| Nucleic Acids Research | Pages |
Characterization of the structure and regulation of the murine gene encoding gut-enriched Krüppel-like factor (Krüppel-like factor 4)
Introduction
Materials And Methods
Genomic library screening
Plasmid constructs
S1 nuclease analysis
Preparation of nuclear extracts and recombinant GKLF protein
DNase I footprint assays
Electrophoretic mobility shift assays (EMSAs)
Site-directed mutagenesis
Transfection and reporter assays
Results
Structure of the murine GKLF gene
GKLF binds to the immediate 5[prime]-flanking region of its own gene
The zinc fingers of GKLF bind to each of the first two GC-boxes as a monomer
The 5[prime]-flanking region of the GKLF gene functions as a promoter and is autoregulated by GKLF
Mutations in one or both of the first two GC-boxes diminish the ability of GKLF to transactivate its own promoter
Discussion
Acknowledgements
References
Characterization of the structure and regulation of the murine gene encoding gut-enriched Krüppel-like factor (Krüppel-like factor 4)
Received September 1, 1999; Revised and Accepted October 8, 1999
DDBJ/EMBL/GenBank accession no. AF117109
ABSTRACT Gut-enriched Krüppel-like factor (GKLF, KLF4) is an epithelial-specific transcription factor whose expression is associated with growth arrest. In order to understand the mechanisms regulating expression of the gene encoding GKLF, we isolated a genomic clone containing murine GKLF. The gene spans 5.3 kb and contains four exons. A major start site of transcription was mapped to an adenine residue 601 nt 5[prime] of the translation initiation codon. An additional 1 kb of the 5[prime]-flanking region was sequenced and found to contain multiple cis-elements homologous to the binding sites of several established transcription factors including Sp1, AP-1, Cdx, GATA, and USF. In particular, three closely spaced GC-boxes 5[prime] of the TATA box resemble the established binding site for GKLF. DNase I protection and electrophoretic mobility shift assays verified that recombinant GKLF bound to each of the three GC-boxes. In co-transfection experiments, GKLF transactivated a reporter gene linked to the GKLF 1 kb 5[prime]-flanking region, as did Sp1, Sp3 and Cdx-2. Mutations of one or both of the first and second GC-boxes in the promoter resulted in diminished transactivation by GKLF. These results demonstrate that the 5[prime]-flanking sequence of the mouse GKLF gene functions as a promoter and is subject to autoregulation by its own gene product.
INTRODUCTION
Gut enriched Krüppel-like factor (GKLF, also called Krüppel-like factor 4 or KLF4; 1) is a recently identified transcription factor that contains three C2H2 zinc fingers (2,3). It belongs to a family of closely related Krüppel-like proteins including lung Krüppel-like factor (LKLF) and erythroid Krüppel-like factor (EKLF) (4-6). In vivo expression of the gene encoding GKLF is specific for epithelial tissues including those lining the gastrointestinal tract (2,3,7), epidermis (3,8) and thymus (9). Expression of GKLF has also been detected in vascular endothelial cells (10). Recent studies of mice with targeted ablation of GKLF indicate that it is required to establish the barrier function of the skin (8). In contrast, the in vivo function of GKLF in the gastrointestinal tract has not been clearly defined. Studies did suggest that GKLF has an important role in regulating genes whose expression is found in differentiated epithelial cells. Examples of identified `target' genes in the gastrointestinal tract include those encoding cytochrome P-450IA1 (CYP1A1) (11) and keratin 4 (7). The mechanisms by which GKLF regulates expression of these two genes, however, are distinctly different. While it is an activator of the keratin 4 gene (7), GKLF suppresses CYP1A1 (11). Thus, GKLF exhibits a pleiotropic regulatory effect in a manner similar to several other zinc finger-containing transcription factors such as YY1 (12,13) and Sp3 (14,15).
Both the in vivo (2,3,8,9) and in vitro (2) patterns of expression of GKLF are indicative of a growth arrest-associated nature. In the intestine, GKLF is preferentially expressed in the post-mitotic population of epithelial cells (2). Similarly, GKLF transcripts are enriched in the mitotically inactive suprabasal layer of the epidermis (3,8) and in quiescent cortical thymus epithelium (9). In cultured cells, levels of GKLF mRNA are increased when cells reach quiescence due to serum deprivation or contact inhibition (2). Conversely, GKLF mRNA levels are decreased during stimulation of serum-starved cells by the addition of fresh serum (2). Furthermore, enforced expression of GKLF results in inhibition of DNA synthesis (2). These results not only suggest that GKLF is potentially a negative regulator of proliferation, but also underscore the fact that expression of GKLF is likely to be stringently controlled at different stages of cellular proliferation.
