The intestinal alkaline phosphatase gene (ALPI) encodes a digestive brush-border enzyme, which is highly upregulated during small intestinal epithelial cell differentiation. To identify new putative promoter motifs responsible for the regulation of ALPI expression during differentiation of the enterocytes, we have conducted a computer-assisted cis-element search of the proximal human ALPI promoter sequence. A putative recognition site for the transcription factor hepatocyte nuclear factor (HNF)-4 was predicted at the positions from −94 to −82 in relation to the translational start site. The ability of HNF-4α to stimulate the expression from the ALPI promoter was investigated in the nonintestinal Hela cell line. Cotransfection with an HNF-4α expression vector demonstrated a direct activation of the ALPI promoter through this −94 to −82 element. EMSA showed that HNF-4α from nuclear extracts of differentiated intestinal epithelial cells (Caco-2) bound with high affinity to the predicted HNF-4 binding site. A 521 bp promoter fragment containing the HNF-4 binding site demonstrated a differentiation-dependent increase in promoter activity in Caco-2 cells. The presence of the HNF-4 binding site was necessary for this increase to occur.
- intestinal alkaline phosphatase
- transcriptional regulation
the small intestinal epithelium is a dynamic structure that undergoes a highly regulated process of cell proliferation, migration, and differentiation (12, 21, 27). A multitude of small intestinal in situ hybridization experiments have been reported (for an overview see the crypt villus in situ hybridization database, see Ref. 18). Such studies have led to the realization that the differentiation process in the absorptive enterocytes in the adult small intestine is associated with the appearance of mRNAs for a large set of enterocyte-specific genes at the epithelial crypt villus junction.
Early studies of the genes encoding the intestinal digestive brush-border hydrolases aminopeptidase N (ANPEP), lactase-phlorizin-hydrolase (LCT), and sucrase-isomaltase (SI) have revealed that the presence of binding sites for the transcription factors hepatocyte nuclear factor 1 (HNF-1) (15, 16, 23, 33) and the caudal-related homeobox (CDX) member CDX-2 (24, 29) are common to the promoters of these genes that are expressed in a differentiation-specific manner along the crypt villus axis in the intestinal epithelium. The more general importance of the HNF-1 and CDX-2 transcription factors for gene expression in differentiated small intestinal epithelial cells have subsequently been stressed by the finding of binding sites for the two factors in the promoters of several other intestinally expressed genes such as calbindin-D9k (11), guanylin (7), and lysophospholipase (25).
HNF-4 belongs to the nuclear receptor superfamily and is expressed in several endodermally derived tissues including the liver, pancreas, and intestine. HNF-4 is known to play a pivotal role in the regulation of the expression of genes encoding proteins involved in hepatic lipid homeostasis (for a review see Ref. 31). In the intestine, HNF-4 is known to be important for the promoter activity of the genes encoding the apolipoproteins (8, 10).
Human intestinal alkaline phosphatase (ALPI) is an intestinal brush-border hydrolase transcriptionally upregulated during enterocyte differentiation. The molecular mechanisms involved in the transcriptional regulation of this gene have recently been addressed (5, 6, 9). Data from the initial characterization of the ALPI promoter have shown that the −179 through −49 segment of the proximal promoter is essential for specific expression during differentiation of the intestinal-like colon carcinoma HT-29 cell line. A cis-element (called IF-III) positioned at −163 to −135 upstream of the translational start site (in the following presentation we will adopt this numbering, which was introduced in Ref. 9) binds Sp1 and Sp3 (6, 9). Subsequently, the gut-enriched Krüppel-like factor 4 (KLF4) has also been demonstrated to bind to the IF-III element. KLF4 is preferentially expressed in differentiated intestinal epithelial cells and the binding of KLF4 to the ALPI promoter has therefore been hypothesized to be important for the differentiation-dependent ALPI expression in the enterocytes (5).
