The Promyelocytic Leukemia Zinc Finger (PLZF ) gene is a novel transcriptional target of the CCAAT-Displacement- Protein (CUX1) repressor Isabelle Fre ´ chette, Mathieu Darsigny, Karine Brochu-Gaudreau, Christine Jones and Franc¸ois Boudreau De ´ partement d’Anatomie et Biologie Cellulaire, Faculte ´ de Me ´ decine et des Sciences de la Sante ´ , Universite ´ de Sherbrooke, Que ´ bec, Canada Introduction The establishment of the functional adult intestine is the result of several steps, beginning with gross mor- phogenesis of the digestive tract and followed by cyto- differentiation of the epithelium and the induction of intestine-specific genes. An exchange of signals between the endoderm and mesoderm leads to successive transi- tions that result in the specialization of the intestinal tube into the small intestine, cecum and colon along the proximal–caudal axis [1]. The coordination of gene transcription is crucial in the orchestration of position- dependent intestinal epithelial development, differentia- tion and homeostasis. An interesting hypothesis is that intestinal disorders, such as colorectal cancer (CRC), could evolve from the subtle accumulation of genetic alterations in the transcriptome that leads to a diver- gence of these cells from their original identities [2]. Colorectal adenomas and carcinomas are often associ- ated with phenotypic and genetic characteristics of the Keywords colorectal cancer; CUX1; intestinal epithelium; PLZF; transcriptional repression Correspondence F. Boudreau, De ´ partement d’Anatomie et Biologie Cellulaire, Faculte ´ de Me ´ decine et des Sciences de la Sante ´ , 3001 12e ave Nord, Sherbrooke, Que ´ bec J1H 5N4, Canada Fax: 001 819 564 5320 Tel: 001 819 820 6876 E-mail: francois.boudreau@usherbrooke.ca (Received 2 June 2010, revised 11 August 2010, accepted 16 August 2010) doi:10.1111/j.1742-4658.2010.07813.x The CCAAT-Displacement-Protein (CUX1) can transcriptionally repress sucrase–isomaltase gene expression, a specific product of enterocytes that becomes re-expressed during human colonic polyposis. Little is known of the gene repertoire that is directly affected by CUX1 in the intestinal epithelial context. This article identifies the Promyelocytic Leukemia Zinc Finger (PLZF) gene as a transcriptional target for the CUX1 repressor. CUX1 interacts in vivo with multiple DNA-binding sites in the 5¢-UTR and promoter of the PLZF gene in colorectal cancer cells, a region that is func- tionally targeted by CUX1 in cotransfection assays. PLZF was found to be induced in colorectal cancer cell lines, correlating with a low detectable level of CUX1, a pattern that was reversed in normal human colonocytes. Reduction of p200CUX1 expression by RNAi in the Caco-2 ⁄ 15 cell line increased PLZF gene transcript expression. Because of the implication of Plzf in the regulation of stem cell maintenance, as well as Wnt and Ras sig- naling, in other systems, our observations suggest that the novel genetic relationship between CUX1 and PLZF could be of relevance to human diseases, such as leukemia, and open up a new field of investigation for the implication of these regulators during intestinal polyposis and cancer. Abbreviations BTB ⁄ POZ, bric-a ` -brac, tramtrack, brad complex ⁄ poxvirus zinc finger; ChIP, chromatin immunoprecipitation; CR, cut repeat; CRC, colorectal cancer; CRESIP, colonic repression of the sucrase–isomaltase gene; CUX1, CCAAT-Displacement-Protein; EMSA, electrophoretic mobility shift assay; HD, homeodomain; PLZF, Promyelocytic Leukemia Zinc Finger; RARa, retinoic acid receptor a; shRNA, short hairpin RNA; SI, sucrase–isomaltase. FEBS Journal 277 (2010) 4241–4253 ª 2010 The Authors Journal compilation ª 2010 FEBS 4241 normal adult small intestinal epithelium [3–6]. During embryonic life, the colon passes through a transient developmental stage that reproduces the small intesti- nal phenotype with the expression of small intestinal- specific genes, such as sucrase–isomaltase (SI) [7]. It has been suggested that the mechanisms involved in fetal colon development are recapitulated during colo- nic neoplasia in adults. An example is the re-expression of the SI gene in the majority of colonic adenomatous polyps and adenocarcinomas [3,4,8]. Therefore, it is clear that the identification of the molecular regulators responsible for intestinal epithelial establishment, determination and maturation is crucial to the provision of a better understanding of the intestinal homeostasis that is challenged during intestinal tumori- genesis. In the past, we have identified a novel regulatory element for colonic repression of the sucrase–isomal- tase gene (CRESIP) [9]. The CCAAT-Displacement- Protein (CUX1) interacts