Wnt Activation and Alternative Promoter Repression of LEF1 in Colon Cancer

Cell culture.

COS-1, SW480, and HT-29 cells were cultured in high-glucose Dulbecco's modified Eagle medium containing 10% fetal bovine serum (Omega Scientific) and 2 mM l-glutamine. Jurkat, Colo 320 HSR, and DLD1 cell lines were cultured in RPMI 1640 with 10% fetal bovine serum and 2 mM l-glutamine. Jurkat media also contained 50 μM 2-mercaptoethanol. D7p11 cells were cultured in DLD1 media supplemented with 500 μg/ml Zeocin (InvivoGen) and 10 μg/ml blastocidin (InvivoGen).

Reporter plasmid constructs.

Nested deletions of promoter 2 were created by exonuclease III and mung bean digestion starting with a fragment containing the last 846 nucleotides of intron 2 and 50 nucleotides of exon 3. Smaller promoter fragments (−846 and smaller) were cloned into pGL2 reporter plasmids that contain the simian virus 40 (SV40) enhancer, with the exception of −27/+60 and −816/+60, which were also inserted into the enhancerless pGL2-basic plasmid. Larger promoter 2 constructs containing upstream exonic and intronic sequences were created by digestion with the indicated restriction enzymes at the following sites: BspEI, −4024/+60; SpeI, −3004/+60; and MscI, −1446/+60. Then, they were cloned into pGL2-basic.

Mutations in the initiator elements in the −27-to-+60 promoter construct were introduced by PCR using the following primers: Inr mutant 1 (GATAGGTACCGAGTGGGAGCATCATTGATTGTTCTTTG), Inr mutant 2 (GATAGGTACCGAGTCAGTGCATGGTAGATTGTTCTTTG), and a double Inr mutant (GATAGGTACCGAGTGGGAGCATGGTAGATTGTTCTTTG). The 3′ primer is complementary to the luciferase gene of the pGL2 vector backbone (Promega). PCR products were inserted into Asp718 and HindIII sites of pGL2-enhancer.

Mutations surrounding the transcription start site were introduced via annealed complementary oligonucleotides of 57 and 61 nucleotides containing the mutations and spanning the promoter. One set of mutations converted the transcription start site region to a GC-rich NotI sequence. The other set is a three-nucleotide mutation that retains the GC-to-AT ratio (see Fig. 6B). Annealed oligonucleotides contained an Asp718 5′ overhang and a blunt end for ligation into Asp718 and Klenow-filled HindIII sites of pGL2 enhancer. The repressor was delimited to −1446/−1281 by internal deletion using blunt-end treatment of BlpI (−1281) and TthIII (−177) sites in intron 2. Fusion of the repressor to LEF1 promoter 1 (−672, +305) was achieved by PCR amplification of repressor sequences and insertion of the 165-nucleotide fragment into LEF1P1 plasmic at a Klenow-filled Asp718 site (−672).

Transient transfection assay.

Jurkat cells were transfected by electroporation using a BTX 600 electroporator. A total of 5 μg of reporter plasmid and 0.5 μg of cytomegalovirus (CMV)-beta-galactosidase plasmid were added to 10 × 10⁶ cells and electroporated with settings of 160 V, 72 Ω, and 2,000 microfarads. For COS-1 and Colo 320 HSR cells, 2.5 × 10⁵ cells were transfected with 0.5 μg of reporter plasmid and 0.1 μg of CMV-beta-galactosidase plasmid by using Effectene reagent according to the manufacturer's protocol (QIAGEN). For activation studies, 200 ng of TCF-1E and 400 ng of β-catenin expression vectors were cotransfected in addition to the reporter gene and CMV-β-galactosidase plasmid. In all cases, the cells were harvested at 18 to 20 h posttransfection and assayed for luciferase activity. Luciferase activity values were normalized using beta-galactosidase activity values.

Western blot analysis.

Cell lysates of 1 × 10⁵ DLD1 cells treated with doxycycline (D7p11 cells; 1 μg/ml) were analyzed for induced expression of dnTCF-1 by Western blotting with a polyclonal LEF-1 antibody (Ab) that detects all LEF/TCF isoforms (18) and secondary anti-rabbit immunoglobulin G (IgG) conjugated to horseradish peroxidase. Blots were developed with ECL reagent (Amersham).

Northern analysis.

RNA from cell lines was extracted by use of TRIzol (Invitrogen) according to the manufacturer's suggested protocol. Twenty micrograms of human thymus RNA (Clontech) and 10-μg portions of Jurkat, HeLa, SW480, Colo 320, DLD1, D7p11, HT-29, and COS-1 RNA were analyzed for LEF1 expression. The blot was probed with a LEF1 cDNA probe and subsequently with a GAPDH cDNA probe as previously described (18).

RNase protection assay.

The RNase protection assay was performed as previously described (18). RNA markers were made with single digestions of the pBluescript vector within the region of the multiple cloning site with EcoRI, XhoI, BamHI, Asp718, and XbaI. The LEF1 RNA probes were made from a subcloned 147-nucleotide fragment spanning 37 nucleotides of intron 2 and 50 nucleotides of exon 3 plus vector sequence in both orientations into pBluescript vector containing a T7 promoter.

Cancer profiling array.

The probing of the cancer profiling array I blot (Clontech) was performed according to manufacturer's suggestions with a few modifications. LEF1 (1.1-kb fragment of the open reading frame) or ubiquitin (Clontech) probes were labeled with ³²P using a Megaprime DNA-labeling kit (Amersham). The cancer array was hybridized at 65°C overnight with 20 × 10⁶ cpm of labeled probe in 20 ml of ExpressHyb (Clontech), 30 μg of COT-1 DNA (Roche), and 150 μg of sonicated salmon sperm DNA in 5× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate).

FISH assay.

Cells were harvested and fixed on slides from either the normal human fibroblast cell line GM5399-E6 (a generous gift from Eric Stanbridge, UC Irvine) or from the Colo 320, SW480, or HT-29 colon cancer cell line for fluorescence in situ hybridization (FISH) analysis. The probe used for FISH was derived from a PAC clone (PAC41H1 [18]) containing the 5′ upstream region and the first 3 exons of the LEF1 locus. This DNA was labeled by a digoxigenin nick translation mix (Roche) according to the manufacturer's suggested protocol. Hybridization was performed according to Cambio's suggested protocol A. LEF1 probe (2 μl) and 1 μl of COT-1 DNA were lyophilized to near dryness (around 1 μl) and then added to 3 μl of the chromosome 4 paint probe (Cambio) and 12 μl of hybridization buffer (Cambio) and denatured at 75°C for 5 minutes. Samples were hybridized overnight at 37°C then washed in 0.4% SSC-0.5% NP-40 at 73°C for 2 min, cooled to room temperature, soaked in 4× SSC, and slightly dried. Samples were blocked with 70 μl of 4× SSC-1% bovine serum albumin (BSA) for 2 min before the addition of digoxigenin-rhodamine antibody (2 μg/ml; Roche) and neutravidin-fluorescein conjugate (10 μg/ml; Molecular Probes) to develop the LEF1 signal and chromosome 4 paint probe at 37°C for 30 min in the dark. Slides were washed in 4× SSC, 4× SSC-0.3% NP-40, and 4× SSC for 2 min each at room temperature and counterstained with DAPI (4′,6′-diamidino-2-phenylindole) at room temperature for 2 min. Samples were mounted with 20 μl of antifade solution, and metaphase spreads visualized with an epifluorescence microscope (Zeiss) by use of standard filters.