The present study was undertaken to investigate whether other transcription factors expressed in intestinal epithelial cells contribute to the differentiation-dependent expression of the ALPI gene in the small intestine. We report here the presence of a hitherto unidentified potential HNF-4 binding site positioned between two regions [intestinal alkaline phosphatase footprint (IF)-IV and IF-V] that are downstream of the IF-III region and that have also previously been shown to interact with nuclear proteins from intestinal cells (9). Moreover, our analyses demonstrate that this HNF-4 binding site at position −94 to −82 in the proximal human ALPI promoter is indeed functional. Using EMSA in combination with a specific HNF-4α antibody, we found that HNF-4α have a high affinity for this motif, and in addition, we show that HNF-4α, via this site, stimulates transcription from a 521 bp ALPI promoter construct in differentiated but not in undifferentiated Caco-2 cells. Moreover, the expression is differentiation-dependent because it increases during differentiation and the increase is absolutely dependent on the presence of the HNF-4 binding site.
MATERIALS AND METHODS
Genomic human DNA (0.5 μg) was used as a template in a PCR using Taq DNA polymerase (Fermentas, Vilnius, Lithuania) to amplify a 521-bp fragment (from positions −566 to −45 using the numbering suggested in Ref. 9) of the ALPI gene. In the 5′ primer (5′-CAGCTAGCCATCTACCTGTGCAAGGGAA-3′) an NheI site (underlined) was introduced, whereas in the 3′ primer (5′-TTAGATCTGAAGTGGGGACACCAGGAACC-3′) a BglII site was added. The PCR fragment was cloned in front of the luciferase reporter gene in the NheI and BglII sites of pGL3-Basic vector (Promega, Madison, WI) to generate the reporter plasmid pALPI_566. In addition, two 5′ promoter deletion constructs with 5′ ends corresponding to positions −143 and −83 in the ALPI gene, respectively, were produced by PCR using the pALPI_566 construct as template. The 5′ primers used for these constructs were 5′-CAGCTAGCAAGATGGACACCAGGGGTGT-3′ and 5′-CAGCTAGCTCCCCTGATTTAAACCCAGG-3′, respectively.
Site-directed mutagenesis of the HNF-4 site in the ALPI promoter was also performed using a PCR strategy. Overlapping PCR fragments carrying the mutation were annealed, and the whole fragment was amplified by using flanking primers. One PCR reaction with the primers 5′-CTAGCAAAATAGGCTGTCCC-3′ (corresponding to nucleotides 4760–4779 of the pGL3-Basic vector), 5′-GTTTAAATCAGGGGAGGACTCGAGCCCTGGCTGTAAATGCTC-3′ (corresponding to nucleotides −68 to −109 of the ALPI gene; underlined nucleotides denote the mutation), and 1 ng of the plasmid pALPI_566 as template generated a 563-bp fragment of the ALPI promoter. A second PCR reaction with the primers 5′-GAGCATTTACAGCCAGGGCTCGAGTCCTCCCCTGATTTAAAC-3′ (corresponding to nucleotides −109 to −68 of the ALPI gene; underlined nucleotides denotes the mutation), 5′-GCCTTATGCAGTTGCTCTCC-3′ (corresponding to nucleotides 101–120 of the pGL3-Basic vector), and 1 ng of the plasmid pALPI_566 as template generated a 186-bp fragment of the ALPI promoter/luciferase fusion in pALPI_566. Each of the generated fragments (1 ng) were combined and used as templates in a third PCR reaction using the flanking primers from the pGL3-Basic vector generating a 706-bp fragment corresponding to the ALPI promoter with a mutation in the HNF-4 site fused to the 5′ end of the luciferase gene. This fragment was digested with the restriction enzymes NheI and BglII, and the released ALPI promoter fragment cloned into the pGL3-Basic vector digested with the same enzymes. All constructs were sequenced using the BigDye Terminator v3.1 Cycle Sequencing Kit and the Avant 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA) to verify the sequences of the cloned DNA fragments.
Cell culture and transfections.
Caco-2 (human colon carcinoma cell line) and Hela (human cervix epithelial carcinoma cell line) cells were cultured in minimal essential medium (Invitrogen Life Technologies, Carlsbad, CA) containing 10% FCS. For transient transfections, Caco-2 and Hela cells were seeded in 24-well plates 1 day before transfection (3 × 104 cells/well). The Exgen500 (Fermentas UAB) transfection reagent was used according to the manufactures instructions. In brief, a total of 1 μg DNA and 3.3 μl Exgen500 were diluted in 150 mM NaCl to a final volume of 50 μl. For each transfection, 250 ng luciferase reporter plasmid and 125 ng of β-galactosidase expression plasmid (pCMVlacZ; Promega) were used. In some transfections 250 ng of a HNF-4α expression plasmid (pRHNF-4 kindly provided by Dr. F. M. Sladeck, University of California Riverside, CA) were added. The total amount of DNA was adjusted to 1 μg by adding pBluescript SK+ (Stratagene, La Jolla, CA). Medium was changed 3 h after transfection and the cells were harvested after 48 h (undifferentiated) or 10 days after transfection (differentiated) as described previously (28, 30). Protein extracts from the harvested cells were analyzed for luciferase and β-galactosidase activities using the Tropix Dual Light system (Applied Biosystems). The luciferase activity was corrected for transfection efficiency by β-galactosidase expression measurements and normalized to the expression of pALPI_566. Four to six replicates of each experiment were performed.
