|Home | About | Journals | Submit | Contact Us | Français|
Krüppel-like factor 4 (KLF4 or GKLF) is an inhibitor of the cell cycle. The gene encoding KLF4 is localized on chromosome 9q, previously shown to exhibit allelic loss in colorectal cancer (CRC). In this study, we show that the mean level of KLF4 mRNA in a panel of 30 CRC was 52% that of paired normal colonic tissues. Similarly, the levels of KLF4 mRNA and protein in a panel of six established CRC cell lines were significantly lower than those of an untransformed colonic epithelial cell line. Using highly polymorphic DNA markers that .ank the KLF4 locus, we found evidence for loss of heterozygosity (LOH) in two of eight surgically resected CRC specimens. In addition, LOH was observed in .ve of six CRC cell lines with one additional cell line exhibiting hemizygous deletion in the KLF4 gene. We also found that the 5′-untranslated region of KLF4 was hypermethylated in a subset of resected CRC specimens and cell lines. Lastly, the open-reading frame of KLF4 in two of three CRC cell lines examined contained several point mutations that resulted in a diminished ability to activate the p21WAF1/Cip1 promoter. These findings indicate that KLF4 is a potential tumor suppressor gene in CRC.
Colorectal cancer (CRC) is the third most common cancer in the United States. An estimated 147 500 new cases of CRC are expected to be diagnosed in 2003 (Jemal et al., 2002; American Cancer Society, 2003). In the same year, it is estimated that CRC will take 57 100 lives, making it the second most common cause of cancer death in men and women. Every American has a 6% lifetime risk of developing CRC. These statistics render CRC a major health concern in this country.
Significant progress has been made in understanding the molecular mechanisms that lead to CRC (Kinzler and Vogelstein, 1996 Kinzler and Vogelstein, 1997; Ilyas et al., 1999; Yang, 1999). Fearon and Vogelstein (1990) presented evidence for a multistep genetic model of colorectal carcinogenesis. This model is based on the understanding that CRC is the result of accumulation of genetic changes (mutations) in key genes, including the inactivation of tumor suppressor genes and aberrant activation of proto-oncogenes. In support of the multistep genetic paradigm is the frequent presence of loss of heterozygosity (LOH) of specific chromosomal loci that harbor tumor suppressor genes as a consequence of a generalized genetic instability often accompanying the development of CRC (Goel et al., 2003). For example, LOH of chromosome 5q is correlated with biallelic inactivation of the tumor suppressor gene, adenomatous polyposis coli (APC), a process that occurs early in the multistep colorectal carcinogenesis (Fearon and Vogelstein, 1990). Other frequently observed LOH in CRC involves chromosomes 1p, 8p, 17p and 18q (Thiagalingam et al., 2001).
Krüppel-like factor 4 (KLF4), also known as gutenriched Krüppel-like factor (GKLF), is a member of the KLF family of zinc-finger-containing transcription factors (Dang et al., 2000a; Bieker, 2001). KLF4 was first identified as a gut epithelially enriched gene, the expression of which is associated with growth arrest (Shields et al., 1996). Constitutive expression of KLF4 suppresses cell proliferation (Shields et al., 1996; Geiman et al., 2000) by blocking G1/S progression of the cell cycle (Chen et al., 2001). Importantly, KLF4 expression is stimulated following DNA damage in a p53-dependent fashion (Zhang et al., 2000) and this stimulation leads to the transcriptional upregulation of the gene encoding the cyclin-dependent kinase inhibitor, p21WAF1/Cip1 (Zhang et al., 2000). Recent genetic evidence suggests that the induction of KLF4 is essential for mediating the G1/S checkpoint function of p53 following DNA damage (Yoon et al., 2003).