DNase I footprinting.

The intronic LEF1 fragment (promoter 2) −77/+6, used for footprint analysis of minimal promoter sequences, was labeled with ³²P at the HindIII site in the polylinker region of the pGL2-enhancer backbone and then digested with Asp718. For upstream regions, a promoter 2 fragment (−300/+60) was cut with BlpI (−177) and labeled with ³²P to create two probes to survey sequences flanking −177. The recombinant LEF-1 protein and DNase I footprint procedure was performed as previously described (18, 19).

ChIP.

Chromatin immunoprecipitation (ChIP) analysis was performed according to the Upstate Biotechnology protocol with modifications. For histone modifications, 5 × 10⁵ cells were treated with 0.37% formaldehyde (wt/vol) at 37°C for 10 min. For ChIP of LEF/TCFs or β-catenin, 2 × 10⁶ cells were treated with 1% formaldehyde (wt/vol). Cross-linking was stopped with the addition of 125 mM glycine for 10 min at room temperature. Cross-linked cells were centrifuged at 1,200 rpm for 10 min at 4°C and washed with 1× phosphate-buffered saline with protease inhibitors and 1 mM phenylmethylsulfonyl fluoride and subsequently resuspended in sodium dodecyl sulfate lysis buffer (Upstate) and incubated on ice for 10 min. Cross-linked DNA was sheared by sonication to an average size of between 350 and 800 nucleotides, diluted 1:10 with ChIP dilution buffer (Upstate), and then precleared with 50% protein A-agarose gel slurry, 80 μg of sonicated salmon sperm, and 133.3 μg of BSA for 2 h at 4°C. Precleared DNA was added to immunoprecipitation reaction mixtures with 10 μg of an antibody to acetylated histone H3 (Upstate), 50 μl of panspecific polyclonal LEF/TCF antibody, or 30 μg of polyclonal β-catenin antibody (Santa Cruz Biotechnology) overnight at 4°C. Ten micrograms of a normal rabbit IgG antibody (Santa Cruz Biotechnology) was used for a negative control immunoprecipitation. The following day, 200 μl of 50% protein A-agarose gel slurry with 80 μg of sonicated salmon sperm and 133.3 μg of BSA was added to both sets of samples and incubated at 4°C for 2 h. Beads were collected by centrifugation at 900 × g for 5 min and then washed according to the Upstate ChIP protocol. Immunoprecipitates were eluted from the beads with 500 μl of elution buffer (1% NaHCO₃, 1% sodium dodecyl sulfate in Tris-EDTA) at 65°C for 15 min. Cross-links were reversed with 30 μl of 5 M NaCl (283 mM final) and incubated at 75°C overnight. DNA was treated with 40 μg of proteinase K and 20 μl of Tris-HCl, pH 6.5, for 2 h at 45°C and subsequently phenol-chloroform extracted and ethanol precipitated. The primers used to survey the 5.5 kb of the LEF1 locus spanning from promoter 1 to promoter 2 are shown below (see Fig. 3 and 4) (sequences available upon request). The amount of input was chosen based on PCR titrations with input sample to establish the linear range for each primer pair and extract (data not shown). Band intensities for all PCRs were determined using the Bio-Rad Gel Doc system and Quantity I software supplied by the manufacturer. The intensities of the PCR products are presented either as a ratio over the intensity obtained from total input supplied to each immunoprecipitation reaction (+Ab/input; used for analysis of histone acetylation) or as a ratio of the PCR product intensities obtained from immunoprecipitations with the specific antibody versus same-species IgG (used for analysis of LEF/TCF and β-catenin binding).

Stable cell lines.

DLD1 and Colo 320 colon cancer cells were used to generate stable cell lines with integrated promoter 2 luciferase reporter constructs. Cells (2 × 10⁶) were transfected with four promoter 2-luciferase reporter plasmids (2 μg; −177/+60, R-177/+60, −816/+60, or −1446/+60), a CMV-beta-galactosidase reporter (0.4 μg), and an RSV-neo plasmid (0.2 μg) by use of Effectene transfection reagent (QIAGEN). Stably transfected DLD1 cells were selected in media containing 400 μg/ml of G418 (A.G. Scientific), and Colo 320 cells were selected with 800 μg/ml of G418. G418-resistant colonies were pooled 3 weeks after transfection and thereafter cells were maintained in media containing 100 μg/ml of G418. Approximately 800 DLD1 colonies and 30 Colo 320 colonies were pooled. For luciferase assays for stable cell lines, 400,000 DLD1 cells or 200,000 Colo 320 cells were seeded in six-well plates 24 h prior to harvest for luciferase and beta-galactosidase assays. The luciferase activities were normalized to the beta-galactosidase activities by use of the above-mentioned harvest and assay methods. For ChIP assays for stable cell lines, 6 × 10⁶ cells were seeded in 15-cm dishes 24 h prior to being cross-linked with 1% formaldehyde at room temperature for 10 min and further processing as described above. Cells from one whole dish were used for each condition (input, normal goat IgG, polyclonal LEF/TCF antibody, or polyclonal goat β-catenin antibody). Immunoprecipitated DNA was resuspended in 50 μl of water, and 1 μl was used for PCRs. Input DNA was resuspended in 200 μl of water, and 1 μl was used for the PCRs. Fifty microliters of the polyclonal LEF/TCF antibody, 20 μg of the polyclonal goat β-catenin antibody (C-18; Santa Cruz Biotechnology), and 20 μg of the normal goat IgG (sc-2028; Santa Cruz Biotechnology) were used in each experiment.

LEF1 expression in primary colon cancer.

We previously reported that LEF1 mRNA could be detected by in situ hybridization analysis of 10 human primary colon carcinomas (18). To determine if this expression is frequent and thus a reliable marker of primary colon carcinoma, we probed a cancer array blot in which cDNA samples from 35 matched normal and colon tumor tissues were spotted side by side (Fig. 1C). Human LEF1 cDNA and human ubiquitin cDNA probes were separately hybridized to the array so that ubiquitin signals could be used to normalize LEF1 signals. Normalized signal intensity values were compared between matched samples and the results plotted as elevation (n-fold) of expression in tumor relative to that in normal tissue (Fig. 1C). Of the 35 tumors, 29 (83%) exhibited an increase of LEF1 expression of at least twofold. Eleven of the 29 samples had expression changes of greater than fivefold. Since previous in situ hybridization results showed epithelial cells of normal colon to be negative (or below the limit of detection) for LEF1 mRNA and transformed epithelial cells to be positive for LEF1, we conclude that LEF1 expression is aberrantly activated from a normally silent locus in the majority of sporadic colon cancers. Interestingly, the overall frequency of expression detected in these cancers is similar to the frequency with which sporadic colon cancer harbors loss of heterozygosity of APC and thus unregulated Wnt signaling (∼80%) (8, 30, 32).

The LEF1 locus is not rearranged in colon cancer cells.