In an attempt to better understand the molecular mechanisms regulating its expression, we identified and isolated a genomic clone containing murine GKLF. The current study depicts the structure of the gene including its flanking regions. We also examined regulation of the GKLF promoter and identified a number of transactivating factors that up-regulate the promoter activity. One of these factors is GKLF itself. This indicates that expression of GKLF is subject to an autoregulatory mechanism and establishes that GKLF is yet another newly identified target gene of GKLF.
MATERIALS AND METHODS
Genomic library screening
A genomic library from the 129SvJ mouse strain in [lambda] FIX II (Stratagene, La Jolla, CA) was screened at high stringency with a full-length cDNA fragment encoding GKLF (2). Two positive plaques were identified among 106 independent plaques. Southern and restriction analyses indicated that one of the two phages contained the entire GKLF gene within a 15 kb insert. Approximately 5 kb of the transcribed region and 1 kb of the 5[prime]-flanking region were sequenced and deposited in GenBank under accession no. AF117109.
Plasmid constructs
The eukaryotic expression constructs containing full-length and truncated GKLF [PMT3-GKLF(1-483) and PMT3-GKLF(1-401), respectively] were previously described (6). Expression constructs containing Sp1 and Sp3 (CMV-Sp1 and CMV-Sp3, respectively) were generously provided by G. Suske (16). CMV-Cdx2 was provided by P. Traber (17). Expression constructs containing Xenopus GATA-4, -5 and -6 were generous gifts from T. Evans (18). The promoterless luciferase reporter plasmid, pGL2-Basic vector, was purchased from Promega (Madison, WI). A SacI-KpnI fragment from the GKLF-containing [lambda] phage that spans 1 kb of the 5[prime]-flanking region and 550 bp of the 5[prime]-untranslated region was subcloned into pGL2-Basic vector to form the -1.0 kb GKLF-pGL2-Luciferase construct. The internal control for transfection, CMV-[beta]-galactosidase, was purchased from Life Technologies (Gaithersburg, MD).
S1 nuclease analysis
S1 nuclease analysis was performed according to a previously published protocol (19). A 312 nt NcoI fragment containing the putative start site of transcription of GKLF was labeled as a probe with [[gamma]-32P]ATP (3000 Ci/mmol; DuPont-New England Nuclear, Wilmington, DE) at the 3[prime]-terminus. Fifty micrograms of RNA isolated from various regions of mouse intestine or liver, or tRNA were annealed to 2.5 × 105 d.p.m. of the probe in 20 µl 80% formamide, 40 mM PIPES, pH 6.4, 1 mM EDTA and 0.4 M NaCl for 16 h at 30°C. A mixture of 0.56 M NaCl, 0.1 M NaOAc, pH 4.5, 9 mM ZnSO4, 6 µg heat-denatured salmon sperm DNA and 300 U S1 nuclease was then added. The reactions were continued for 1 h at 30°C, chilled on ice, and terminated by addition of 80 µl 4 M NH4OAC, pH 8.0, 20 mM EDTA and 40 µg/ml tRNA. The products were analyzed on a 6.0% polyacrylamide-7 M urea sequencing gel and visualized by autoradiography.
Preparation of nuclear extracts and recombinant GKLF protein
Nuclear extracts containing full-length GKLF were prepared from COS-1 cells transfected with PMT3-GKLF. Extracts from cells transfected with PMT3 alone were used as controls. Transfected cells were rinsed with ice-cold phosphate-buffered saline (PBS), scraped and pelleted at 400 g for 5 min at 4°C. The pellets were washed with 4 pack cell volumes (p.c.v.) of a solution containing 10 mM Tris-HCl, pH 7.8, 1.5 mM MgCl2, 10 mM KCl, and a cocktail of protease inhibitors including 1 µg/ml leupeptin, 1 µg/ml pepstatin, 1 µg/ml aprotinin and 20 µM phenylmethylsulfonyl fluoride. Following a 10 min incubation on ice, the cells were lysed by 10 strokes of a Dounce homogenizer. Nuclei were collected by centrifugation at 4500 g for 5 min at 4°C and resuspended in 2 p.c.v. of a lysis solution containing 420 mM KCl, 20 mM Tris-HCl, pH 7.8, 1.5 mM MgCl2, 0.5 mM DTT, 20% glycerol and the protease inhibitor cocktail. After incubation for 1 h at 4°C with gentle agitation, the solution was centrifuged at 10 000 g for 30 min at 4°C. The supernatant was dialyzed twice against 500 ml of 20 mM Tris-HCl, pH 7.8, 100 mM KCl, 0.2 mM EDTA, 0.5 mM DTT, 20% glycerol and protease inhibitors at 4°C for 4 h. Extracts were divided into small aliquots and stored at -70°C.