Generation of pools of stably transfected Caco-2 cells carrying either the pALPI_566 or the pALPI_566ΔHNF-4 promoter constructs was done essentially as previously described (14). The ALPI promoter/reporter construct (20.4 μg), the CMVLACZ construct (1.8 μg), and the neomycin resistance expression plasmid (1.8 μg) RSVNEO (3) was mixed with 120 μl Exgen500 according the manufacturer's protocol and added to exponentially growing Caco-2 cells. Forty-eight hours after the addition of the DNA, the cells were trypsinized and divided among three 75-cm2 culture flasks in minimal essential medium supplemented with 10% FCS, and 1 mg/ml G418. After 3 wk of selection, ∼30 clones were visible per flask. Clones from each flask were released by trypsination, pooled, and passaged once before they were seeded at a density of 2 × 104 cells/cm2. Cells were harvested either 48 h after seeding (undifferentiated) or 2 wk after seeding (differentiated). Protein extracts and measurements of luciferase and β-galactosidase activities were performed as described above. The APLI promoter-driven luciferase activity was normalized to the measured β-galactosidase activity. Three different pools of cells were analyzed for each of the two ALPI promoter constructs.
Nuclear extracts and EMSA.
Nuclear extracts were prepared from (1) exponentially growing (2 days after seeding) and from 14 days postconfluent, Caco-2 cultures followed by concentration using ammonium sulfate precipitation as described previously (17). DNA mobility shift assays were performed with the following double-stranded oligonucleotides HNF-4 (5′-CCAGGGGCAAAGTCCTCCCCTGA-3′) and SP-1 (5′-AGGGGAGGACTTTGCCCCTGGCT-3′). Two complementary oligonucleotides (1,000 pmol each) were annealed, and 5 pmol of the annealed oligonucleotides were 5′ end-labeled with [γ32P]ATP using T4 polynucleotide kinase. Labeled products were purified on a Sephadex G-25 column (Amersham-Bioscience, Uppsala, Sweden). Nuclear extracts (3 μg for the experiments shown in Figs. 1 and 2 and 2 μg for the experiment shown in Fig. 4) obtained from Caco-2 cells were incubated with 1,000 ng of the nonspecific competitor poly(dI-dC) for 10 min on ice in 1× binding buffer (in mM: 25 Tris·HCl, pH 7.5, 5 MgCl, 60 KCl, 0.5 EDTA, 1 DTT, and 0.5 PMSF plus 5% Ficoll 400 and 2.5% glycerol). We then added 10 fmol of 32P-labeled probes, and incubation was prolonged for 20 min on ice (in a final volume of 12 μl). In some assays, unlabeled double-stranded oligonucleotides were added in 100-fold molar excess along with the labeled probe. Where indicated, 2 μg of specific HNF-4α antibody (SC-8987; Santa Cruz Biotechnology, Santa Cruz, CA) was added to the incubation mixture, and the incubation was prolonged for 10 min at 37°C before addition of the probe. The samples were then loaded onto a 5% polyacrylamide gel and electrophoresed at 120 V for 1.5 h in 0.5 × 45 mM Tris-borate, pH 8.3, 1 mM EDTA (TBE) buffer. The gel was dried and used to expose a PhosphorImager screen, which was scanned using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).