with CRESIP and represses SI gene transcription based on transfection and Cux1 deletion studies. These findings suggest that CUX1 could represent an important regulator of colonic epi- thelium homeostasis, but there is limited knowledge about the nature of the molecular pathways susceptible to direct influence by its transcriptional activity. The Cux1 gene encodes for a transcriptional repres- sor that belongs to the homeodomain protein family and contains four evolutionarily conserved DNA-bind- ing domains, three cut repeats (CR1, CR2 and CR3) and a cut homeodomain (HD) [10]. CUX1 contains two repressive regions (R1 and R2) at the carboxy ter- minus that account for its repressive activity [11]. Sev- eral Cux1 isoforms are produced from the proteolytic processing of p200Cux1 (e.g. p110Cux1) as well as alternative splicing (e.g. p75Cux1) [12]. Many in vitro gene targets of CUX1 are implicated in the control of the cell cycle [10,13,14]. Genetic deletion of Cux1 in mice has confirmed its crucial role during development. Cux1 hypomorphic mice are smaller than wild-type mice [15,16] and display organ-specific phenotypes, including growth retardation, delayed differentiation of lung epithelia, curly whiskers and altered hair follicle morphogenesis [17]. These mutant mice also show a deficiency in the hematopoiesis system with decreases in lymphocyte T- and B-cell populations, as well as an excessive production of myeloid cells in the liver, bone marrow and peripheral blood, a feature that is com- monly observed during the progression of human myeloid leukemia [16]. In the search for novel transcriptional targets of CUX1, we investigated the colonic changes in the gene repertoire of Cux1 mutant mice by gene expression analysis. This approach identified the Promyelocytic Leukemia Zinc Finger (PLZF) gene as a putative gene target of CUX1. PLZF was originally identified as a t(11;17) reciprocal chromosomal translocation with the retinoic acid receptor a (RARa) gene in individuals with acute promyelocytic leukemia [18]. PLZF is also expressed in immature hematopoietic cells and is downregulated after differentiation [19]. PLZF is a transcription factor composed of a repression domain BTB ⁄ POZ (bric-a ` -brac, tramtrack, brad complex ⁄ pox- virus zinc finger) and nine Kru ¨ ppel zinc fingers [20]. This repression domain allows PLZF to recruit core- pressors such as SMRT (silencing mediator of retinoid and thyroid hormone receptor), N-CoR (nuclear recep- tor corepressor), mSin3a (murine Swi-independent 3a) and HDAC (histone deacetylase) to exert its transcrip- tional action [21]. Recent genetic experimental approaches have shown a function for Plzf in the maintenance of spermatogonial stem cells [22,23], as well as axial skeletal patterning with alterations in the expression of Hox genes [24]. Other genetic evidence has identified the Caenorhabditis elegans PLZF ortho- log eor-1 as a positive regulator of the Ras and Wnt pathways [25]. Interestingly, the deregulation of these pathways is a common step in the molecular cascade that promotes CRC [26]. In this article, we provide evidence that PLZF is a direct target gene of CUX1. Expression analyses show a reciprocal expression pattern of CUX1 and PLZF among CRC cell lines and normal isolated colono- cytes. Because of the well-documented role of PLZF in acute promyelocytic leukemia [27], as well as its impli- cation in the regulation of several pathways well docu- mented to be upregulated in CRC, we suggest that the modulation of PLZF via CUX1 action may be func- tionally relevant to cancer pathology. Results and Discussion CUX1 physically interacts with the PLZF gene both in vitro and in vivo CUX1 is a crucial molecular contributor to SI gene transcriptional repression in colonocytes during murine post-natal development [9]. Although SI gene re-expres- sion is commonly observed in human colonic polyposis [4,8,28], the transcriptional effect of CUX1 on cancer- related genes has not been explored in the intestinal context. We thus decided to perform a gene expression analysis with colon isolated from newborn Cux1 mutant and control mice in order to identify novel CUX1 gene targets. A nonstatistical analysis of gene transcript vari- ations among pooled mutant versus pooled control CUX1 represses PLZF transcriptional activity I. Fre ´ chette et al. 4242 FEBS Journal 277 (2010) 4241–4253 ª 2010 The Authors Journal compilation ª 2010 FEBS samples predicted that the Plzf gene would be one of the most induced genes in the Cux1 mutant pooled sam- ples (Table S1). A quantitative RT-PCR analysis revealed that Plzf mRNA expression was induced in some, but not all, Cux1 mutant versus control colonic samples. However, this