Consistent with a DNA-damage checkpoint function for KLF4, we and others have reported that KLF4 expression is altered in various models of intestinal tumorigenesis. For example, the level of KLF4 transcript is significantly decreased in intestinal adenomas of multiple intestinal neoplasia (APCMin/+) mice and in colonic adenomas of familial adenomatous polyposis (FAP) patients when compared to the respective adjacent normal mucosa (Ton-That et al., 1997; Dang et al., 2000b). Similarly, KLF4 mRNA level is decreased in sporadic colonic adenomas and carcinomas of patients when compared to normal colonic tissues (Shie et al., 2000b). The reason for this decrease is in part due to the fact that KLF4 is regulated by APC (Dang et al., 2001), a gene that is mutated in FAP (Groden et al., 1991; Kinzler et al., 1991; Nishisho et al., 1991) and more than 80% of sporadic CRC (Miyoshi et al., 1992; Powell et al., 1992). Moreover, the fact that allelic loss of chromosome 9q, in which KLF4 resides (Garrett-Sinha et al., 1996), is observed in 25–50% of sporadic CRC (Gryfe et al., 1997) raises the possibility that KLF4 itself is a tumor suppressor gene for CRC. In this study, we present evidence that the expression of KLF4 is significantly reduced in a panel of sporadic CRC and colon cancer cell lines. We also show evidence for LOH of the KLF4 locus in a subset of surgically resected CRC specimens and cell lines. Additional explanation for the reduction in KLF4 expression is provided by hypermethylation in the 5′-untranslated region (5′-UTR) of KLF4 in a portion of the CRC specimens and colon cancer cell lines as well as point mutations in the reading frame of KLF4 in two cell lines. These findings support the notion that KLF4 is a tumor suppressor gene in CRC.
Figure 1a shows the strategy for synthesizing specific 3′-untranslated region (3′-UTR) probes for KLF4 and KLF5, respectively. KLF5 is a closely related Krüppellike factor to KLF4 that is also expressed in the intestinal epithelium (Conkright et al., 1999; Ohnishi et al., 2000), but is associated with proliferation (Sun et al., 2001) rather than growth arrest as in KLF4 (Shields et al., 1996). When hybridized under high stringency to RNA obtained from a human colon cancer cell line HCT116, the probes specifically detected the corresponding KLF4 and KLF5 mRNAs (Figure 1b). Results of Northern blot analysis for human p21WAF1/Cip1 and the internal control ubiquitin were also shown when the RNA was probed with the respective full-length cDNAs (Figure 1b).
To determine the relative level of expression of KLF4 in CRC compared to normal colon, we analysed a cancer-profile array (BD BioSciences Clontech) that contained cDNAs from 30 pairs of adenocarcinoma and matched adjacent normal tissues from patients who underwent resection for CRC (11 males and 19 females; ages ranged from 40 to 77 years). The same array was also probed with the 3′-UTR KLF5 probe and fulllength p21WAF1/Cip1 and ubiquitin probes (Figure 2a). Figure 2b shows the mean relative expression ratios for KLF4, KLF5 and p21WAF1/Cip1 after normalization to the signals from the ubiquitin control. The mean level of expression of KLF4 in CRC was 52% of that in matched adjacent normal colonic tissues (P <0.0005). There was also a statistically significant reduction in the level of expression of p21WAF1/Cip1 in CRC when compared to normal tissues (74%; P <0.005). In contrast, no statistically significant difference was noted for KLF5 between CRC and matched normal tissues (P = 0.3).
We also determined the levels of KLF4 mRNA in six different cell lines derived from human colon cancers by Northern blot analysis (Figure 3a). The same blot also contained RNA isolated from an untransformed normal human colonic epithelial cell line, FHC. As seen in Figure 3b, the relative level of KLF4 mRNA in each of the CRC cell line was significantly lower than that in the normal FHC cell line. Moreover, only one cell line, T84, contained detectable amounts of KLF4 protein when examined by Western blot analysis, although its level was not as high as that in FHC (Figure 3c). Results from Figures 2 and and33 indicate that the expression of KLF4 is reduced in both surgically resected CRC specimens and cell lines.