Chromatin immunoprecipitation (ChIP) was performed essentially as described in Ref. 2. In brief, 48-h cultures and 14 days postconfluent Caco-2 cell cultures were fixed by addition of 37% formaldehyde to a final concentration of 1% followed by 10 min of incubation at room temperature. Cells were subsequently harvested by scraping in SDS buffer [50 mM Tris·HCl, pH 8.1, 1% SDS, 100 mM NaCl, 5 mM EDTA, and protease inhibitor cocktail (Sigma-Aldrich, St. Louis, MO) in a 100-fold dilution], and counted. Cells were sonicated to achieve an average genomic DNA size of 500–1,000 bp. The sonicated mixture was stored frozen (−80°C) in 300-μl aliquots. Each sample was diluted in 1,200 μl ChIP dilution buffer [1% Triton X-100, 150 mM NaCl, 2 mM EDTA, 20 mM Tris·HCl, pH 8.1, and protease inhibitor cocktail (Sigma-Aldrich) and mixed with 25 μl protein A beads (50% slurry Protein A-Sepharose; Amersham-Bioscience, Uppsala, Sweden) in Tris-EDTA buffer containing 0.2 mg/ml sonicated calf sperm DNA, and 0.5 mg/ml lipid-free BSA] followed by centrifugation at 20,000 g for 30 min. The supernatant was immunoprecipitated with 2 μg HNF-4α antibody (SC-8987; Santa Cruz Biotechnology) at 4°C with continuous mixing overnight. Debris was pelleted (20,000 g for 20 min), and the supernatant was transferred to a new tube containing 25 μl protein A beads [50% slurry protein A-Sepharose (Amersham-Bioscience) in Tris-EDTA buffer containing 0.2 mg/ml sonicated calf sperm DNA and 0.5 mg/ml lipid-free BSA] and allowed to incubate for 2 h at 4°C with continuous mixing. The precipitate was harvested, and the pelleted protein A-Sepharose was resuspended, incubated on ice for 4 min, and washed once in buffer 1 (0.1% SDS, 1% Triton X-100, 150 mM NaCl, 2 mM EDTA, 20 mM Tris·HCl, pH 8.1, protease inhibitors), twice in buffer 2 (0.1% SDS, 1% Triton X-100, 500 mM NaCl, 2 mM EDTA, 20 mM Tris·HCl, pH 8.1), once in buffer 3 (0.25 LiCl, 1% deoxycholate, 1% Nonidet P-40, 0.5 mM EGTA, 10 mM Tris·HCl, pH 8.1, protease inhibitors) and two times in Tris-EDTA buffer. Finally, the DNA was eluted by resuspending the pellet in 300 μl 1% SDS and 0.1 M NaHCO3 followed by an incubation at 65°C for 30 min and a brief centrifugation. To the supernatant, 12 μl 5M NaCl was added and decross-linking continued overnight at 65°C followed by the addition of 250 μl proteinase K and incubation at 55°C for 2 h. Subsequently, 55 μl of 4 M LiCl was added. The DNA was then purified by phenol extraction and ethanol precipitation.
Quantitative PCR was performed on 5 μl ChIP DNA with the LightCycler FastStart DNA MasterPLUS SYBR Green I (Roche, Basel, Switzerland) system in accordance with the manufacturer's instructions. Ten picomoles of each primer for human ALPI (5′-TTCAGCAAGCTTGGCTTCAGGT-3′ and 5′CAGGCTGCCTGGGTTTAAATCA-3′) and human ANPEP (5′-GCCAAGGCTACAGATGAAGG-3′ and 5′-TCCCTGACCTCTCCATGAAC-3′) were used to generate an ALPI promoter fragment of 127 bp (including the region of potential HNF-4 binding) and a 147 bp product of the ANPEP intron 1. Before the actual quantifications of ChIP DNA were performed, a control procedure was carried out for the specificity of the amplification with the LightCycler instrument. The PCR products obtained by the LightCycler PCR amplification protocol were spun out of the capillary tubes and analyzed on a 3% agarose gel to verify their size. The sequences of the products were also determined by direct sequencing of the PCR products and their identity with the expected products confirmed. A melting point analysis is routinely carried out at the end of the LightCycler PCR amplification protocol and the melting curve profile of the analyzed PCR products for the ALPI promoter and the ANPEP intron 1 was recorded and stored. For the actual quantifications of ChIP DNA, a total of 40 cycles were performed on the LightCycler (Roche, Basel, Switzerland), and resulting amplification curves were analyzed using the second derivative method setting in the LightCycler software (Roche) to determine the crossing point. The number of template copies was determined using serial dilutions of the purified PCR products as templates in separate PCR reactions. In all ChIP real-time PCR quantifications the melting curve analysis showed only the expected peak.