observation was not significant overall (data not illustrated). The variability of Plzf induction of expression among different Cux1 mutant individuals could be explained by putative compensa- tory mechanisms, as well as the possibility that the Cux1 mutation results in a hypomorphic allele because of the production of an N-terminal, 160-kDa truncated form of Cux1 in mutant mice, as already observed for other CUX1 gene targets [9,16]. To further determine whether this prediction could be of importance in human genetic regulation, we promptly investigated the presence of potential interac- tion sites for the transcription factor CUX1 in the PLZF gene. The human sequence of the PLZF gene has been characterized previously and the transcrip- tional start position has been well identified [29]. A computer analysis of the human PLZF gene using matinspector matrix software [30] predicted 15 poten- tial binding sites for the transcription factor CUX1 within 6.6 kb of the 5¢-UTR and flanking regions of the gene (Fig. 1). Noteworthy, the 5¢-UTR genomic region contained a high-density cluster of 12 predicted binding sites for CUX1 (Fig. 1). We then evaluated the capacity of CUX1 to interact with different potential binding sites of the genomic 5¢-UTR and promoter regions of the PLZF gene. Electrophoretic mobility shift assay (EMSA) was performed using double- stranded 32 P-labelled probes that corresponded to the 15 predicted CUX1 interacting sites included within the 5¢-UTR and the promoter region of PLZF. Nuclear extracts isolated from HEK293T cells, trans- fected or not with an expression vector for CUX1, were used for the assays. Overexpression of CUX1 protein in HEK293T cells led to the formation of sev- eral retarded complexes that harbored similar patterns with the binding sites 1, 2, 4, 5, 6, 7, 10, 12, 13 and 15, as opposed to HEK293T crude nuclear extracts (Fig. 2A). The pattern of retarded complexes between HEK293T crude and CUX1-enriched nuclear extracts was similar for sites 8, 9, 11 and 14, suggesting that these sites could be of better affinity for CUX1 recog- nition in vitro (Fig. 2A). The binding pattern for site 3 was somewhat different from that of the other sites, suggesting that the presence of other putative elements could be interfering with CUX1 interaction with this region (Fig. 2A). In order to better characterize the nature of the different patterns for the CUX1-retarded complexes, some sites representative for these different patterns were analyzed further. The introduction of mutations in the CUX1 consensus binding site resulted in a loss of CUX1-formed complexes, as observed for sites 8, 10, 12, 13, 14 and 15 and, to a lesser extent, for site 3 (Fig. 2B). The addition of an excess amount of nonlabelled oligonucleotide for these specific sites competed for the formation of CUX1-related complex, whereas corresponding mutated oligonucleotides vali- dated for the absence of CUX1 interaction did not compete, as illustrated for sites 8, 10, 12, 13, 14 and 15 in Fig. 2C. The addition of affinity-purified polyclonal CUX1 antibodies to the binding reaction for each of these positive interacting sites resulted in a supershift Fig. 1. Identification of 15 putative CUX1 binding sites in 6.6 kb of the PLZF gene. (A) Schematic representation of the PLZF gene with its predicted binding sites for the tran- scription factor CUX1. (B) Nucleotide sequence of the 15 potential binding sites for CUX1 in the PLZF gene. Base pairs in capital letters denote the CUX1 core sequence and italics show high conservation content [30]. I. Fre ´ chette et al. CUX1 represses PLZF transcriptional activity FEBS Journal 277 (2010) 4241–4253 ª 2010 The Authors Journal compilation ª 2010 FEBS 4243 of the four major specific CUX1-containing complexes relative to IgG control lanes, as observed for sites 5, 7, 8, 10 and 14 (Fig. 2D). These observations suggested that these sites can be recognized by multiple CUX1 cleaved forms specifically produced from the peG- FP ⁄ CUX1 expression vector, as observed in transfect- ed HEK293T cells (Fig. 2A, right panel). However, the exact contribution of these cleaved forms could not be distinguished in these conditions. To further verify whether the in vitro interaction of CUX1 with the PLZF gene could be reflected in the cel- lular context in vivo, chromatin immunoprecipitation A B CD Fig. 2. The eGFP ⁄ CUX1 fusion protein interacts with different 5¢-UTR and promoter binding sites of the PLZF gene in vitro. (A) Nuclear pro- tein extracts from transfected HEK293T cells (empty vector, lanes 2; eGFP ⁄ CUX1, lanes 3) were used for EMSA with labelled oligonucleo- tides. A condition with no protein included within the binding reaction