To explore the reasons responsible for the decreased KLF4 expression in CRC, we examined surgically resected CRC specimens and matched adjacent normal colonic tissues for evidence of LOH using several highly polymorphic DNA makers flanking the KLF4 locus on chromosome 9q (Figure 4b). As seen in Figure 4a, CRC from two of eight patients tested demonstrated LOH in all three markers. Similarly, five of six colon cancer cell lines had evidence of LOH in one or more markers (Figure 4a). A single cell line, RKO, showed no evidence of LOH, although the level of KLF4 mRNA was one of the lowest among the panel of colon cancer cell lines investigated (Figure 3a). To further examine the mechanism that accounts for this discrepancy, we performed Southern blot analysis of genomic DNA extracted from the six cell lines using full-length KLF4 cDNA probe. As seen in Figure 5a, b, with the exception of RKO, each cell line exhibited a single band when the DNA was digested with EcoRI and HindIII, respectively. In contrast, two bands were apparent when DNA from RKO was used (Figure 5a). This suggests that there is a hemizygous deletion of the KLF4 gene in RKO cells, as illustrated in Figure 5b.
To investigate additional mechanisms responsible for the decreased expression of KLF4 in CRC, methylationspecific polymerase chain reaction (MSP) using primers specific for the 5′-UTR of KLF4 was performed on genomic DNA extracted from surgically resected CRC specimens and matched normal colonic tissues as well as from the six colon cancer cell lines. The region selected for MSP is between nucleotides 79 and 355 in the 5′-UTR and contains a CpG island (Garrett-Sinha et al., 1996). As seen in Figure 6a, tumors from two of the eight patients (patient nos. 5 and 8) demonstrated evidence of methylation in the KLF4 5′-UTR. None of the paired adjacent normal tissues had methylation in the same region. In addition, one of the six colon cancer cell lines, RKO, had evidence of methylation (Figure 6b). Direct DNA sequencing of the PCR products using methylation-specific primers revealed that every one of the 22 CpG dinucleotides in the amplified region of the KLF4 sequence was methylated (data not shown). These results suggest that hypermethylation may contribute to the reduction in KLF4 expression in a subset of CRC and colon cancer cell lines.
To determine whether mutations in KLF4 may be present in CRC, we sequenced the entire coding region of KLF4 in three different colon cancer cell lines, HCT116, HT29 and RKO, using reverse transcription– polymerase chain reaction (RT–PCR). The sequences were compared to those deposited in GenBank (no. NM004235), which were derived from cloned KLF4 cDNA from normal tissues (Yet et al., 1998). Two of six clones of cDNA amplified from HCT116 had point mutations in the same codon, resulting in a conservative change in amino acid in one and no change in the other (Figure 7a). All six cDNA clones obtained from HT29 demonstrated point mutations in multiple locations, resulting in several amino-acid changes (Figure 7b). No mutation was detected in cDNA amplified from RKO cells (results not shown).
To further determine whether mutations in the KLF4 coding region from HT29 could affect the function of the protein, we cloned the mutated cDNA sequence and wild-type sequence into expression vectors and analysed the function of the expressed proteins by immunocytochemistry and promoter activation in transfected cells. As seen in Figure 8a, wild-type KLF4 is evenly distributed in the nucleus, excluding the nucleoli (panel a). In contrast, the mutant KLF4 obtained from HT29 showed a flocculated distribution in the nucleus (panel d). When cotransfected with a p21WAF1/Cip1 promoterluciferase construct, the mutant KLF4 exhibited a diminished ability to transactivate the promoter when compared to the wild-type KLF4 (Figure 8b). These results suggest that mutations in KLF4 from HT29 cells adversely affect the function of KLF4.