Aberrant expression of genetic loci in cancer arises from multiple mechanisms, including gene rearrangement, amplification, and loss of proper gene regulation. Indeed, aberrant differential promoter usage in the CMYC locus is due to relocation of the locus to the IgH enhancer and to gene amplification (27). The human LEF1 gene is located at chromosome 4q23-25, a region not known to be a hot spot for rearrangements in colon cancer. Nevertheless, we explored the possibility that aberrant P1 activation is due to chromosomal rearrangement. With a large LEF1 genomic fragment from the human gene, FISH analysis of metaphase chromosome spreads was used to determine whether the LEF1 locus is grossly amplified or its chromosomal region rearranged in human colon cancer cell lines (Fig. 2). A chromosome 4 paint probe was used to check for chromosome integrity. For comparison, metaphase spreads of normal human fibroblasts were probed because the LEF1 gene should be normal and diploid in these cells (GM5399-E6) (Fig. 2A). The results show that the human LEF1 locus is diploid and not rearranged in two human colon cancer cell lines that express LEF1 mRNA and that in a third human colon cancer cell line (HT-29), the LEF1 gene is triploid due to a partial third chromosome 4 (Fig. 2B to D). Even though the LEF1 gene is triploid in HT-29 cells, it is nevertheless not expressed very well (weak expression is detectable by reverse transcription-PCR amplification only) (L. Arce, unpublished data). Thus, in at least two colon cancer cell lines where LEF1 expression is easily detected, the locus is diploid and not grossly rearranged or amplified. Based on these results, we conclude that LEF1 expression in colon cancer is due to aberrant gene regulation and not to large-scale damage of the locus.

LEF1 is a Wnt target in colon cancer cells.

Since our expression analyses show that P1 is activated in colon cancer at a frequency similar to the frequency with which mutations of Wnt signaling components are detected, there may be a link between activated Wnt signaling and misregulation of LEF1 P1. Indeed, we have previously reported that LEF1 P1 can be activated ∼14-fold by TCF-β-catenin complexes in transient transfections and we have identified LEF/TCF binding sites at the promoter and downstream at +190 and +283 (Fig. 1A) (3, 18). Another group has also delimited a Wnt-responsive element in the LEF1 promoter, and this element is important for in vivo expression in some tissues (11, 13). To test whether the endogenous LEF1 locus reflects the Wnt pathway regulation implied by the transient transfections, we used DLD1 colon cancer cells modified with a tetracycline-inducible dnTCF system. Overexpression of dnTCF protein efficiently suppresses expression of Wnt target genes by displacing TCF-β-catenin complexes from Wnt response elements in the genome (5, 43). Figure 3A shows that at 2 and 4 h after the addition of doxycycline to the media, DLD1 cells express low but detectable amounts of dnTCF-1. Surprisingly, even with this small amount of dnTCF-1 protein, a decrease in LEF1 mRNA (P1 and P3 derived) is detected. After 8 hours of induction, LEF1 gene expression is essentially shut off. No change is seen in the levels of GAPDH mRNA, suggesting that during the course of the experiment, the cells are viable and the effect of dnTCF-1 is specific. The rapid change in LEF1 mRNA suggests that P1 (as well as P3) is a direct target of the Wnt pathway and is activated by the endogenous pool of TCF-β-catenin complexes that are abundant in colon cancer cells.

TCF-β-catenin occupancy of endogenous promoter 1 and promoter 2.

We used ChIP assays to assess whether TCF-β-catenin complexes are directly bound to LEF1 and to locate their sites of occupancy (Fig. 3B). Antibodies specific for LEF/TCFs or β-catenin were used with extracts from two colon cancer cell lines that express LEF-1 (Colo 320, DLD1). For the analysis, 11 oligonucleotide primer pairs were designed for PCR amplification of LEF1 genomic sequences from the immunoprecipitates. These primer pairs were designed such that they could amplify tandem regions of the locus to cover both P1 and P2 and the entire 5.5-kb region in between (Fig. 3B). Sonicated extracts from formaldehyde-cross-linked cells were incubated with antibodies for LEF/TCFs, β-catenin, or rabbit serum IgG. Amplification of LEF1 sequences from the immunoprecipitates was performed under conditions where the cycle number and the amount of template assured product generation in the linear range (data not shown). In addition, the entire experiment was repeated a total of four times with independently prepared extracts. The representative results shown in Fig. 3B indicate that LEF/TCF-β-catenin complexes are bound to one or more of the LEF/TCF binding sites in P1 in both DLD1 and Colo 320 cells. A region in intron 1 also contains TCF-β-catenin in Colo 320 cells and TCF in DLD1 cells. In contrast, no TCF-β-catenin complexes were detected beyond exon 2.

Dynamic, β-catenin-dependent acetylation of LEF1 chromatin.

Myriad proteins are known to associate with β-catenin and to be important for transcription activation; these proteins include components of chromatin-modifying or -remodeling complexes (4, 6, 7, 16, 22, 35, 38, 40). One of these proteins is p300, a well-known histone acetyltransferase (HAT) that can covalently modify the tails of nucleosomal histones H3 and H4. The HAT activities of p300 have been shown to be involved in the activation of some Wnt target genes (16, 38). Others have shown using colon cancer cells that β-catenin recruits ATP-dependent chromatin-remodeling complexes to influence histone acetylation (12). Given these links between β-catenin and chromatin acetylation, we used the ChIP assay to determine the extent of histone H3 and H4 acetylation over P1 and P2 and then assessed the degree to which this acetylation was dependent upon recruited β-catenin (Fig. 4 and 5).

The acetylation profiles of LEF1 P1 and P2 in colon cancer cells where only P1 is active (DLD1; colon cancer) were analyzed and compared to profiles for cells where both promoters are active (Jurkat; mature T lymphocyte) as well as to those where they are mostly inactive (HT-29; colon cancer). Acetylation was detected with polyclonal antibodies that recognize singly or multiply acetylated histone H3 (Fig. 4) or histone H4 (data not shown). The results show that as expected, an inactive LEF1 locus is not acetylated at any point along the approximately 5.5-kb region that spans both promoters (HT-29), whereas acetylation is readily detected over both P1 and P2 in Jurkat cells. The pattern for histone H4 acetylation is identical to that for histone H3 (data not shown). While acetylation is detected over both promoters, the level of histone H3 acetylation over P2 (20% of input) is significantly lower than that over P1 (60% of input). We conclude that the ChIP assay was detecting bona fide euchromatin acetylation within this part of the LEF1 locus and that such modifications are generally correlated with LEF1 transcription activity. For the DLD1 colon cancer cell line where only promoter 1 is actively transcribed, the pattern of histone acetylation was different from that seen in Jurkat cells (Fig. 4). Histone H3 acetylation of P1 in DLD1 cells was markedly high (60 to 100% of input), but acetylation dropped abruptly near the intron 1/exon 2 boundary and remained low, but significant, over the silent P2.

To test whether acetylated P1 chromatin is dependent upon continual β-catenin recruitment, we induced expression of dnTCF in DLD1 cells by adding doxycycline to the media (Fig. 5). After 4 or 24 h of treatment, cells were cross-linked and processed for immunoprecipitation with acetylated histone H3-specific antibodies. If TCF-β-catenin recruits coactivators with histone acetyltransferase activities such as p300 or CBP, then removal of such complexes by competition with the repressive dnTCF-1 should alter the acetylation profile. Figure 5 shows that induction of dnTCF-1 rapidly and dramatically lowers acetylation over P1 at least three- to fivefold as early as 4 hours after the addition of doxycycline. Acetylation over the GAPDH promoter is unchanged, showing that the effect is specific (Fig. 5, inset). Acetylation over P2 is lowered slightly but because acetylation levels are low to begin with, the magnitude of the decrease is not statistically significant. These data show that histone acetylation over the LEF1 locus in colon cancer cells is inextricably linked to transcription of LEF1 P1 and that the TCF-β-catenin complexes detected by ChIP are functional and active. We conclude that promoter 1 transcription requires constant, continual recruitment of β-catenin and its associated HAT activities to Wnt response elements (+191, +283).