The MatInspector professional program 5.1 (20) was used in a computer-assisted search for the presence of possible transcription factor binding sites within the sequence of the 521-bp ALPI promoter segment cloned (ALPI_566).
HNF-4 binds to a cis-element in the human ALPI promoter.
To identify potential motifs that could confer enterocyte-specific regulation of the human ALPI promoter, we performed a computer-assisted analysis of the proximal upstream region of this gene. No motifs for HNF-1 or CDX-2 were predicted within this region; however, a putative binding site for HNF-4 was identified at positions from −94 to −82, which is between the previously identified IF-IV and IF-V regions (Fig. 1) that interact with proteins from nuclei of HT-29 colon carcinoma cells (9). The ability of HNF-4 to regulate and bind to the ALPI proximal promoter was therefore investigated. The colon carcinoma cell line Caco-2 was chosen as a cell model for small intestinal epithelial cells in the subsequent analysis because preconfluent Caco-2 cells resemble undifferentiated crypt cells and postconfluent Caco-2 cells after 1–2 wk in culture display-differentiated enterocyte characteristics that also include the expression alkaline phosphatase activity (19).
EMSA and super-shift assays were carried out using nuclear extracts from 14-days postconfluent Caco-2 cells, an HNF-4α specific antibody and a double-stranded oligonucleotide probe corresponding to positions −97 to −75 in the ALPI promoter. Proteins present in the Caco-2 nuclear extracts bind to the radiolabeled probe, as revealed by the presence of one intensive band (Fig. 1, lane 1). The band disappears after the addition of the same unlabeled oligonucleotide in 100-fold molar excess (Fig. 1, lane 2). On contrary, binding of nuclear extracts to the probe is not affected by the addition of an unlabeled unrelated double-stranded oligonucleotide corresponding to the Sp1 motif described in the aminopeptidase N promoter (15) (Fig. 1, lane 3). This suggests that the band represents a sequence-specific protein DNA interaction. The addition of an HNF-4α-specific antibody results in a further decrease in the mobility of the protein-DNA complex. Thus the HNF-4α antibody produces a complete super shift of the radioactive probe with the Caco-2 nuclear extract (Fig. 1, lane 4). This indicates that HNF-4α is the dominating protein, which binds to the double-stranded ALPI HNF-4 oligonucleotide probe in nuclear extracts from differentiated Caco-2 cells.
The effect of HNF-4 on ALPI promoter activity was studied in the nonintestinal Hela cell line that does not express ALPI or HNF-4 endogenously. Transient cotransfection assays were carried out with four reporter constructs including three ALPI promoter constructs with a different length of 5′ flanking sequences (with 5′ ends at positions −566, −143, and −83 in the ALPI upstream region, respectively) and a mutated version of the longest ALPI promoter (ending at position −566) construct carrying a mutation in the predicted HNF-4 binding site. All ALPI promoter constructs were placed in front of the firefly luciferase gene. As shown in Fig. 2, cotransfection with the HNF-4 expression plasmid increased significantly the transcription from the −521 and −143 ALPI promoter deletion constructs that both included the predicted HNF-4 binding site in the ALPI promoter. In contrast, the activity of both the −83 ALPI promoter deletion construct and the −521 promoter construct with a mutation in the predicted HNF-4 binding site were not stimulated when cotransfected with the HNF-4 expression construct. Moreover, as seen also in Fig. 2, a double-stranded oligonucleotide with a sequence corresponding to the mutated ALPI HNF-4 binding site used in the promoter construct pALPI_566ΔHNF-4 cannot compete out the binding of nuclear proteins from differentiated Caco-2 cells to the ALPI HNF-4 probe. These results show that HNF-4 can activate transcription of the ALPI promoter and that this activation is dependent on the presence of the predicted HNF-4 binding site.
HNF-4 stimulates ALPI promoter activity in differentiated Caco-2 cells.
