding of EF-C to this site is weak. Identical results were found when the binding of monomeric or dimeric RFX-1 to each site was compared (data not shown).

Figure 11: Competition binding and off-rates of purified EF-C with the Py EF-C and DRA RFX sites. A, binding reactions contained the Py EF-C or DRA RFX binding site probe and either no competitor DNA (lane0) or increasing molar concentrations (5-, 25-, and 100-fold molar excess) of homologous cold competitor DNAs added simultaneously with the probe DNA. B, EF-C binding reactions were established using the Py EF-C and DRA RFX-1 binding site probes. Following a 30-min binding reaction, an aliquot of the sample was loaded on a mobility shift gel (t). A 500-fold molar excess of Py EF-C wild type competitor DNA was then added to the binding reaction and samples were withdrawn at 15-min intervals and loaded on a mobility shift gel (t, t, t, and t) to follow the decay of the original bound complexes. Lanet + competitor contained a 500-fold molar excess of the EF-C wild type competitor DNA in the initial binding reaction.

DISCUSSION

Nuclear factor EF-C binds as a dimer to inverted repeat sequences in the polyomavirus and hepatitis B virus enhancer regions ( Fig. 1(A and B), 3, and 4). EF-C corresponds to two seemingly unrelated activities that have been described. The first is MDBP, a protein that binds to certain DNA sites only when methylated at CpG dinucleotide base pairs(8, 9, 10) . The second is RFX-1, a member of a family of related transcription factors that form homo- and heterodimers and that bind to the conserved and functionally important X box in MHC class II antigen promoter regions(14) . In this report, we have compared the binding of EF-C and RFX-1 in vitro to wild type and mutant binding sites in the HBV and Py enhancer regions with the function of these sites in vivo. These mutations result in the generation of EF-C half-sites that mimic the RFX-1 binding site in the MHC class II promoter. Our results demonstrate that the interaction of dimeric EF-C and RFX-1 with an intact inverted repeat is required for the functional activity of viral enhancer regions. This conclusion is supported by the observation that a point mutation in the Py enhancer region (PM12) that does not reduce the binding of monomeric RFX-1 to this site but does dramatically reduced dimeric RFX-1 and EF-C binding in vitro was greatly reduced for enhancer activity in vivo. Additionally, while the HBV EF-C site is critical for optimal enhancer activity in vivo, no interaction of monomeric RFX-1 was detected with the HBV enhancer in vitro while dimeric RFX-1 and EF-C bound efficiently.

The RFA RFX-1 binding site has a 9-nucleotide pyrimidine stretch on the upper strand to the left of the EF-C half-site denoted in Fig. 7. Because dimethyl sulfate and DEPC interference assays would not reveal contacts of a protein within these nucleotides and because two T residues are positioned at an appropriate location relative to the EF-C consensus inverted repeat in this region (see Fig. 1C, labeled with asterisks in Fig. 7), it was possible that both partners of an EF-C/RFX-1 dimer contacted DNA at the DRA RFX site, similar to the interaction of the dimers with the viral EF-C inverted repeats. OPCU chemical footprinting assays, however, showed that a narrower segment of DNA was protected at the DRA RFX site with bound EF-C, which does not include protection of the aforementioned T residues. This result strongly suggests that the interaction of EF-C and RFX-1 with the DRA MHC class II promoter truly represents half-site interaction. While monomeric RFX-1 binds to the Py EF-C site, binding of monomeric RFX-1 to the HBV EF-C site was not detected. These sites differ notably in the 3 nucleotides in the spacer region between the arms of the EF-C inverted repeat. These binding results as well as the pattern of protein interaction evident from DEPC interference assays suggests that the composition of the spacer nucleotides strongly influences the binding of monomeric, but not dimeric, EF-C/RFX-1 to these sites. We note that the Py EF-C site and DRA RFX-1 binding site are identical at nine consecutive positions including five base pairs of the EF-C half-site and four base pairs corresponding to the spacer region. The fact that the intact EF-C inverted repeat binding site displayed very stable binding of EF-C and dimeric RFX-1 while half-site binding was unstable suggests that viral enhancer regions have evolved dimeric binding sites for stable EF-C/RFX-1 interaction to effectively compete with the cellular genome for binding of limiting concentrations of this transcription factor.