KLF4 was initially identified as a transcription factor the expression of which is associated with the growth arrest process induced by conditions such as serum deprivation and contact inhibition (Shields et al., 1996). KLF4 was also found to have an inhibitory effect on cell proliferation when overexpressed in transfected cells (Shields et al., 1996; Geiman et al., 2000). Recent studies showed that KLF4 expression is induced upon DNA damage and that such induction leads to a block in the transition between the G1 and S phase of the cell cycle (Zhang et al., 2000; Chen et al., 2001; Yoon et al., 2003). Moreover, the DNA damage-induced activation of KLF4 is dependent on p53 and that KLF4 activation results in the induction of the p21WAF1/Cip1 gene (Zhang et al., 2000; Chen et al., 2001; Yoon et al., 2003). Indeed, genetic and biochemical evidence suggests that KLF4 is an essential mediator of p53 in controlling the G1/S progression of the cell cycle in response to DNA damage (Yoon et al., 2003). Other studies have also documented a cell cycle-inhibitory function of KLF4 by showing its ability to repress the genes encoding cyclin D1 and ornithine decarboxylase (Shie et al., 2000a; Chen et al., 2002). These observations provide strong evidence that KLF4 has an important checkpoint function, in a manner similar to other established tumor suppressors such as p53 (Baker et al., 1990a, b; Hollstein et al., 1991).
Owing to the growth-inhibitory effect of KLF4, we and others have investigated the expression of KLF4 during intestinal tumorigenesis. We demonstrated that the level of KLF4 transcript is significantly decreased in the intestine of APCMin/+ mice during a period of tumor formation when compared to age-matched control littermates (Ton-That et al., 1997). Moreover, we showed that the level of KLF4 transcript is highest in normal-appearing intestinal tissues and decreases as the size of the adenomas increases in tumors derived from the intestine of APCMin/+ mice (Dang et al., 2000b). A similar decrease is present in the colonic adenomas from FAP patients when compared to the normal-appearing colonic mucosa (Dang et al., 2000b). A related study showed a similar decrease in KLF4 mRNA levels in sporadic adenomas and cancer of the colon (Shie et al., 2000b). These findings are consistent with those derived from the present study in which KLF4 transcript levels in 30 CRC are reduced by close to 50% when compared to matched normal colonic tissues in a highly statistically significant manner. A similar reduction in p21WAF1/Cip1 transcript levels was also observed in CRC, raising the possibility that the reduction in p21WAF1/Cip1 expression is a direct consequence of the reduced KLF4 expression in CRC. Our current study also demonstrated that the expression of KLF4 in six of the colon cancer cell lines surveyed is much lower than that in an untransformed colonic epithelial cell line. These findings are in support of a putative role for KLF4 as a tumor suppressor.
The present study is the first to provide evidence for a role of KLF4 in functioning as a tumor suppressor gene in CRC. They include LOH of the chromosomal loci flanking KLF4 and hypermethylation of the KLF4 5′-UTR in a subset of surgically resected CRC specimens. Although we did not document a direct correlation between the decrease in KLF4 expression and the surgical CRC specimens due to technical limitations, the fact that each of the above scenarios has been implicated in contributing to a tumor suppressor phenotype provides support for the potential role of KLF4 as a tumor suppressor in CRC. Additional support for KLF4’s role as a tumor suppressor is provided by studies involving CRC cell lines. Thus, five of six CRC cell lines examined demonstrated LOH near the KLF4 locus. In addition, hypermethylation, hemizygous deletion and point mutations affecting functions of KLF4, are detected among the cell lines, all of which have significantly reduced levels of KLF4 expression when compared to a nontransformed colonic epithelial cell line. Some of the observed LOH in the CRC cell lines are noncontinuous (e.g. SW480 and T84), suggesting that additional mechanisms such as point mutations may be implicated in the inactivation of the KLF4 gene in these cells. Be that as it may, each of the mechanisms documented above is likely to contribute to the reduction in KLF4 expression in CRC in addition to the previous observation that KLF4 is a downstream target of two CRC tumor suppressors, APC (Dang et al., 2001; Stone et al., 2002) and p53 (Zhang et al., 2000; Yoon et al., 2003).