was also tested for each probe (lanes 1). The retarded complexes are indicated by arrows. The right panel is a Western blot analysis representing the different isoforms produced from p200CUX1 in HEK293T cells transfected with the peGFP-CUX1 expression vector. (B) EMSA was carried out as described in (A), except that the consensus site for CUX1 for each oligonucleotide was mutated. (C) Competition of site 3, 8, 10, 12, 13, 14 and 15 complexes were obtained with the addition of a 10-fold molar excess of nonlabelled wild-type sites or CUX1 mutated site oligonucleotides. (D) Supershift experiments of the interacting sites 3, 5, 7, 8, 10 and 14 were performed by including a specific CUX1 antibody or rabbit IgG in the binding reactions. CUX1 represses PLZF transcriptional activity I. Fre ´ chette et al. 4244 FEBS Journal 277 (2010) 4241–4253 ª 2010 The Authors Journal compilation ª 2010 FEBS (ChIP) experiments were performed. The human Caco- 2 ⁄ 15 cell line was used as the p200CUX1 protein was shown to be expressed in these cultured cells [9]. Sub- confluent Caco-2 ⁄ 15 cell chromatin was cross-linked and immunoprecipitated with the CUX1 antibody. PCR amplifications of different PLZF regions were per- formed on the immunoprecipitated purified chromatin (Fig. 3B). CUX1 consistently interacted with a specific region of the genomic 5¢-UTR that contained the in vitro validated CUX1 binding sites 5–7 (Chip2, Fig. 3A, B) and a specific region of the promoter that contained the validated binding site 15 (Chip3, Fig. 3A, B). Because of the proximity of sites 13, 14 and 15, it remains probable that these additional sites also contrib- ute to the enrichment of the promoter region as revealed by this assay. As a negative control, a portion of the PLZF gene for which CUX1 was not predicted to inter- act (Chip1) was not immunoprecipitated with the CUX1 antibody as determined by PCR (Fig. 3B, C). Overall, these data confirmed that CUX1 can interact strongly with the genomic 5¢-UTR and promoter regions of the PLZF gene both in vitro and in vivo. CUX1 represses the transcriptional activity of the genomic 5¢-UTR and promoter regions of the human PLZF gene Because the genomic 5¢-UTR may be of importance in gene transcriptional regulation [31–34], we decided to test whether this region was functionally responsive to CUX1 transcriptional action. A 3-kb 5¢-UTR that con- tained eight interacting CUX1 binding sites was PCR amplified and subcloned into a luciferase reporter vec- tor. To measure the transcriptional effect of CUX1 on the 5¢-UTR genomic region, murine (CMV-Cux1) and human (eGFP ⁄ CUX1) expression vectors were cotransfected with the pGL3basic ⁄ PLZF 5¢-UTR reporter construct in HEK293T cells. The addition of CUX1 expression vectors to the cotransfection assay resulted in a two-fold reduction in 5¢-UTR ⁄ luciferase activity (Fig. 4A). A western analysis confirmed that both the murine Cux1 and human CUX1 proteins were synthesized when cotransfected in HEK293T cells, with the production of multiple isoforms of the proteins, as reported previously (Fig. 4B) [35]. The murine Cux1 protein was shorter than the human CUX1 protein because of the use of an N-terminally truncated mouse Cux1 construct [36]. The effect of CUX1 was next verified on a 776-bp portion of the PLZF gene promoter that contained the initiation start site for transcription (Fig. 4C). The addition of increasing amounts of CUX1 expression vector to the cotransfection assay resulted in a two-fold maximal reduction of the PLZF promoter activity in HEK293T cells (Fig. 4D). Individual abrogation of CUX1 inter- acting sites 13 and 15 of the PLZF promoter did not influence significantly the repressive activity of CUX1, in contrast with site 14, for which deletion resulted in a reduction of CUX1-dependent repression of the PLZF promoter (Fig. 4D, E). Simultaneous mutations of CUX1 interacting sites 13, 14 and 15 completely abolished the repressive effect of CUX1 on the PLZF promoter (Fig. 4D, E). Taken together, these data demonstrate that CUX1 has the ability to occupy and repress PLZF gene transcriptional activity. The loca- tion of several CUX1 elements within the 5¢-UTR was surprising because this region typically contains post- transcriptional elements controlling RNA stability or translation. Only a few transcriptional response ele- ments have been identified in 5¢-UTR genomic regions [31–34]. Interestingly, a recent report identified a single RHOX5 repressor element within the 5¢-UTR region of the netrin-1 receptor gene Unc5c that accounted for repression of the gene within Sertoli cells [34]. As we identified multiple CUX1 elements within the 5¢-UTR of the PLZF gene, we speculate that CUX1 could