Since the predominant form of EF-C detected in cellular extracts by gel mobility shift assay ( Fig. 8and Fig. 11, for example) and in gel filtration experiments (data not shown) is that of a protein dimer, one would anticipate that certain cellular promoter and enhancer regions would display inverted repeat binding sites. Such binding sites have been uncovered by computer analysis and found to bind MDBP(21) . Additionally, we have found a perfect EF-C inverted repeat, which is virtually identical to the Py EF-C binding site, within the enhancer region of the CoA synthetase transcriptional control region; this site binds EF-C with high affinity (data not shown), although the functional importance of this site is not clear at the moment. Other functionally important EF-C/RFX-1 half-sites have been found in cellular promoter regions. RFX-1 binds to the rpL30 ribosomal protein-encoding gene at the element, a sequence that strongly resembles the RFX-1 half-site in the DRA promoter(22) . The unstable binding of EF-C/RFX-1 to a half-site suggests that an additional activity may be required to stabilize RFX-1 interaction with the MHC class II and rpL30 promoter regions. With the MHC class II promoter, a second regulatory element, the X2 box, overlaps the RFX-1 binding site and binds members of the leucine zipper family of transcription factors(23) . Perhaps RFX-1 interacts with one or more of these activities. Additionally, it is clear that RFX-1 forms heterodimers with other family members, RFX-2 and RFX-3(14) . While RFX-2 and RFX-3 show the same binding specificity as RFX-1 is competition and footprinting experiments(14) , it is possible that specific heterodimeric combinations of these activities interact with a half-site with higher affinity. Finally, a recent report has shown that an additional RFX family member, which has not been cloned, binds cooperatively with NF-Y, a protein that binds to MHC class II promoter Y box located 25 base pairs toward the transcription initiation site(24) . With the rpL30 ribosomal protein-encoding gene, the element functions in conjunction with neighboring transcription factor binding sites(22, 25) , and it is possible that similar cooperative interactions between RFX-1 and other proteins may occur. Perhaps through the use of different types of binding sites with varying binding affinities, the expression of certain sets of genes is regulated is cell-type specific manners dependent on the availability of other binding partners (e.g. RFX and NF-Y). Viruses may wish to circumvent such a limitation and therefore have evolved a site that binds a ubiquitous enhancer factor in a autonomous manner.

FOOTNOTES

*

This work was supported in part by United States Public Health Service Grant AI29427 (to P. H.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked ``advertisement'' in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

§

Supported by United States Public Health Service Training Grant CA09176 from the NCI, National Institutes of Health.

¶

To whom correspondence should be addressed. Tel.: 516-632-8813; Fax: 516-632-8891.

()

The abbreviations used are: Py, polyomavirus; HBV, hepatitis B virus; MHC, major histocompatibility complex; nt, nucleotides(s); DTT, dithiothreitol; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]1-propanesulfonic acid; DEPC, diethyl pyrocarbonate; OPCU, 1,10phenanthroline-copper.