Regional hypermethylation plays an important role in the alteration of gene expression in carcinoma formation and in the progression of carcinoma (Bird, 1996; Jubb et al., 2001; Jones and Baylin, 2002). We found evidence for hypermethylation in the KLF4 5′-UTR in a subset of surgically resected CRC specimens. Interestingly, the same specimens also demonstrated LOH in the KLF4 allele. Similarly, the lone colon cancer cell line, RKO, which demonstrated KLF4 methylation, also had a hemizygous deletion of the KLF4 gene. Whether these results indicate a preferential association between LOH or gene deletion and hypermethylation of KLF4 in CRC is unclear at this time. Nonetheless, the presence of both defects in the same tumor specimen is likely to lead to a significant reduction in KLF4 expression.
The finding that the colon cancer cell line, RKO, has the lowest KLF4 mRNA level among all colon cancer cell lines surveyed merits some additional discussion. The same observation has also been noted previously (Dang et al., 2001). The latter study demonstrated that the reduced level of KLF4 mRNA is due to a dominant-negative CDX2 mutant present in RKO cells (Dang et al., 2001). The present study provides additional explanations for the nearly negligible level of KLF4 transcript in RKO cells that include the hemizygously deleted and hypermethylated KLF4 gene. Combined together, these mechanisms may contribute to the aggressive tumor phenotype of RKO cells as they are defective in the checkpoint function of KLF4. This is consistent with our recent finding that re-expression of KLF4 in RKO cells significantly reduced tumorigenicity of this cell line as demonstrated by reduced colony formation in vitro and tumor growth in vivo (Dang et al., 2003).
Paraffin-embedded tissue samples of colon cancer and normal adjacent colonic tissues were obtained from patients who underwent surgical resection for adenocarcinoma of the colon. All patient identifiers have been removed prior to the study, which was approved by the Emory University Institution Review Board (IRB #446-2002). The colon cancer profiling arrays containing cDNA amplified from paired CRC and normal tissues were purchased from BD BioSciences Clontech (Palo Alto, CA, USA). All colon cancer cell lines were purchased from the American Type Culture Collection (ATCC) (Manassas, VA, USA) and were propagated in the media according to the conditions recommended by ATCC at 37°C in a 5% CO2 atmosphere.
RNA was isolated using the Trizol method (Invitrogen Inc., Carlsbad, CA, USA). Total RNA (5 μg) was treated with DNase I per protocol from Ambion, Inc. (Austin, TX, USA). In total, 1 μg of the treated RNA was used for first-strand cDNA synthesis using the SuperScript First-Strand Synthesis System for RT–PCR from Invitrogen. The resulting firststrand cDNA was subsequently treated with RNase H (Invitrogen). A measure of 2 μl of this cDNA were included in each PCR. Primer sequences for the respective cDNA were as follows: 5′-AAAAGACAAAAATCAAAGAACAGA-3′ and 5′-CATATTTATTAAGGCAAGCAACTA-3′ (KLF4 3′-UTR probe); 5′-TTGGCTTCGTTTCTTCTCTTCGTTGACTTT-3′ and 5′-TCTCTTCTGGCAGTGTGGGTCAT-3′ (KLF4 full coding region); and 5′-GTTGCCATTTTCAGCCACCAGAGT-3′ and 5′-GCCAGTTTAGAAGCAATTGTAGCAGCATAG-3′ (KLF5 3′-UTR probe). The Fast-Start Taq Polymerase (Roche Molecular Biochemicals, Indianapolis, ID, USA) was used for RT–PCR. In total, 35 cycles were used to amplify the various cDNAs. Reactions were subjected to 94°C for 5 min, followed by 35 cycles at 94°C, 30 s; 55°C, 30 s; and 72°C, 60 s. All reactions were then subjected to a further extension of 7 min at 72°C. PCR products were visualized on a 1.5% nondenaturing agarose gel stained with ethidium bromide.