Activation and repression of P2.

One possibility for the striking differential activation of P1 and P2 in colon cancer is that P1 is a target of the Wnt pathway but P2 is not. Indeed, there are no TCF-β-catenin complexes binding beyond intron 1 in colon cancer cells (Fig. 3B). Alternatively, since both promoters appear to be coordinately expressed in normal cells, it is possible that P2 is also a target of the Wnt pathway but is not accessible or responsive to TCF-β-catenin complexes in colon cancer. In order to test these two models of misregulation, it was necessary to delimit P2, including its transcription start site(s) and basal promoter. To localize P2, a set of nested deletions containing exon 1 through exon 3 was subcloned into a luciferase reporter plasmid for transient transfection analysis in three different cell lines. Figure 6A shows that regardless of which cell line is used, a basal promoter can be delimited to a region near the intron 2/exon 3 boundary. Finer deletions and mutations within an SV40 enhancer plasmid backbone defined a minimal TATA-less promoter (Fig. 6B). Transcription start site analysis (Fig. 7A) determined that the main start site occurs 10 nucleotides upstream of the intron 2/exon 3 boundary; thus, the minimal promoter fragment in Fig. 6A is shown as −27/+60. In brief, P2 contains multiple transcription start sites that depend upon an Initiator-like element; a second Initiator and a weakly conserved TATA box are not necessary, and their mutation or deletion had no effect on promoter activity (Fig. 6B). DNase I footprinting with recombinant LEF-1 protein showed that there are TCF/LEF binding sites near the basal promoter (Fig. 7B), with the strongest binding sites located at −216, −33, and −3. Thus, the footprinting assay suggests that P2 carries potential for Wnt regulation. Figure 7B summarizes all of these findings and shows by comparison with the minimal P1 promoter that the two basal promoters share some similar features. Both are TATA-less promoters with TCF/LEF binding sites located at proximal sites.

During the course of our promoter analysis, we noted that P2 promoter activity was strikingly high in the colon cancer cell line (Colo 320) compared to its more modest activities in a T-lymphocyte line (Jurkat) and a monkey kidney cell line (COS-1) (Fig. 6A). Thus, even though endogenous P2 is silent in colon cancer cells, there are specific factors present in the nucleus sufficient to make transfected, episomal P2 a highly active promoter for transcription. The strong promoter activity in colon cancer cells could be due to a colon-specific set of factors that recognize P2 and/or the aberrantly high level of TCF-β-catenin complexes that might bind to mapped LEF/TCF binding sites in the promoter (TCF-β-catenin complexes are not abundant in COS-1 and Jurkat cells). However, we also noted that high promoter activity was evident only in smaller promoter fragments (−816/+60, −27/+60) and not in larger, 5′-extended fragments that contain a complete exon 2 and intron 2 (although the colon cancer-specific activity is masked in plasmids that contain an SV40 enhancer [Fig. 6B]). Inclusion of these upstream sequences appeared to dramatically reduce promoter activity and prevent factors from activating this promoter. We devised two sets of experiments to determine whether P2 might be repressed by upstream sequences and whether such repression might override regulation by the Wnt pathway or other colon-specific factors (Fig. 8 and 9).

To determine whether the LEF/TCF binding sites in promoter 2 could function as WREs, a transient cotransfection analysis was performed with COS-1 cells with P2 reporter plasmids and expression vectors for TCF-1 and β-catenin (Fig. 8A). The results show that P2 is activated by TCF-β-catenin complexes but that the best activation is observed with smaller promoter fragments, such as −816/+60 and −27/+60 (12-fold and 10-fold, respectively). These activation levels are similar to those found in our previous reports of TCF-β-catenin activation of LEF1 promoter 1 (see also LEF1P1 in Fig. 8B) (3, 18). The fact that P2 can be activated by TCF-β-catenin complexes suggests a similarity for P1 and P2 that goes beyond their basal regulatory elements and into their potential for regulation by the Wnt pathway. These data also provide at least part of the explanation for the very high activity levels of transfected P2 plasmids in colon cancer cells in which there are high levels of TCF-β-catenin complexes versus the modest activity of transfected P2 plasmids in Jurkat T lymphocytes and COS-1 cells. Clearly, all of the factors and elements, including activation by the Wnt pathway, are present and available for P2 to be as active as P1 in colon. Despite this potential, P2 is a relatively weak promoter when upstream intron 2 sequences from −1446 and beyond are present (Fig. 6A and 8A). Even when TCF and β-catenin are overexpressed, they activate P2 only weakly (e.g., −3004/+60 is activated only 2.9-fold) (Fig. 8A). This suggests that upstream sequences carry a repressive activity that prevents TCF-β-catenin complexes from fully activating P2.

The repressor region was delimited to a small stretch of conserved nucleotides spanning the exon 2/intron 2 boundary between −1446 and −1281 (Fig. 8B). Fusion of these 165 nucleotides to −177 of P2 almost completely repressed basal activity. In Colo 320 cells, the 165-nucleotide repressor sequence reduced P2 activity 18-fold (levels of −177/+60 and R-177/+60 were 81.3-fold and 4.4-fold, respectively, above that for the empty vector; data not shown). Clearly, the repressor is able to work equally well when far upstream or when juxtaposed to the basal promoter. This region was also tested for its ability to repress P1 (LEF1P1 constructs in Fig. 8B). Placement of the 165-nucleotide fragment upstream (−672) of the main transcription start site of P1 in either orientation repressed basal P1 by only 50% at most (Fig. 8B). Weak activity towards P1 suggests that the repressor is selective for P2 and targets a unique feature of the basal promoter rather than shared features such as TCF-β-catenin complexes. The isolated repressor was also tested for its ability to prevent TCF-β-catenin from activating P1 or P2 (Fig. 8B). P2 was strongly activated by TCF-1-β-catenin, but when the repressor was present, activation was much reduced (Fig. 8B, compare the 85-fold activation level for −177/+60 with the 13-fold activation level for R-177/+60). In contrast, the repressor did not prevent activation of P1 but in fact appeared to enhance activation when fused to the promoter in the forward orientation (25-fold for LEF1P1, 31-fold and 56-fold for RLEFP1) (Fig. 8B). Taken together, these data define and localize a strong P2-specific repressor that may prevent transcription in colon cancer by suppressing the basal promoter rather than TCF-β-catenin complexes.

P2 regulation in the colon cancer genome.