To clarify the role of HNF-4 on the regulation of the differentiation-dependent ALPI gene expression in an intestinal cell line, transfection experiments were performed in Caco-2 cells with the wild-type (pALPI_566) as well as with the pALPI_566ΔHNF-4 ALPI promoter construct with the HNF-4 binding site mutated. Two different transfection strategies were used. In a first series of transfections a recently developed transient transfection protocol (28, 30) for the analysis of differentiation-dependent promoter activity in Caco-2 cells was used. Transfected cells were harvested at their undifferentiated state (48 h after transfection) or in their differentiated state (10 days after transfection) and assayed for reporter gene luciferase activity. As seen in Fig. 3A the wild-type 521 bp ALPI promoter construct is able to mediate a differentiation-dependent twofold increase in promoter activity in Caco-2 cells. In contrast, the activity of the ALPI promoter construct carrying the HNF-4 binding site mutation is decreased during differentiation of Caco-2 cells. The overall differentiation-dependent increase in ALPI promoter activity in Caco-2 cells was, however, relatively low. In transient transfections, the reporter construct remains as episomal DNA in the cells. One explanation might therefore be that integration of the reporter construct into the genome of Caco-2 cells and subsequent organization into chromatin is required to yield a high effect of differentiation on the promoter activity. A second series of transfections were therefore carried out in combination with an expression construct yielding resistance to the G418 selective agent. Pools of the derived stably transfected cellular clones were subsequently analyzed for differentiation-dependent activity from the pALPI_566 wild-type construct and the pALPI_56ΔHNF-4 mutated construct. As seen in Fig. 3B the promoter activity of the wild-type ALPI promoter construct was strongly increased (almost 300-fold) in differentiated Caco-2 cells compared with the undifferentiated Caco-2 cells. The ALPI promoter construct with the mutated HNF-4 site demonstrated only, however, a small twofold increase in promoter activity in the differentiated Caco-2 cells. These results support a role of the HNF-4 site in the ALPI promoter for its differentiation-dependent activity in intestinal epithelial cells.
ALPI HNF-4 binding site is occupied in both undifferentiated and differentiated Caco-2 cells.
The change in the binding of nuclear proteins to the predicted ALPI HNF-4 binding site during differentiation of the Caco-2 cells was investigated by EMSA analysis. As seen in Fig. 4, there was no significant change in the binding to the ALPI HNF-4 site during differentiation of the Caco-2 cells. To investigate further the dynamics of HNF-4 association with the endogenous ALPI promoter in Caco-2 cells, we performed ChIP with an antibody specific to HNF-4α. The ChIP procedure involves the cross-linking of nuclear proteins to the DNA in vivo. The cross-linked chromatin is subsequently fragmented and immunoprecipitated using an antibody specific to the transcription factor under investigation. A key point in the procedure is the quantification of an enrichment of specific DNA regions carrying binding sites for the factor under investigation. The enrichment is most efficiently quantified by a quantitative PCR procedure. In this regard, it is important that only the specific DNA fragment is amplified and subsequently quantified. For the present study, two sets of primers were designed: one set of primers that can be used in PCR to amplify the ALPI promoter region including the predicted HNF-4 binding site and another set of primers that the can be used for PCR amplification of a region in intron 1 of the aminopeptidase N gene (ANPEP). This intronic ANPEP region serves as a control and does not contain a predicted HNF-4 binding site. With the input DNA used for ChIP, the primers specifically amplified the predicted regions (Fig. 5) using the LightCycler real-time PCR instrument. In addition, a melting curve analysis (Fig. 5) showed that each fragment has a well-defined melting point. Melting point analysis was subsequently used to verify that the quantitative measurements were indeed performed on the correct and specific PCR products.
As seen in Fig. 6, quantitative PCR on ChIP DNA from both differentiated and undifferentiated Caco-2 cells shows that the ALPI promoter fragments are specifically enriched by immunoprecipitation with the HNF-4 antibody compared with the ANPEP intron 1 region not suspected to bind HNF-4. This suggests that HNF-4α is associated with the ALPI promoter in both undifferentiated and differentiated Caco-2 cells. Taken together with the EMSA analysis in Fig. 4, results suggest that little or no change occurs in the binding of HNF-4 to the binding site in ALPI promoter during differentiation of the Caco-2 cells.
Role of HNF-4 in transcriptional activation of intestinal ALPI gene expression.