DNA samples were isolated using the QIAamp DNA Mini Kit from Qiagen (Valencia, CA, USA). Extraction of genomic DNA from cell lines was conducted according to standard procedures. Southern blot analysis was performed using 5 μg of genomic DNA digested with EcoRI and HindIII. A fulllength KLF4 cDNA spanning the entire ORF was generated by RT–PCR with the aid of primers for the full coding region as described above was used as the probe.
Western blot analysis was performed as previously described (Yoon et al., 2003). The blots were analysed against KLF4 (Shields et al., 1996) and β-actin using the respective primary and secondary antibodies.
Northern blot analysis was performed using 20 μg total RNA isolated from various colon cancer cell lines. The DNA probe was labeled by random priming with [α-32P]dATP (Perkin-Elmer, Boston, MA, USA). Blots were hybridized at 68°C for 1 h in QuikHyb solution (Stratagene, La Jolla, CA, USA) and washed according to a standard protocol.
The colon cancer profiling array (BD Biosciences Clontech; Palo Alto, CA, USA) was hybridized with the 3′-UTR cDNA probe for KLF4 or KLF5. A full-length cDNA clone for p21WAF1/CIP1 was a generous gift from Dr B Vogelstein (Johns Hopkins Oncology Center, Baltimore, MD, USA). Human ubiquitin were used as an internal loading control (Clontech). The slide was prehybridized for 30m with 3ml of ExpressHyb solution (Clontech) at 68°C. The radioactively labeled cDNA probe totaling 6 × 106 c.p.m. was mixed with 300 μg of sheared salmon sperm DNA (Stratagene). The mixture was boiled for 5 min, quenched on ice for 2 min and mixed with 3ml of fresh ExpressHyb solution. The filter was hybridized overnight at 68°C and washed twice at room temperature for 30 min with 2 × SSC/0.5% SDS, once with 2 × SSC/0.5% SDS at 60°C for 30m and once with 0.1 × SSC/0.5% SDS at 60°C for 30m and finally for 5m at room temperature with 2 × SSC. All filters and slides were exposed to a phosphor imager, Typhoon 9200, and the signal intensities quantified with the Image Quant software (Molecular Dynamics, Piscataway, NJ, USA).
Three highly polymorphic microsatellite markers: D9S53 (nine q22.3–31), D9S105 (nine q22.3–31) and D9S302 (nine q31–33) were selected for LOH analysis (Human MapPairs, Research Genetics, Huntsville, AL, USA). The total distance from D9S53 to D9S302 is estimated at 13.48 cm or 9.5Mb (http://cedar.genetics.soton.ac.uk/public_html/read.html). Primers specific for each marker were tested in matched normal colon and colon tumor DNA obtained from eight patients. Total reaction volumes were 25 μl containing 40 ng DNA, 200 μm dNTPs, 10mm Tris-HCl, pH 7.8, 50mm KCl, 1.5mm MgCl2, 0.3 μm each fluorescently labeled primer and 0.5U of Taq DNA polymerase (Promega, Madison, WI, USA). After 5 min of initial denaturation at 95°C, amplification was performed for 30 cycles (95°C for 30 s, 58°C for 30 s and 72°C for 30 s) with a final extension cycle at 72°C for 7 min. Products were electrophoresed on 4% agarose gels to confirm the specificity of the amplification reactions. The amplified products were then separated by capillary electrophoresis using an ABI Prism 3100 Genetic Analyzer (Applied Biosystems; Foster City, CA, USA). Following electrophoresis, the raw genotype files were analysed using ABI GeneMapper software (Applied Biosystems). LOH was scored by allelic imbalance between the normal and tumor tissue samples and defined by more than 50% loss of intensity of one allele in the tumor sample when compared with the same allele from the matched normal tissue.