A repressor that acts upon a basal promoter could prevent TCF-β-catenin complexes from activating the promoter but not necessarily prevent their binding to Wnt response elements. However, since endogenous P2 is not occupied by TCF-β-catenin complexes in colon cancer cells at any of the four LEF/TCF sites identified (Fig. 3B), we devised experiments to test whether the repressor is also a cause of the vacant Wnt response elements in P2. We also wished to examine P2 promoter and repressor activity within the context of the colon cancer genome, since transient transfections assess promoter activity on episomal plasmids that are not likely to have authentic chromatin structures. Stable DLD1 and Colo 320 colon cancer cell lines were created with various P2 luciferase reporter plasmids integrated into the genome (Fig. 9). To create these stable cell lines, four different P2 reporter plasmids were each cotransfected along with separate plasmids expressing beta-galactosidase and a neomycin resistance gene, and G418-resistant colonies were pooled and expanded. The use of pooled stable integrants corrected for variations in reporter gene activity due to position effects from the site of integration, and the use of a cointroduced beta-galactosidase plasmid enabled luciferase activities to be normalized for comparison among the four stable cell lines. Integrated P2 reporter plasmids −816/+60 and −177/+60 were highly active in DLD1 and Colo 320s, although the small −177/+60 promoter was less active in DLD1 cells (Fig. 8A). Thus, even though endogenous P2 of the LEF1 locus is silent in colon cancer cells, the basal promoter can be quite active when taken out of its native context.

In contrast to what was seen for the active P2 reporter plasmids, inclusion of the 165-nucleotide repressor region inhibited P2 activity as efficiently as it did in the transient transfections. Both the large and small promoters had reduced activity when the repressor was present (compare −1446/+60 and R-177/+60 with −816/+60 and −177/+60), and with the exception of the small −177/+60 promoter in DLD1 cells, the magnitudes of P2 repression in both cell lines were approximately eightfold, a level similar to the levels of repression observed in the transient transfection assays (Fig. 6A and 8B). Interestingly, P2 activity was repressed to nearly identical levels for any of the constructs in both cell lines, suggesting that the repression mechanism lowers transcription output to a minimal basal level regardless of the number of Wnt response elements or other positive-acting regulatory elements that are present.

We also used chromatin immunoprecipitation assays of the stable DLD1 cell lines to assess the occupancy of the identified LEF/TCF binding sites in the absence or presence of the repressor (Fig. 9B). As before, antibodies to β-catenin and LEF/TCFs were used to immunoprecipitate sonicated chromatin from extracts of the formaldehyde-treated stable cells. PCR primers specific to the integrated P2 reporter plasmid were used to detect only the integrated P2 and not the silent endogenous P2, and negative controls included the use of IgG and PCR primers to other regions of the integrated reporter plasmid. The data are shown as the means of three independent ChIP experiments. Although the data exhibited some variation, particularly for β-catenin, the results consistently show that TCF-β-catenin complexes bind to unrepressed P2 integrants. Levels of occupancy for the large active promoter fragment (−816/+60) are higher than for the small active promoter construct (−177/+60), likely reflecting the fact that there is an additional strong LEF/TCF binding site at −216. We conclude that part of the P2 activity is due to TCF-β-catenin complexes that bind to LEF/TCF sites, and we propose that these sites function as bona fide Wnt response elements. In the presence of the repressor region, however, occupancy is much reduced. This is most evident for R-177/+60, where the level of TCF-β-catenin binding is low to undetectable. Thus, it appears that TCF-β-catenin complexes are not able to engage in stable association with a repressed P2. Two general mechanisms could account for these observations: one is that the repressor specifically acts upon TCF-β-catenin complexes and disallows binding, and the other is that the repressor acts upon a different promoter element to lower responsiveness or basal activity and thus destabilizes the association of TCF-β-catenin complexes with their binding sites. Our data suggest that the latter is the case, because between −1446/+60 and −177/+60, TCF-β-catenin occupancy levels are very similar but promoter activities are different (Fig. 9A and B). Furthermore, the repressor represses basal P1 only weakly and does not prevent activation by TCF-β-catenin complexes (Fig. 8B). If the repressor were acting directly upon TCF-β-catenin complexes to displace or otherwise silence them, one would expect a similar repressor effect on P1. We conclude that the repressor acts upon a specific promoter feature of P2 and that repression of this feature destabilizes TCF-β-catenin occupancy, leading to an overall reduction in transcription.

Active promoter 1.

We have shown conclusively that LEF1 P1 is a dynamically regulated Wnt target gene in colon cancer (Fig. 9C). Transcription from this promoter is completely dependent on TCF-β-catenin complexes because expression of dnTCF-1 rapidly shuts down transcription of the endogenous locus and most of the chromatin acetylation (Fig. 3 and 5). Normally, LEF1 expression is detected in a broad spectrum of cells and tissues during embryonic development, but its postnatal expression is more restricted to cycling and differentiating cells in tissues such as thymus, bone marrow, and hair follicles (29, 44, 46). Wnt signals are important for proper lymphocyte and hair follicle development in adults, and we propose that LEF1 expression is influenced by Wnts in these tissues. Other groups have examined Wnt regulation of LEF1 in vitro and in vivo (11, 13, 26). A 2.5-kb fragment of the LEF1 promoter is responsive to Wnt1 and Wnt3a in 293 cells, and a novel Wnt response element was delimited in this context. This element does not contain LEF/TCF binding sites but is nevertheless involved in LacZ marker expression in transgenic mice in several types of Wnt-responsive tissue where endogenous LEF1 is expressed (e.g., mesenchymal cells of hair follicles, mammary gland) (11, 13, 26). It is not known whether this element mediates Wnt regulation directly or indirectly, and the element is not necessary for the TCF-β-catenin activation we observe or for activation by LEF-1-β-catenin-PITX2 complexes (3, 18, 41). Therefore, at least two modes of Wnt regulation have been uncovered, and it is possible that Wnt regulation of LEF1 occurs via these different modes in a tissue-specific fashion. Our ChIP analysis focused on regions surrounding P1 and detected high levels of TCF-β-catenin occupancy at downstream elements (+191, +283) (Fig. 3). We propose in our model of regulation that these sites, which match canonical Wnt response elements and are protected by LEF-1 and TCF-1 protein in DNase I footprinting assays, are primarily responsible for LEF1 transcription in colon cancer cells (Fig. 9C). It is possible that other regulatory elements or distal regions are involved in LEF-1 expression in cancer because even though epithelial cells of colon crypts are directed by Wnt signals for proper cell growth and differentiation, LEF1 transcription is not activated by these signals and mRNA is not detected in normal cells. It is clearly the case that in colon the LEF1 locus is silenced to such a degree that it is not accessible for TCF-β-catenin occupancy. When and how the LEF1 locus becomes accessible to aberrant Wnt signals in colon cancer are an important problem to understand.

Silent promoter 2.

Our ChIP assays detect heavy acetylation of the chromatin over P1 in colon cancer cells, but they also show that silent P2 chromatin is modestly acetylated. In fact, the level of acetylation over P2 is not much different from that of the same region in Jurkat cells where P2 is active (Fig. 4). Most or all of this acetylation is dependent upon TCF-β-catenin complexes binding to sites in P1 and intron 1 because induction of dnTCF-1 causes a rapid three- to fivefold drop in acetylation levels over the P1/P2 region (Fig. 5). It is possible that the low level of acetylation over P2 is derived from elongating RNA polymerase II complexes that initiate upstream at P1 in response to β-catenin recruitment at promoter 1 and intron 1. HAT activities that travel with elongating forms of RNA polymerase II have been identified for Saccharomyces cerevisiae (Elp3) and Drosophila melanogaster (TAC1) (37, 45). Despite this acetylation and despite the fact that P2 is transcribed through by promoter 1-initiated RNA polymerase II-type polymerases, the promoter remains silent and TCF-β-catenin complexes do not bind to mapped elements. Thus, P2 is clearly a separately regulated domain in cancer. We propose that P1 and P2 are both targeted by Wnt signaling in normal cells for expression, and consistent with this, we detect low but significant occupation of LEF/TCFs over P1 and P2 in a lymphocyte cell line where both promoters are active (A. Bernstein, unpublished data). If P1 and P2 are normally coexpressed, the loss of this coordinate control might be due to the repressor identified here—a clear issue of epigenetic regulation.