The preparation of recombinant protein containing the zinc finger portion of GKLF (amino acids 350-483) from pET-16b-GKLF-transformed bacteria was described previously (11,20). Briefly, 1 mM isopropyl-[beta]-D-thiogalactopyranoside was added to logarithmically growing bacteria for 4 h to induce GKLF production. Cell lysates were prepared by treating pelleted bacteria with lysis buffer (20 mM Tris-HCl, pH 7.9, 0.5 M NaCl, 6 M urea, 5 mM imidazole and the protease inhibitor cocktail) on ice for 30 min, followed by sonication and purification on a Ni-NTA-agarose column (Qiagen, Santa Clarita, CA) equilibrated with the lysis buffer. After extensive washing, bound proteins were eluted with the same buffer except that the concentration of imidazole was increased to 1 M. The eluted protein was serially dialyzed against a solution of 10 mM Tris-HCl, pH 7.4, 100 mM NaCl, 10 µM ZnCl2, 10% glycerol and gradually decreasing concentrations of urea from 6 M to nil.
DNase I footprint assays
The procedure for the DNase I protection (footprint) assay was previously described (21). The 312 nt NcoI fragment that encompasses the start site of transcription of GKLF was labeled on the coding strand with [[gamma]-32P]ATP and T4 polynucleotide kinase. The labeled fragment was purified on a native polyacrylamide gel and subsequently concentrated using a NACS column (Life Technologies). The binding reaction was performed with 2 ng (~100 000 d.p.m.) of purified probe and increasing amounts of recombinant GKLF protein in a buffer containing 20 mM HEPES, pH 7.4, 300 mM KCl, 5 mM MgCl2, 25 mM DTT and 1 µg poly(dI·dC) for 20 min at 4°C. The DNA was digested with a diluted (<0.05 µg/µl) solution of DNase I (Worthington Biochemical, Freehold, NJ) for 1 min at room temperature, and the reaction terminated with a stop buffer containing 1% SDS, 20 µM EDTA, 200 mM NaCl and 250 µg/ml yeast tRNA. After phenol/chloroform extraction and ethanol precipitation, the samples were resuspended in 99% formamide, denatured at 70°C for 5 min and resolved on a sequencing gel.
Electrophoretic mobility shift assays (EMSAs)
EMSAs were performed according to previously published protocols (11,20). Probes included the 312 nt NcoI restriction fragment in the 5[prime]-end of the GKLF gene and a double-stranded synthetic oligonucleotide containing the sequence between nt -134 and -87 of GKLF. Sequences of the wild-type and mutated oligonucleotides are as follows with the mutated nucleotides in bold, GC-box 1 in italic and GC-box 2 underlined: wild-type, 5[prime]-GCTGCGGGAAGGCGGGGAGAAGAAAGG-CAGGGGGCGGGGCCTGGCGGC-3[prime]; mutant 1, 5[prime]-GCTG-CGGGAATTCCGGGAGAAGAAAGGCAGGGGGCGGGG-CCTGGCGGC-3[prime]; mutant 2, 5[prime]-GCTGCGGGAAGGCGG-GGAGAAGAAAGGCAGGGATCCGGGCCTGGCGGC-3[prime]; mutant 3, 5[prime]-GCTGCGGGAATTCCGGGAGAAGAAAGG-CAGGGATCCGGGCCTGGCGGC-3[prime].
Probes were labeled with [[gamma]-32P]ATP and T4 polynucleotide kinase. Between 0.1 and 0.5 pmol of labeled probes were typically used in each reaction. Nuclear extracts from GKLF-transfected cells or recombinant GKLF were incubated in a volume of 30 µl containing 20 mM Tris-HCl, pH 7.2, 10 mM MgCl2, 150 mM NaCl, 20 mM KCl, 5 µM ZnCl2, 0.5 mM DTT, 5% glycerol and 1 µg poly(dI·dC) at room temperature for 30 min. The labeled probe was then added to the reaction and the incubation continued for another 30 min at 4°C. The DNA and DNA-protein complexes were resolved from one another on a native 6% polyacrylamide gel electrophoresis and visualized by autoradiography.