ACKNOWLEDGEMENTS

REFERENCES

1. Ondek, B., Gloss, L., and Herr, W. (1988) Nature 333, 40-45 [CrossRef][Medline] [Order article via Infotrieve]
2. Fromental, C., Kanno, M., Nomiyama, H., and Chambon, P. (1988) Cell 54, 943-953 [CrossRef][Medline] [Order article via Infotrieve]
3. Ostapchuk, P., Diffley, J. F. X., Bruder, J. T., Stillman, B., Levine, A. J., and Hearing, P. (1986) Proc. Natl. Acad. Sci. U. S. A 83, 8550-8554 [Abstract/Free Full Text]
4. Ostapchuk, P., Scheirle, G., and Hearing, P. (1989) Mol. Cell. Biol. 9, 2787-2797 [Abstract/Free Full Text]
5. Bolwig, G. M., and Hearing, P. (1991) J. Virol. 65, 1884-1892 [Abstract/Free Full Text]
6. Garcia, A. D., Ostapchuk, P., and Hearing, P. (1993) J. Virol. 67, 3940-3950 [Abstract/Free Full Text]
7. Dikstein, R., Faktor, O., Ben-Levy, R., and Shaul, Y. (1990) Mol. Cell. Biol. 10, 3683-3689 [Abstract/Free Full Text]
8. Huang, L. H., Wang, R., Gaga-Sosa, M. A., Shenoy, S., and Ehrlich, M. (1984) Nature 308, 293-295 [CrossRef][Medline] [Order article via Infotrieve]
9. Khan, R., Zhang, X. Y., Supakar, P. C., Ehrlich, K. C., and Ehrlich, M. (1988) J. Biol. Chem. 263, 14374-14383 [Abstract/Free Full Text]
10. Supakar, P. C., Zhang, X.-Y., Githens, S., Khan, R., Ehrlich, K. C., and Ehrlich, M. (1989) Nucleic Acids Res. 17, 8611-8628 [Abstract/Free Full Text]
11. Garcia, A. D., Ostapchuk, P., and Hearing, P. (1991) Virology 182, 857-860 [CrossRef][Medline] [Order article via Infotrieve]
12. Siegrist, C. A., Durand, B., Emery, P., David, E., Hearing, P., Mach, B., and Reith, W. (1993) Mol. Cell. Biol. 13, 6375-6384 [Abstract/Free Full Text]
13. Zhang, X. Y., Ferrat, N. J., Asiedu, C. K., Samac, S., Peterlin, B. M., and Ehrlich, M. (1993) Mol. Cell. Biol. 13, 6810-6818 [Abstract/Free Full Text]
14. Reith, W., Ucla, C., Barras, E., Gaud, A., Durand, B., Herrero-Sanchez, C., Kobr, M., and Mach, B. (1994) Mol. Cell. Biol. 14, 1230-1244 [Abstract/Free Full Text]
15. Reith, W., Herrero-Sanchez, C., Kobr, M., Silacci, P., Berte, C., Barras, E., Fey, S., and Mach, B. (1990) Genes & Dev. 4, 1528-1540
16. Herrero-Sanchez, C., Reith, W., Silacci, P., and Mach, B. (1992) Mol. Cell. Biol. 12, 4076-4083 [Abstract/Free Full Text]
17. Voliva, C. F., Aronhein, A., Walker, M. D., and Peterlin, B. M. (1992) Mol. Cell. Biol. 12, 2383-2390 [Abstract/Free Full Text]
18. Kunkel, T. A. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 488-492 [Abstract/Free Full Text]
19. Dignam, J. D., Lebovitz, R. M., and Roeder, R. G. (1983) Nucleic Acids Res. 11, 1475-1489 [Abstract/Free Full Text]
20. Bolwig, G. M., Bruder, J. T., and Hearing, P. (1992) Nucleic Acids Res. 20, 6555-6564 [Abstract/Free Full Text]
21. Zhang, X.-Y., Asiedu, C. K., Supakar, P. C., Khan, R., Ehrlich, K. C., and Ehrlich, M. (1990) Nucleic Acids Res. 18, 6253-6260 [Abstract/Free Full Text]
22. Safrany, G., and Perry, R. P. (1993) Gene (Amst.) 132, 279-283 [CrossRef][Medline] [Order article via Infotrieve]
23. Liou, H. C., Boothby, M. R., and Glimcher, L. (1988) Science 242, 69-72 [Abstract/Free Full Text]
24. Reith, W., Siegrist, C. A., Durand, B., Barras, E., and Mach, B. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 554-558 [Abstract/Free Full Text]
25. Hariharan, N., Kelley, D. E., and Perry, R. P. (1989) Genes & Dev. 3, 1789-1800

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				Y. Lubelsky, N. Reuven, and Y. Shaul Autorepression of Rfx1 Gene Expression: Functional Conservation from Yeast to Humans in Response to DNA Replication Arrest Mol. Cell. Biol., December 1, 2005; 25(23): 10665 - 10673. [Abstract] [Full Text] [PDF]

				R. Tierney, H. Kirby, J. Nagra, A. Rickinson, and A. Bell The Epstein-Barr Virus Promoter Initiating B-Cell Transformation Is Activated by RFX Proteins and the B-Cell-Specific Activator Protein BSAP/Pax5 J. Virol., November 15, 2000; 74(22): 10458 - 10467. [Abstract] [Full Text]

				E. K. Schmitt and U. Kuck The Fungal CPCR1 Protein, Which Binds Specifically to beta -Lactam Biosynthesis Genes, Is Related to Human Regulatory Factor X Transcription Factors J. Biol. Chem., March 24, 2000; 275(13): 9348 - 9357. [Abstract] [Full Text] [PDF]

				Y. Katan-Khaykovich and Y. Shaul RFX1, a Single DNA-binding Protein with a Split Dimerization Domain, Generates Alternative Complexes J. Biol. Chem., September 18, 1998; 273(38): 24504 - 24512. [Abstract] [Full Text] [PDF]

				K. Roder, S. S. Wolf, K.-F. Beck, and M. Schweizer Cooperative Binding of NF-Y and Sp1 at the DNase I-hypersensitive Site, Fatty Acid Synthase Insulin-responsive Element 1, Located at -500 in the Rat Fatty Acid Synthase Promoter J. Biol. Chem., August 22, 1997; 272(34): 21616 - 21624. [Abstract] [Full Text] [PDF]

				S. Habib and S. E. Hasnain A 38-kDa Host Factor Interacts with Functionally Important Motifs within the Autographa californica Multinucleocapsid Nuclear Polyhedrosis Virus Homologous Region (hr1) DNA Sequence J. Biol. Chem., November 8, 1996; 271(45): 28250 - 28258. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by David, E. Articles by Hearing, P. Search for Related Content PubMed PubMed Citation Articles by David, E. Articles by Hearing, P. Social Bookmarking                      What's this?