Several different types of regulatory elements can impose differential regulation on closely linked promoters. Boundary elements or insulators are regulatory elements that separate different domains for regulation, and they lie at the sharp boundaries between active and silent chromatin. Indeed, the acetylation levels between P1 and P2 change sharply near the region where the P2 repressor has been delimited (Fig. 4 and 5). However, our data suggest it is unlikely that an insulator/boundary element exists within the delimited 165-nucleotide repressor region, because (i) active repression can be achieved on transiently transfected plasmids and (ii) repression occurs in the absence of any upstream enhancer, negating two of the criteria for assigning insulator/boundary function. In addition, we do not detect inhibitory methylation of histones (e.g., histone H3-K9) that correlate with an inactive chromatin environment anywhere along the 5.5-kb region, nor do we detect any significant binding of CTCF, a well-characterized insulator factor (T. Li and N. Yokoyama, unpublished observation). While our data do not completely rule out the possibility of an insulator element, they suggest that the sharp drop in acetylation could be due to skewed, polar actions of the repressor for P2.

There are numerous other ways in which P2 could be repressed and made unresponsive to the Wnt pathway. For example, TCF-β-catenin complexes may not be able to bind because repressive chromatin obscures binding sites or because transcription activators required by TCF for cooperative binding to sites in P2 may be missing. Alternatively, TCFs, which are known to recruit corepressors such as CtBP or TLE (Groucho), could be engaged in active repression by exchanging β-catenin for these coregulators in a repressor-specified way. As a final possibility, P2 may be actively repressed by the distal repressor region in a more direct way, such as by the basal promoter. Our data rule out all of these possibilities except the last. A missing cooperative factor in colon cancer cells is unlikely, because endogenous TCF-β-catenin complexes bind to transiently transfected as well as genome-integrated P2 reporter plasmids, resulting in a very high level of promoter activity. Also, it is unlikely that TCF-recruited TLE or CtBP repressors are involved, because we detected almost no TCF binding to P2 at its most repressed level. Finally, it is unlikely that repressive chromatin structures occlude P2 Wnt response elements, because repression occurs on transiently transfected plasmids which are not carrying authentic chromatin structures. Instead, our data support a model of active primary repression of the basal promoter and an interesting secondary loss of TCF-β-catenin binding (Fig. 9C). Notably, the upstream repressor region inhibits P2 reporter activity in cells that do not contain nuclear TCF-β-catenin complexes (Fig. 6). Also, the repressor region only modestly represses promoter 1 and does not prevent activation by TCF-β-catenin complexes (Fig. 8B). We conclude that the lower level of TCF-β-catenin complexes binding to P2 when the repressor is present is a result, rather than a cause, of promoter repression. Clearly, one effect can reinforce the other and so there are at least two contributing factors to a silent promoter 2: active repression of the basal promoter and the absence of TCF-β-catenin complexes. It is interesting to note that TCF-β-catenin complexes do not bind to their sites when P2 is repressed, because this observation runs counter to the general notion that LEF/TCFs remain bound to Wnt target genes and silence them by recruiting transcription repressors. A comprehensive survey of TCF-β-catenin and TLE or CtBP occupancy of Wnt target genes in their active and silent states will be an important test of this generally accepted model.

Alternative promoters and differential expression in cancer.

The results reported here contribute to the understanding of dual promoter genes and illustrate how two promoters that produce polypeptides with important functional differences are differentially regulated in an abnormal setting such as cancer. The first promoter (LEF1 P1) is ectopically activated by the very signal transduction pathway that it works in. The second promoter (LEF1 P2), which produces a polypeptide that suppresses the signal transduction pathway, carries the same potential for regulation by the cognate pathway but is nevertheless untouched in cancer and is actively repressed. The result is the selective expression of an isoform that promotes greater signal strength in the nucleus.

Of the list of characterized dual promoter genes for transcription factors, the p53-related p73 and p63 have genetic structures similar to those of LEF1 and TCF7 (20). Alternative promoters in the third intron produce N-terminally truncated polypeptides missing the transcription activation domain (ΔNp73, ΔNp63). ΔNp73 and ΔNp63 function as dominant negatives, as they are incorporated into p53, p63, and p73 tetramers and prevent activation of target genes. Full-length p53 and p73 activate transcription of this internal promoter, establishing an important negative regulatory feedback loop (15, 28). Interestingly, like that of the LEF1 locus, the pattern of p73 transcription is aberrant in cancer. In this case, the intronic promoter for ΔNp73 is greatly upregulated over that of the first promoter for full-length p73 to the point where overproduction of ΔNp73 suppresses the tumor suppressor functions of full-length, wild-type p53 and p73 (51). In the case of colon cancer, selective expression of full-length LEF1 may upset the careful balance between TCF-4 and dnTCF-1, and this may contribute to colon carcinogenesis. Indeed, LEF-1 can transform normal chicken embryo fibroblasts when its full-length open reading frame is fused to transcription activation domains (1, 2).

It is unlikely that LEF1, p73, p63, and CMYC are the only genes in which switches in promoter activity profiles may contribute to cancer. A recent in silico analysis of approximately 20,000 human gene loci and associated expressed sequence tag sequences led the authors of that study to estimate that ∼18% of all human genes contain alternative promoters (25). Just like alternative splicing, alternative promoters are an important genetic feature that contributes to protein and regulatory diversity and may be exploited in aberrant settings such as cancer.