Site-directed mutagenesis
For site-directed mutagenesis, the 312 nt NcoI fragment in the 5[prime]-end of the GKLF gene was first subcloned into the plasmid pBluescript (Stratagene). Site-directed mutagenesis was performed using the QuikChange mutagenesis kit purchased from Stratagene and pairs of oligonucleotides containing the mutant sequences in the preceding section. Briefly, 50 ng template DNA was incubated in a 50 µl solution of 20 mM Tris-HCl, pH 8.8, 10 mM KCl, 10 mM (NH4)2SO4, 2 mM MgSO4, 125 ng each of the sense and antisense strands of the mutant oligonucleotides, 0.1% Triton X-100, 0.5% dimethylsulfoxide (DMSO), 0.1 mg/ml bovine serum albumin and 12.5 U Pfu DNA polymerase in a PCR of 95°C for 30 s, 55°C for 1 min, 68°C for 8 min for a total of 18 cycles. At the end of PCR, products were treated with 10 U DpnI for 1 h at 37°C. Ten microliters were used to transform Epicurian Coli XL1-Blue supercompetent cells (Stratagene) and the transformants screened by restriction digestion followed by sequencing to identify the mutants. The wild-type and mutated GKLF promoters were then unidirectionally recloned into the pGL2-Basic vector reporter plasmid and analyzed by transfection.
Transfection and reporter assays
Transient transfection by lipofection of Chinese hamster ovary (CHO) cells with the various DNA constructs was previously described (2,6). All transfections included the internal standard CMV-[beta]-galactosidase to normalize for transfection efficiency. Luciferase and galactosidase assays were performed as described (2,6,11).
RESULTS
Structure of the murine GKLF gene
Figure 1 depicts the structure of murine GKLF, contained within a single [lambda] phage clone. The gene has four exons, each containing a portion of the translated region. Over 6 kb of DNA of the gene was sequenced. All three exon-intron and intron-exon boundaries conform to the previously established consensus sequences (22; Table 1).
Figure 1. The structure of the murine GKLF gene. The four exons of the murine GKLF gene are identified by Roman numerals. The translated region or open reading frame is depicted in black. The locations of restriction sites for several endonucleases are labeled: Nc, NcoI; N, NotI; K, KpnI; H, HindIII; Xh, XhoI; Sm, SmaI; Xb, XbaI; Bg, BglII; B, BamHI.
Table 1. Nucleotide sequences at exon-intron junctions of the mouse GKLF gene
Upper case letters represent exon sequences and lower case letters represent intron sequences. Met is the initiation methionine codon. nt, nucleotides.
S1 nuclease analysis was performed on RNA specimens derived from various regions of mouse intestine or from liver to localize the start site of transcription of GKLF. The samples were incubated with a labeled NcoI DNA fragment that contained the putative transcription start site. As seen in Figure 2, two DNA fragments, one major and the other minor, were protected from digestion by the S1 nuclease when RNA from the small (lanes 2 and 3) or large (lanes 4 and 5) intestine was used. In contrast, neither tRNA (lane 1) nor RNA from the liver (lane 6) yielded any protected products. This result is consistent with our previous finding that GKLF is expressed in the intestinal tract but not in the liver (2). In addition, knowing the length of the protected fragments, we positioned the major start site of transcription to an adenine residue 601 nt 5[prime] of the beginning of the open reading frame.
Figure 2. S1 nuclease analysis of the start site of transcription of GKLF. Fifty micrograms of tRNA, RNA isolated from the proximal or distal portion of the mouse small and large intestine or from the liver were hybridized to the end-labeled NcoI DNA fragment in the 5[prime]-end of GKLF under the conditions described in Materials and Methods. Following hybridization, the RNA-DNA hybrids were digested with S1 nuclease and the products resolved on a sequencing gel. The accompanying sequence ladder was from the NcoI fragment. SI, small intestine; LI, large intestine.
Figure 3 shows the sequence for 1 kb of the 5[prime]-flanking region and 601 bp of the 5[prime]-untranslated region of GKLF. A single TATA box is located at 31 nt 5[prime] of the major start site of transcription. We also compared the sequences in Figure 3 with a database for those similar to established binding sites of known transcription factors. Beginning from the 5[prime]-end, there is a single site for Cdx-2 that partially overlaps an AP-1 binding site. Also, two GATA binding sites that flank another AP-1 site were identified. Further downstream, there is an inverted Sp1 site followed by an E-box. Lastly, three closely spaced GC-boxes, named boxes 1-3, are present just 5[prime] of the TATA box. No significant homology was noted to any transcription factor binding sites in the 5[prime]-untranslated region.
 |
 |
Figure 3. Nucleotide sequence of the 5[prime]-flanking region of the GKLF gene. The DNA sequence for 1000 bp of the 5[prime]-flanking region and 601 bp of the 5[prime]-untranslated region of the GKLF gene is shown. This sequence was compared with a computer database (TFSEARCH, http://pdaptrc.twcp.or.jp ) for sites similar to established transcription factor binding sites. A score of 90 was used as the lower threshold for homology. Shown are the sites that matched the sequences in the database. The three GC-boxes are numbered 5[prime]->3[prime]. The start site of transcription is designated +1 and identified by the rightward pointing arrow. Met indicates the initiation methionine of the open reading frame and rev indicates reverse. The two NcoI sites used in the S1 nuclease analysis, EMSA and DNase I footprint analysis are labeled.