MSP was performed on genomic DNA using 0.4 μg. DNA was modified with bisulfite as previously described (Herman et al., 1996). For unmethylated KLF4 5′-UTR, the forward primer is 5′-ATTAGTTTGTAGTTTTGTGTTATGG-3′ and the reverse primer is 5′-TCTTTAAATTAAATA TAACTTAAAAACATC-3′, which were designed to amplify a 275 bp sequence between nucleotides +80 and +355 in the 5′-UTR of the human KLF4 gene (Garrett-Sinha et al., 1996). For methylated KLF4 5′-UTR, the forward primer is 5′-GATTAGTTCGTAGTTTCGCGTTAC-3′ and the reverse primer is 5′-AAATTAAATATAACTTAAAAACGT-3′, which were designed to amplify a 271 bp sequence between nucleotides +79 and +350 in the 5′-UTR of the human KLF4. The negative control was water. The positive control for the methylation-specific reaction was in vitro methylated DNA (IVD) of normal human genomic DNA from Promega (Madison, WI, USA) that was treated with Sss1 methylase per protocol from New England Biolabs (Beverly, MA, USA). This treatment renders all cytosine residues in a CpG dinucleotide methylated. Each PCR reaction of 50 μl consisted of 40 ng DNA and 200 nm each of the forward and reverse primer in 1 × Hot Start Taq Master Mix from Qiagen (Valencia, CA, USA). Each PCR reaction was hot started at 94°C for 15 min and then denatured at 95°C for 30 s, annealed at 57°C for 30 s and extended at 72°C for 30 s, for a total of 35 cycles. PCR products were visualized on a 3% agarose gel stained with ethidium bromide. The products were also directly sequenced in order to determine the exact locations of CpG methylation in the amplified region.
First-strand cDNA was synthesized from total RNA isolated from HCT116, HT29 and RKO cell lines. Two primers, 5′-TTGGCTTCGTTTCTTCTCTTCGTTGACTTT-3′ and 5′-TTCTCTTCTGGCAGTGTGGGTCAT-3′, were used to amplify the full coding region of KLF4. The resultant PCR products were purified using the Qiaquick PCR Purification Kit (Qiagen) and sent to the Molecular Genetics Facility at the University of Georgia, Athens, GA, USA for sequencing. Once confirmed by sequencing, wild-type or mutated fulllength KLF4 cDNA fragment was subcloned into the eucaryotic expression vector pcDNA3.1D/V5-His-TOPO vector (Invitrogen) for further transfection studies.
Transient transfection of NIH3T3 cells using lipofection were performed as previously described (Shields et al., 1996). Briefly, a luciferase reporter linked to 2.4 kb of the 5′-flanking region of the p21WAF1/Cip1 gene (Zhang et al., 2000) was cotransfected with pcDNA3.1/wild-type KLF4, pcDNA3.1/mutant KLF4 (HT29) or pcDNA3.1 vector alone. Luciferase activities were measured 24 h following transfection and normalized to the activity of an internal control, pRL-CMV (Promega) using the Dual-Luciferase Assay (Promega).
Immunofluorescence was performed on NIH3T3 cells on coverslips and transfected with pcDNA3.1/wild type. KLF4 or pcDNA3.1/mutant KLF4 (HT29). At 48 h following transfection, cells were washed twice in phosphate-buffered saline (PBS) and incubated in methanol at −20°C for 15 min and then washed twice in cold PBS. Cells were blocked for 1 h with 0.5% bovine serum albumin at 4°C. Cells were then incubated for 1 h with V5-FITC antibody (Invitrogen) at 1 : 500 dilutions and then washed three times for 5 min each with PBS. The nuclear stain ToPRO3 (Molecular Probes) was added to the final PBS wash (200 ng/ml). Coverslips were then mounted with mounting media (Sigma) and observed under an LSM 510 Zeiss confocal microscope.
We thank Dr Bert Vogelstein for providing the full-length cDNA probe for p21WAF1/Cip1. This work was in part supported by grants from the National Institutes of Health (DK52230, DK64399, CA84197). VWY is the recipient of a Georgia Cancer Coalition Distinguished Cancer Clinician Scientist Award.