• 1.Aoki, M., A. Hecht, U. Kruse, R. Kemler, and P. K. Vogt. 1999. Nuclear endpoint of Wnt signaling: neoplastic transformation induced by transactivating lymphoid-enhancing factor 1. Proc. Natl. Acad. Sci. USA 96:139-144. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 2.Aoki, M., V. Sobek, D. J. Maslyar, A. Hecht, and P. K. Vogt. 2002. Oncogenic transformation by beta-catenin: deletion analysis and characterization of selected target genes. Oncogene 21:6983-6991. [DOI] [PubMed] [Google Scholar]
• 3.Atcha, F. A., J. E. Munguia, T. W. Li, K. Hovanes, and M. L. Waterman. 2003. A new beta-catenin-dependent activation domain in T cell factor. J. Biol. Chem. 278:16169-16175. [DOI] [PubMed] [Google Scholar]
• 4.Barker, N., A. Hurlstone, H. Musisi, A. Miles, M. Bienz, and H. Clevers. 2001. The chromatin remodelling factor Brg-1 interacts with beta-catenin to promote target gene activation. EMBO J. 20:4935-4943. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 5.Batlle, E., J. T. Henderson, H. Beghtel, M. M. van den Born, E. Sancho, G. Huls, J. Meeldijk, J. Robertson, M. van de Wetering, T. Pawson, and H. Clevers. 2002. β-Catenin and TCF mediate cell positioning in the intestinal epithelium by controlling the expression of EphB/EphrinB. Cell 111:251-263. [DOI] [PubMed] [Google Scholar]
• 6.Bauer, A., S. Chauvet, O. Huber, F. Usseglio, U. Rothbacher, D. Aragnol, R. Kemler, and J. Pradel. 2000. Pontin52 and reptin52 function as antagonistic regulators of beta-catenin signalling activity. EMBO J. 19:6121-6130. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 7.Bauer, A., O. Huber, and R. Kemler. 1998. Pontin52, an interaction partner of beta-catenin, binds to the TATA box binding protein. Proc. Natl. Acad. Sci. USA 95:14787-14792. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 8.Bienz, M., and H. Clevers. 2000. Linking colorectal cancer to Wnt signaling. Cell 103:311-320. [DOI] [PubMed] [Google Scholar]
• 9.Brantjes, H., J. Roose, M. van De Wetering, and H. Clevers. 2001. All TCF HMG box transcription factors interact with Groucho-related co-repressors. Nucleic Acids Res. 29:1410-1419. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 10.Chuvpilo, S., E. Jankevics, D. Tyrsin, A. Akimzhanov, D. Moroz, M. K. Jha, J. Schulze-Luehrmann, B. Santner-Nanan, E. Feoktistova, T. Konig, A. Avots, E. Schmitt, F. Berberich-Siebelt, A. Schimpl, and E. Serfling. 2002. Autoregulation of NFATc1/A expression facilitates effector T cells to escape from rapid apoptosis. Immunity 16:881-895. [DOI] [PubMed] [Google Scholar]
• 11.Driskell, R. R., X. Liu, M. Luo, M. Filali, W. Zhou, D. Abbott, N. Cheng, C. Moothart, C. D. Sigmund, and J. F. Engelhardt. 2004. Wnt-responsive element controls Lef-1 promoter expression during submucosal gland morphogenesis. Am. J. Physiol. Lung Cell. Mol. Physiol. 287:L752-L763. [DOI] [PubMed] [Google Scholar]
• 12.Feng, Y., N. Lee, and E. R. Fearon. 2003. TIP49 regulates beta-catenin-mediated neoplastic transformation and T-cell factor target gene induction via effects on chromatin remodeling. Cancer Res. 63:8726-8734. [PubMed] [Google Scholar]
• 13.Filali, M., N. Cheng, D. Abbott, V. Leontiev, and J. F. Engelhardt. 2002. Wnt-3A/beta-catenin signaling induces transcription from the LEF-1 promoter. J. Biol. Chem. 277:33398-33410. [DOI] [PubMed] [Google Scholar]
• 14.Gregorieff, A., R. Grosschedl, and H. Clevers. 2004. Hindgut defects and transformation of the gastro-intestinal tract in Tcf4(-/-)/Tcf1(-/-) embryos. EMBO J. 23:1825-1833. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 15.Grob, T. J., U. Novak, C. Maisse, D. Barcaroli, A. U. Luthi, F. Pirnia, B. Hugli, H. U. Graber, V. De Laurenzi, M. F. Fey, G. Melino, and A. Tobler. 2001. Human delta Np73 regulates a dominant negative feedback loop for TAp73 and p53. Cell Death Differ. 8:1213-1223. [DOI] [PubMed] [Google Scholar]
• 16.Hecht, A., K. Vleminckx, M. P. Stemmler, F. van Roy, and R. Kemler. 2000. The p300/CBP acetyltransferases function as transcriptional coactivators of beta-catenin in vertebrates. EMBO J. 19:1839-1850. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 17.Henriksen, M. A., A. Betz, M. V. Fuccillo, and J. E. Darnell, Jr. 2002. Negative regulation of STAT92E by an N-terminally truncated STAT protein derived from an alternative promoter site. Genes Dev. 16:2379-2389. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 18.Hovanes, K., T. W. Li, J. E. Munguia, T. Truong, T. Milovanovic, J. Lawrence Marsh, R. F. Holcombe, and M. L. Waterman. 2001. Beta-catenin-sensitive isoforms of lymphoid enhancer factor-1 are selectively expressed in colon cancer. Nat. Genet. 28:53-57. [DOI] [PubMed] [Google Scholar]
• 19.Hovanes, K., T. W. H. Li, and M. L. Waterman. 2000. The human LEF-1 gene contains a promoter preferentially active in lymphocytes and encodes multiple isoforms derived from alternative splicing. Nucleic Acids Res. 28:1994-2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 20.Irwin, M. S., and W. G. Kaelin. 2001. p53 family update: p73 and p63 develop their own identities. Cell Growth Differ. 12:337-349. [PubMed] [Google Scholar]
• 21.Jimenez, J., G. M. Jang, B. L. Semler, and M. L. Waterman. 2005. An internal ribosome entry site mediates translation of lymphoid enhancer factor-1. RNA 11:1385-1399. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 22.Kim, J. H., B. Kim, L. Cai, H. J. Choi, K. A. Ohgi, C. Tran, C. Chen, C. H. Chung, O. Huber, D. W. Rose, C. L. Sawyers, M. G. Rosenfeld, and S. H. Baek. 2005. Transcriptional regulation of a metastasis suppressor gene by Tip60 and beta-catenin complexes. Nature 434:921-926. [DOI] [PubMed] [Google Scholar]
• 23.Korinek, V., N. Barker, P. Moerer, E. van Donselaar, G. Huls, P. J. Peters, and H. Clevers. 1998. Depletion of epithelial stem-cell compartments in the small intestine of mice lacking Tcf-4. Nat. Genet. 19:379-383. [DOI] [PubMed] [Google Scholar]
• 24.Korinek, V., N. Barker, P. Morin, D. van Wichen, R. de Weger, K. Kinzler, B. Vogelstein, and H. Clevers. 1997. Constitutive transcriptional activation by a beta-catenin-Tcf complex in APC-/- colon carcinoma. Science 275:1784-1787. [DOI] [PubMed] [Google Scholar]
• 25.Landry, J. R., D. L. Mager, and B. T. Wilhelm. 2003. Complex controls: the role of alternative promoters in mammalian genomes. Trends Genet. 19:640-648. [DOI] [PubMed] [Google Scholar]
• 26.Liu, X., R. R. Driskell, M. Luo, D. Abbott, M. Filali, N. Cheng, C. D. Sigmund, and J. F. Engelhardt. 2004. Characterization of Lef-1 promoter segments that facilitate inductive developmental expression in skin. J. Investig. Dermatol. 123:264-274. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 27.Marcu, K. B., A. J. Patel, and Y. Yang. 1997. Differential regulation of the c-MYC P1 and P2 promoters in the absence of functional tumor suppressors: implications for mechanisms of deregulated MYC transcription. Curr. Top. Microbiol. Immunol. 224:47-56. [DOI] [PubMed] [Google Scholar]