GKLF binds to the immediate 5[prime]-flanking region of its own gene
We previously determined that GKLF binds to a relatively GC-rich sequence (20), which resembles the three GC-boxes in the immediate 5[prime]-flanking region of GKLF. To determine whether GKLF interacts with the 5[prime]-flanking region of its own gene, we performed EMSA on a labeled 312 nt NcoI fragment that contained all three GC-boxes (Fig. 4C). The source of GKLF was a full-length protein derived from transfected cells (Fig. 4A) or partially purified, bacterially produced recombinant protein that contained the zinc fingers (ZnF) of GKLF (Fig. 4B). As shown, both full-length GKLF (Fig. 4A, lane 2) and the zinc finger portion of GKLF (Fig. 4B, lanes 2-6) bound to the probe. A progressive retardation in mobility of the complexes was also noted when increasing amounts of recombinant protein were used (Fig. 4B). The results indicate that GKLF binds to the 5[prime]-flanking region of its own gene and that this interaction occurs either at multiple sites or multiple molecules of GKLF may bind to a single site.
Figure 4. EMSA between the 5[prime]-flanking region of the GKLF gene and GKLF. The 312 nt NcoI fragment at the 5[prime]-end of GKLF (C) was labeled as a probe. Twenty micrograms of nuclear extracts derived from COS-1 cells transfected with an expression construct containing full-length GKLF (PMT3-GKLF; lane 2) or vector alone (PMT3; lane 3) were used in the analysis shown in (A). In (B), increasing amounts (between 0 and 500 ng) of recombinant GKLF containing the zinc finger (ZnF) portion were used.
To further localize the site(s) of interaction between GKLF and the 5[prime]-end of its gene, we performed DNase I footprint experiments using a singly end-labeled NcoI fragment as a probe. As shown in Figure 5, all three GC-boxes were protected from DNase I digestion by the recombinant protein containing the zinc finger region of GKLF. The results of the footprint experiments therefore demonstrate the presence of multiple binding sites for GKLF in the 5[prime]-flanking region of GKLF.
Figure 5. DNase I footprint analysis of the 5[prime]-flanking region of the GKLF gene. The 312 nt NcoI fragment of the GKLF gene was labeled at its 5[prime]-end and used as a probe. Increasing amounts of recombinant protein containing the zinc fingers of GKLF were used in the experiments. The locations of the three GC-boxes coincided with the three footprinted areas and are identified as such.
The zinc fingers of GKLF bind to each of the first two GC-boxes as a monomer
The first two of the three GC-boxes in the 5[prime]-flanking region of GKLF are in close proximity to each other. To further examine the interaction between GKLF and these two elements, we performed EMSA using a radiolabeled oligonucleotide containing sequence between -134 and -87 of the GKLF 5[prime]-flanking region. As shown in Figure 6, two DNA-protein complexes were formed when either a constant amount of the probe was incubated with increasing amounts of recombinant GKLF (Fig. 6A) or a constant amount of the recombinant protein was incubated with increasing amounts of the probe (Fig. 6B). The protein composition of each of the two complexes was revealed by the results in Figure 7, in which one or both of the GC-boxes was mutated and analyzed by EMSA. As seen, while the formation of band 2 persisted with each of the two singly mutated oligonucleotides (Fig. 7A, mutants 1 and 2), band 1 largely failed to form. In contrast, both bands 1 and 2 were significantly diminished when both GC-boxes were mutated (Fig. 7A, mutant 3). These results suggest that the zinc fingers of GKLF bind to each of the two GC-boxes as a monomer, as depicted in Figure 6C. The binding equilibria between recombinant GKLF and each of the four probes were analyzed by quantifying the band intensities (Fig. 7B). The estimated Kd between GKLF and band 1 was 1.8 × 10-9 M and that between GKLF and band 2 was 1.3 × 10-9 M, both similar to the previously measured Kd between GKLF and another GC-rich binding site (11). It is also apparent that although the Kd between mutant 1 or mutant 2 and band 2 was largely unchanged (Fig. 7B, right), that between each of the two mutants and band 1 was significantly increased (Fig. 7B, left). These results confirm the preceding observation that band 1 represents the binding of two molecules of GKLF to each of the oligonucleotide probes.