In the present study, we have extended the investigations of the human ALPI proximal promoter and identified a previously unreported HNF-4 cis-element at the positions from −94 to −82. We discovered the HNF-4 binding site during a computer search with the MatInspector program (20) of the ALPI promoter for binding sites defined by position-specific weight matrices available in the Transfac database (32). The matrix with the ID: V$HNF4_01; gave a very high score (0.909) for the −94 to −82 ALPI promoter region. The consensus HNF-4 binding sequence defined by the V$HNF4_01 matrix is 5′-RGGNCAAAGKTCA-3′ (R = purine; K = G, T; N = G, A, C, T). This ALPI HNF-4 binding site is located immediately between the IF-IV and IF-V regions previously reported to interact with nuclear proteins from HT-29 cells. In our analysis, we have used the Caco-2 cell line, and this might explain why the HNF-4 region was not reported to be protected in the DNase I foot printing carried out by Kim et al. (9). Recently, the IF-III region (from positions −163 to −135) was reported to interact with the KLF4 transcription factor, and it was speculated that this interaction might be important for the increase in ALPI promoter activity during enterocytes differentiation (5). In the present work, we show that the region from positions −566 to −45 of the ALPI promoter displays a differentiation-dependent increase in promoter activity in Caco-2 cells. This increase is abolished by mutation of the HNF-4 site. The IF-III region with the KLF4 site is therefore not sufficient to secure a differentiation-dependent increase in promoter activity in Caco-2 cells, although the site is likely to contribute to the overall activity of the promoter. Data from the ChIP experiment showed binding of HNF-4 at the ALPI promoter in both undifferentiated and differentiated Caco-2 cells in vivo, which suggests that HNF-4 associates with the ALPI promoter before differentiation and before the increase in ALPI transcription that occurs at differentiation. This apparent paradox might be explained by the fact that HNF-4 can associate with both the transcriptional coactivators p160 and p300 as well as with the transcriptional corepressor NCOR2 (also known as SMRT) (26).
Similar findings of occupancy of an HNF-4 binding site before promoter activation have been found for the association of HNF-4 with the α1-antitrypsin (α1-AT) promoter during differentiation of Caco-2 cells (22). In the case of the α1-AT promoter, the HNF-4 binding was increased during differentiation, but the binding of HNF-4 to the α1-AT promoter was maximal 2 days before the promoter was switched on. The eventual trigger of α1-AT transcription was thought to be chromatin remodeling (22), and the acetylation of histone 3 in nucelosomes on the α1-AT promoter was increased during Caco-2 differentiation before transcriptional initiation (22). For the ALPI promoter, an increase in histone H3 acetylation in nucleosomes surrounding the ALPI has similarly been demonstrated (6). Moreover, in the present study, it is demonstrated that the ALPI promoter responds much stronger to the differentiation of the Caco-2 cells when it is integrated into the genome compared with its differentiation-dependent response when it is present as episomal DNA. The strongly increased activity observed for the ALPI promoter in response to Caco-2 differentiation did, however, also depend on the presence of the HNF-4 binding site when the construct was integrated into the host cell genome. All of these findings might suggest that HNF-4 recruit additional chromatin modifying cofactors to the ALPI promoter as well as to other intestinal HNF-4 binding promoters during differentiation. The possibility that repressors are recruited in the undifferentiated state and activators in the differentiated state is appealing although not yet proven.
HNF-4 as a regulator of an intestinal lipid absorption program?
As stated in the introduction, the studies on especially the SI and LCT genes encoding the intestinal disaccharidases sucrase-isomaltase and lactase-phlorizin hydrolase suggest that the combination of CDX-2 and HNF-1 binding sites are typical to one subset of intestinal specific promoters.
In the liver, HNF-4 is essential for hepatic lipid homeostasis. Liver-specific inactivation of the HNF-4 gene results in defect lipoprotein synthesis, hepatic lipid accumulation and reduced serum cholesterol and triglyceride levels (4). This knowledge therefore might suggest that genes encoding proteins involved in intestinal lipid absorption might constitute another subset of HNF-4 responsive genes. It has recently been established that inactivation of the ALPI expression in mice, leads to increased fat absorption during a high-fat diet (13), thus directly proving a role for the ALPI gene in intestinal lipid absorption. The ALPI gene therefore seems to be a valid member of the proposed subset of small intestinal HNF-4 target genes encoding proteins involved in lipid absorption.
This work was supported by grants from The Danish Medical Research Council, The Novo Nordic Foundation, The Lundbeck Foundation, and the Alfred Nielsen and Wife's foundation.
We thank L.-L. Laustsen and J. Møller for excellent technical assistance and Dr. F. Sladek for the kind gift of the HNF-4α expression vector.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society