• 28.Nakagawa, T., M. Takahashi, T. Ozaki, K. Watanabe Ki, S. Todo, H. Mizuguchi, T. Hayakawa, and A. Nakagawara. 2002. Autoinhibitory regulation of p73 by ΔNp73 to modulate cell survival and death through a p73-specific target element within the ΔNp73 promoter. Mol. Cell. Biol. 22:2575-2585. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 29.Oosterwegel, M., M. van de Wetering, J. Timmerman, A. Kruisbeek, O. Destree, F. Meijlink, and H. Clevers. 1993. Differential expression of the HMG box factors TCF-1 and LEF-1 during murine embryogenesis. Development 118:439-448. [DOI] [PubMed] [Google Scholar]
• 30.Polakis, P. 2000. Wnt signaling and cancer. Genes Dev. 14:1837-1851. [PubMed] [Google Scholar]
• 31.Porfiri, E., B. Rubinfeld, I. Albert, K. Hovanes, M. Waterman, and P. Polakis. 1997. Induction of a beta-catenin-LEF-1 complex by wnt-1 and transforming mutants of beta-catenin. Oncogene 15:2833-2839. [DOI] [PubMed] [Google Scholar]
• 32.Powell, S. M., N. Zilz, Y. Beazer-Barclay, T. M. Bryan, S. R. Hamilton, S. N. Thibodeau, B. Vogelstein, and K. W. Kinzler. 1992. APC mutations occur early during colorectal tumorigenesis. Nature 359:235-237. [DOI] [PubMed] [Google Scholar]
• 33.Pozniak, C. D., S. Radinovic, A. Yang, F. McKeon, D. R. Kaplan, and F. D. Miller. 2000. An anti-apoptotic role for the p53 family member, p73, during developmental neuron death. Science 289:304-306. [DOI] [PubMed] [Google Scholar]
• 34.Roose, J., G. Huls, M. van Beest, P. Moerer, K. van der Horn, R. Goldschmeding, T. Logtenberg, and H. Clevers. 1999. Synergy between tumor suppressor APC and the beta-catenin-Tcf4 target Tcf1. Science 285:1923-1926. [DOI] [PubMed] [Google Scholar]
• 35.Sierra, J., T. Yoshida, C. A. Joazeiro, and K. A. Jones. 2006. The APC tumor suppressor counteracts beta-catenin activation and H3K4 methylation at Wnt target genes. Genes Dev. 20:586-600. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 36.Smale, S. T., A. Jain, J. Kaufmann, K. H. Emami, K. Lo, and I. P. Garraway. 1998. The initiator element: a paradigm for core promoter heterogeneity within metazoan protein-coding genes. Cold Spring Harbor Symp. Quant. Biol. 63:21-31. [DOI] [PubMed] [Google Scholar]
• 37.Smith, S. T., S. Petruk, Y. Sedkov, E. Cho, S. Tillib, E. Canaani, and A. Mazo. 2004. Modulation of heat shock gene expression by the TAC1 chromatin-modifying complex. Nat. Cell Biol. 6:162-167. [DOI] [PubMed] [Google Scholar]
• 38.Takemaru, K. I., and R. T. Moon. 2000. The transcriptional coactivator CBP interacts with beta-catenin to activate gene expression. J. Cell Biol. 149:249-254. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 39.Tontonoz, P., J. B. Kim, R. A. Graves, and B. M. Spiegelman. 1993. ADD1: a novel helix-loop-helix transcription factor associated with adipocyte determination and differentiation. Mol. Cell Biol. 13:4753-4759. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 40.Tutter, A. V., C. J. Fryer, and K. A. Jones. 2001. Chromatin-specific regulation of LEF-1-beta-catenin transcription activation and inhibition in vitro. Genes Dev. 15:3342-3354. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 41.Vadlamudi, U., H. M. Espinoza, M. Ganga, D. M. Martin, X. Liu, J. F. Engelhardt, and B. A. Amendt. 2005. PITX2, beta-catenin and LEF-1 interact to synergistically regulate the LEF-1 promoter. J. Cell Sci. 118:1129-1137. [DOI] [PubMed] [Google Scholar]
• 42.van de Wetering, M., J. Castrop, V. Korinek, and H. Clevers. 1996. Extensive alternative splicing and dual promoter usage generate Tcf-1 protein isoforms with differential transcription control properties. Mol. Cell. Biol. 16:745-752. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 43.van de Wetering, M., E. Sancho, C. Verweij, W. de Lau, I. Oving, A. Hurlstone, K. van der Horn, E. Batlle, D. Coudreuse, A. P. Haramis, M. Tjon-Pon-Fong, P. Moerer, M. van den Born, G. Soete, S. Pals, M. Eilers, R. Medema, and H. Clevers. 2002. The beta-catenin/TCF-4 complex imposes a crypt progenitor phenotype on colorectal cancer cells. Cell 111:241-250. [DOI] [PubMed] [Google Scholar]
• 44.van Genderen, C., R. Okamura, I. Farinas, R. Quo, T. Parslow, L. Bruhn, and R. Grosschedl. 1994. Development of several organs that require inductive epithelial-mesenchymal interactions is impaired in LEF-1-deficient mice. Genes Dev. 8:2691-2703. [DOI] [PubMed] [Google Scholar]
• 45.Wittschieben, B. O., G. Otero, T. de Bizemont, J. Fellows, H. Erdjument-Bromage, R. Ohba, Y. Li, C. D. Allis, P. Tempst, and J. Q. Svejstrup. 1999. A novel histone acetyltransferase is an integral subunit of elongating RNA polymerase II holoenzyme. Mol. Cell 4:123-128. [DOI] [PubMed] [Google Scholar]
• 46.Wong, M. H., J. Huelsken, W. Birchmeier, and J. I. Gordon. 2002. Selection of multipotent stem cells during morphogenesis of small intestinal crypts of Lieberkuhn is perturbed by stimulation of Lef-1/beta-catenin signaling. J. Biol. Chem. 277:15843-15850. [DOI] [PubMed] [Google Scholar]
• 47.Xiao, Q., G. Claassen, J. Shi, S. Adachi, J. Sedivy, and S. R. Hann. 1998. Transactivation-defective c-MycS retains the ability to regulate proliferation and apoptosis. Genes Dev. 12:3803-3808. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 48.Yamanaka, R., G. D. Kim, H. S. Radomska, J. Lekstrom-Himes, L. T. Smith, P. Antonson, D. G. Tenen, and K. G. Xanthopoulos. 1997. CCAAT/enhancer binding protein epsilon is preferentially up-regulated during granulocytic differentiation and its functional versatility is determined by alternative use of promoters and differential splicing. Proc. Natl. Acad. Sci. USA 94:6462-6467. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 49.Yang, A., M. Kaghad, Y. Wang, E. Gillett, M. D. Fleming, V. Dotsch, N. C. Andrews, D. Caput, and F. McKeon. 1998. p63, a p53 homolog at 3q27-29, encodes multiple products with transactivating, death-inducing, and dominant-negative activities. Mol. Cell 2:305-316. [DOI] [PubMed] [Google Scholar]
• 50.Yokoyama, C., X. Wang, M. R. Briggs, A. Admon, J. Wu, X. Hua, J. L. Goldstein, and M. S. Brown. 1993. SREBP-1, a basic-helix-loop-helix-leucine zipper protein that controls transcription of the low density lipoprotein receptor gene. Cell 75:187-197. [PubMed] [Google Scholar]
• 51.Zaika, A. I., N. Slade, S. H. Erster, C. Sansome, T. W. Joseph, M. Pearl, E. Chalas, and U. M. Moll. 2002. ΔNp73, a dominant-negative inhibitor of wild-type p53 and TAp73, is up-regulated in human tumors. J. Exp. Med. 196:765-780. [DOI] [PMC free article] [PubMed] [Google Scholar]