Figure 6. EMSA of GKLF and an oligonucleotide containing the first two GC-boxes of the GKLF 5[prime]-flanking region. A double-stranded oligonucleotide containing the DNA between nt -134 and -87 of the 5[prime]-flanking region of GKLF was labeled. (A) EMSA was performed with 0.1 pmol probe and no protein (lane 1) or increasing amounts (between 100 and 400 ng) of recombinant GKLF (lanes 2-5). (B) An aliquot of 200 ng of the recombinant GKLF was used in each reaction which contained increasing amounts (0.01-0.5 pmol) of the probe. The locations of bands 1 and 2 are labeled. (C) The relationship between GKLF and each of the two bands. F indicates free probe.
A
 |
B
 |
Figure 7. EMSA of GKLF and wild-type or mutated oligonucleotides between nt -134 and -87 of the GKLF 5[prime]-flanking region. (A) EMSA performed with 200 ng of recombinant GKLF and increasing amounts (0.01-0.5 pmol) of labeled probe with either wild-type or mutated sequences. (B) Binding equilibrium between GKLF and each of the four oligonucleotides. The band intensity of band 1 or 2 and the respective probe was quantified by densitometry and plotted against the amount of probe used. The band that had the highest intensity is defined as 100%.
The 5[prime]-flanking region of the GKLF gene functions as a promoter and is autoregulated by GKLF
To determine whether 1 kb of the 5[prime]-flanking region of GKLF functions as a promoter, a fusion construct in which the sequence between nt -1000 and +550 of the 5[prime]-flanking region of GKLF linked in the sense orientation to a luciferase reporter in the promoterless pGL2-Basic vector DNA was generated. Upon transfection into CHO cells, this fusion construct (called -1.0 kb GKLF-pGL2-Luciferase) caused a dose-dependent increase in luciferase activity when compared to the pGL2-Basic vector (not shown). This suggests that the 1 kb 5[prime]-flanking region of GKLF functions as a promoter. To determine whether GKLF may influence the promoter activity of its own gene, we performed co-transfection experiments with the -1.0 kb GKLF-pGL2-Luciferase reporter and an expression construct containing either wild-type or mutated GKLF. As shown in Figure 8A, wild-type (PMT3-GKLF), but not a mutant GKLF in which its zinc fingers were deleted [PMT3-GKLF([Delta]ZnF)], induced reporter activity. Similarly, an expression construct containing Sp1, Sp3 or Cdx2 was able to transactivate the GKLF promoter (Fig. 8B). However, none of the three expression constructs containing GATA-4, -5 or -6 stimulated reporter activity despite the presence of two GATA-binding sequences in the GKLF promoter (not shown).
Figure 8. GKLF, Sp1, Sp3 and Cdx2 transactivate the GKLF promoter. Five micrograms each of the -1.0 kb GKLF-pGL2-Luciferase reporter DNA and the various effector constructs as shown in the figure were co-transfected into CHO cells. The fold induction was calculated by dividing the reporter activity from cells transfected with the transcription factor-containing constructs by that with the empty vector, PMT3 in (A) and CMV in (B). *P < 0.05 and [dagger]P < 0.01 when compared to the vector control by two-tailed Students' t-test.
Mutations in one or both of the first two GC-boxes diminish the ability of GKLF to transactivate its own promoter
Mutations in either one or both of the first two GC-boxes of the GKLF promoter caused a decreased interaction with GKLF. To determine whether this decreased interaction correlates with a decreased ability of GKLF to transactivate its promoter, we generated two mutant promoter-reporter constructs in which either GC-box 1 (mutant 1) or both GC-boxes 1 and 2 (mutant 3) were mutated. When these constructs were co-transfected into CHO cells with the control PMT3 vector, there was a slight (20%) decrease in basal promoter activity in mutant 1 when compared to the wild-type promoter (Fig. 9, lanes 1 and 3). Mutations in both GC-boxes further lowered the basal activity to 60% of that of the wild-type promoter (Fig. 9, lanes 1 and 5). When PMT3-GKLF was included in the transfection, there was a 3.7-fold induction of the wild-type promoter (lane 2). In contrast, the degree of induction by GKLF was attenuated for mutant 1 (2.3-fold; lane 4) and even more so for mutant 3 (1.4-fold; lane 6). These results suggest that mutations in either GC-box 1 or both GC-boxes 1 and 2 of the GKLF promoter had a deleterious effect on the ability of GKLF to transactivate its own promoter, indicating that both GC-boxes are necessary for optimal transactivation by GKLF.
Figure 9. Mutations in the first two GC-boxes diminish the ability of GKLF to transactivate its promoter. Generation of the wild-type and mutated promoter-reporter constructs were described in Materials and Methods. The wild-type construct contained the NcoI-NcoI fragment of GKLF (between nt -168 and +144) linked to the luciferase reporter. GC-box 1 in the promoter was mutated in mutant 1. Both GC-boxes 1 and 2 were mutated in mutant 3. CHO cells were co-transfected with either PMT3 or PMT3-GKLF and an equal amount of the wild-type or mutated reporter construct. The relative activity of the wild-type reporter co-transfected with the PMT3 vector control was taken as 1.
DISCUSSION
GKLF belongs to a rapidly expanding family of eukaryotic transcription factors containing the Krüppel homology (1). In descending orders of similarity, GKLF-related Krüppel-like proteins include LKLF (4), EKLF (5), BTEB2/IKLF (23,24), Zf9/CPBP (25,26), UKLF (27), BKLF (28) and FKLF (29). These proteins exhibit a diverse range of activities in regulating gene expression and are often involved in the control of cell- or tissue-specific proliferation or differentiation. GKLF, for example, appears to be an important regulator of epithelial tissue-specific expression of genes (7,8,11). Its physiological significance is demonstrated by the recent observation that late stage differentiation structures of the skin of mice deficient in GKLF are selectively perturbed, resulting in an inappropriate skin barrier function in the newborn (8). In addition, expression of GKLF is temporally associated with specific stages of the cell cycle (2). These findings suggest that expression of the gene encoding GKLF is likely to be stringently regulated. Further knowledge about this regulation may help advance our understanding of the biological function of GKLF.
The in vivo functions of several Krüppel-like factors including GKLF (8), EKLF (30-32) and LKLF (33-36) have been partly unraveled due to the availability of transgenic mice with selective gene ablation. However, the genomic regulation of only one gene, EKLF, has been characterized in some detail (37-39). More recently, genomic structures of human (40) and mouse (41) LKLF have become available. Isolation and characterization of the GKLF gene described herein offers the opportunity to examine and compare the regulation of expression of three closely related members of this group of seemingly important genes.
A distinct feature of the GKLF gene is the highly GC-rich nature of the sequence near its 5[prime]-end. Thus, the G+C content of the 1000 nt 5[prime]-flanking region is 67% and that of the 5[prime]-untranslated region is 63%. Moreover, the bulk of the GC residues are concentrated in the region between nt -600 and +300 of the gene where the G+C content is 82%. In fact, this region of the gene is a CpG island (42,43) with a frequency of CG dinucleotides at a high 14.6%. In view of the importance of methylation of CG dinucleotides in the regulation of expression of many eukaryotic genes, it is possible that methylation is also involved in the regulation of GKLF. Of note is that a similar CpG island is present in the 5[prime]-end of the gene encoding LKLF (40). In contrast, the 5[prime]-end of the EKLF gene does not have this feature (37). These findings would suggest that although the three Krüppel proteins are highly related (6), perhaps the mechanisms regulating expression of GKLF and LKLF are more conserved than EKLF.
As a first step in understanding the mechanisms regulating GKLF expression, we functionally tested its 1 kb 5[prime]-flanking region and showed that it behaves as a promoter. We also showed that the promoter is positively regulated by several established transcription factors including Sp1, Sp3 and Cdx-2, the latter being an intestine-specific transcription factor (17). However, despite the presence of several cis-elements resembling binding sites for GATA proteins, none of the GATA factors that we tested had an effect on the GKLF promoter. This is in distinct contrast to that of the EKLF gene, which has been shown to be regulated by GATA factors (37-39). Studies to date have not demonstrated a role for GATA factors in regulating expression of LKLF (40,41).
Another intriguing finding of the present study is the ability of GKLF to interact with the three GC-boxes immediately 5[prime] of the TATA box of its own gene. A result of this interaction is sequence-specific transactivation of the promoter by GKLF. Mutations of either a single or multiple sites in the GC-boxes resulted in a diminished ability of GKLF to transactivate the promoter. These findings indicate that GKLF participates in an autoregulatory circuit that positively controls expression of its own gene. We have recently observed that GKLF is a transactivator of the gene encoding the cyclin-dependent kinase inhibitor p21WAF1/Cip1 through a specific GC-rich cis-element present in the p21WAF1/Cip1 promoter (unpublished observations). Conversely, GKLF was shown to repress the promoter of the cyclin D1 gene (44). Thus, GKLF appears to be involved in the regulation of several genes with important functions in the control of the cell cycle. It may act positively to regulate growth-suppressive genes such as p21WAF1/Cip1 and GKLF, and negatively on a growth-promoting gene encoding cyclin D1. The net result of this combination may help explain the abundant expression of GKLF in the growth-arrested state and indeed suggests that GKLF may be necessary for the maintenance of cells in this particular state of the cell cycle.
ACKNOWLEDGEMENTS
We thank Drs Evans, Suske and Traber for providing plasmid DNA. This work was in part supported from grants from the National Institutes of Health (DK44484 to V.W.Y. and DK53839 to K.H.K.).
REFERENCES
*To whom correspondence should be addressed at: Department of Medicine, Ross 918, The Johns Hopkins University School of Medicine, 720 Rutland Avenue, Baltimore, MD 21205, USA. Tel: +1 410 955 9691; Fax: +1 410 955 9677; Email: vyang{at}welch.jhu.edu
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Liu, H. Zhang, L. Zhu, L. Zhao, and Y. Dong Kruppel-Like Factor 4 Is a Novel Mediator of Selenium in Growth Inhibition Mol. Cancer Res., February 1, 2008; 6(2): 306 - 313. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Wei, X. Wang, B. Gan, A. M. Urvalek, Z. K. Melkoumian, J.-L. Guan, and J. Zhao Sumoylation Delimits KLF8 Transcriptional Activity Associated with the Cell Cycle Regulation J. Biol. Chem., June 16, 2006; 281(24): 16664 - 16671. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Kang, O. Nielsen, C. Fenger, G. Leslie, U. Holmskov, and K. B.M. Reid Induction of DMBT1 expression by reduced ERK activity during a gastric mucosa differentiation-like process and its association with human gastric cancer Carcinogenesis, June 1, 2005; 26(6): 1129 - 1137. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W.-W. Huang, Y. Yin, Q. Bi, T.-C. Chiang, N. Garner, J. Vuoristo, J. A. McLachlan, and L. Ma Developmental Diethylstilbestrol Exposure Alters Genetic Pathways of Uterine Cytodifferentiation Mol. Endocrinol., March 1, 2005; 19(3): 669 - 682. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. S. Yoon and V. W. Yang Requirement of Kruppel-like Factor 4 in Preventing Entry into Mitosis following DNA Damage J. Biol. Chem., February 6, 2004; 279(6): 5035 - 5041. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Jaubert, J. Cheng, and J. A. Segre Ectopic expression of Kruppel like factor 4 (Klf4) accelerates formation of the epidermal permeability barrier Development, June 15, 2003; 130(12): 2767 - 2777. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. T. Dang, W. Zhao, C. S. Mahatan, D. E. Geiman, and V. W. Yang Opposing effects of Kruppel-like factor 4 (gut-enriched Kruppel-like factor) and Kruppel-like factor 5 (intestinal-enriched Kruppel-like factor) on the promoter of the Kruppel-like factor 4 gene Nucleic Acids Res., July 1, 2002; 30(13): 2736 - 2741. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. P. Katz, N. Perreault, B. G. Goldstein, C. S. Lee, P. A. Labosky, V. W. Yang, and K. H. Kaestner The zinc-finger transcription factor Klf4 is required for terminal differentiation of goblet cells in the colon Development, January 6, 2002; 129(11): 2619 - 2628. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. A. Miller, E. A. Eklund, M. L. Peddinghaus, Z. Cao, N. Fernandes, P. W. Turk, B. Thimmapaya, and S. A. Weitzman Kruppel-like Factor 4 Regulates Laminin alpha 3A Expression in Mammary Epithelial Cells J. Biol. Chem., November 9, 2001; 276(46): 42863 - 42868. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. T. Ziemer, D. Pennica, and A. J. Levine Identification of a Mouse Homolog of the Human BTEB2 Transcription Factor as a {beta}-Catenin-Independent Wnt-1-Responsive Gene Mol. Cell. Biol., January 15, 2001; 21(2): 562 - 574. [Abstract] [Full Text]
|
|
|
|
|
|
W. Zhang, D. E. Geiman, J. M. Shields, D. T. Dang, C. S. Mahatan, K. H. Kaestner, J. R. Biggs, A. S. Kraft, and V. W. Yang The Gut-enriched Kruppel-like Factor (Kruppel-like Factor 4) Mediates the Transactivating Effect of p53 on the p21WAF1/Cip1 Promoter J. Biol. Chem., June 9, 2000; 275(24): 18391 - 18398. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. Bieker Kruppel-like Factors: Three Fingers in Many Pies J. Biol. Chem., September 7, 2001; 276(37): 34355 - 34358. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||