interfere with PLZF gene transcription by forming a physical barrier in the PLZF genomic region that cor- responds to the 5¢-UTR, causing a limitation of PLZF transcriptional initiation and gene processing via the transcriptional basal machinery. PLZF exhibits a reciprocal expression pattern to CUX1 in human colonocytes and CRC cell lines The expression of CUX1 is altered in breast cancer [37] and uterine leiomyoma [38,39]. The profile of CUX1 expression in colon cancer has not been reported. We thus aimed to compare the level of CUX1 protein expression among human CRC cell lines in comparison with other cell lines of cancer ori- gin. p200CUX1 was weaker in CRC cell lines (Colo205, T84, Caco-2 ⁄ 15 and DLD-1 cells) when compared with colonocytes isolated from human fetuses that were used as normal controls (Fig. 5A). This reduction was not observed in the liver cancer cell line HepG2, the pancreatic cancer cell line MIA PaCa- 2 and the transformed kidney cell line HEK293T (Fig. 5A). This indicates that the p200CUX1 protein reduction observed in the CRC context is most likely not a general phenomenon of cancer. We then asked whether the profile of PLZF expression was interre- lated with CUX1 expression in these cell lines. PLZF protein was detected in every CRC cell line, but was not detectable in isolated normal colonocytes (Fig. 5A), a pattern that was confirmed at the gene I. Fre ´ chette et al. CUX1 represses PLZF transcriptional activity FEBS Journal 277 (2010) 4241–4253 ª 2010 The Authors Journal compilation ª 2010 FEBS 4245 transcript level (Fig. 5B). However, we cannot exclude the possibility that a minor subpopulation of epithelial crypt cells, for example stem cells, could support PLZF expression. Although PLZF expression was weakly detected in HepG2 and MIA PaCa-2 cancer cell lines that harbored high levels of p200CUX1 A B C Fig. 3. CUX1 interacts with the PLZF gene in vivo. (A) Consensus CUX1 binding sites of the PLZF sequence are indicated in brackets. Bases in capital letters denote the CUX1 core sequence and italics show high conservation content in comparison with the consensus sequence. Primer sets used for the ChIP experiments are indicated by bold arrows (Chip2 ⁄ 3up and Chip2 ⁄ 3dw). For clarity, primer sets Chip1up and Chip1dw are not represented in the figure. Primer Chip1up started at position +3165 bp on the PLZF sequence and primer Chip1dw at posi- tion +3304 bp. (B) ChIP assays were performed with subconfluent Caco-2 ⁄ 15 cell chromatin. Chromatin was immunoprecipitated with the CUX1 antibody or with rabbit IgG as a negative control. Purified immunoprecipitated chromatin was subjected to PCR amplification of three independent regions of the PLZF gene (1–3); 10% of the chromatin extract (Input) was also amplified by PCR to determine the amount of DNA prior to immunoprecipitation. A representative result of three to four independent experiments is illustrated. (C) Chromatin immunopre- cipitated in (B) was analysed by quantitative real-time PCR with oligonucleotides amplifying three independent regions of the PLZF gene (1–3). Results are expressed as the fold increase over normal rabbit IgG normalized to input. CUX1 represses PLZF transcriptional activity I. Fre ´ chette et al. 4246 FEBS Journal 277 (2010) 4241–4253 ª 2010 The Authors Journal compilation ª 2010 FEBS AB C E D Fig. 4. CUX1 functionally influences the PLZF 5¢-UTR and promoter genomic regions in transient reporter assays. (A) HEK293T cells were transiently cotransfected with 0 or 0.4 lg of the CMV-Cux1 (murine) or peGFP-CUX1 (human) expression vector and 0.2 lg of the pGL3ba- sic ⁄ PLZF 5¢-UTR luciferase reporter construct. The pcDNA3.1 plasmid was used as an empty control vector to calibrate for the addition of the expression vectors. Cells were harvested after transfection and analyzed for luciferase activity. Data were normalized with Renilla values. Results obtained in triplicate were reported as a percentage of the controls (means ± SD) and are representative of three independent experiments. (B) A western blot with CUX1 and actin antibodies was performed on pooled lysates used for the luciferase detection in (A). The molecular masses for each Cux1 and CUX1 protein isoform are indicated. (C) Schematic representation of the wild-type, single, double and triple mutants of the 776-bp region of the PLZF promoter. The bold crosses represent the site(s) mutated in each construct. (D) HEK293T cells were transfected with increasing amounts of the peGFP-CUX1 (0, 5 and 12 ng) expression vector and 0.2 lg of the different pGL3basic ⁄ PLZF 776-bp promoter luciferase constructs illustrated in (C). The pcDNA3.1 plasmid was used as an empty control vector to calibrate for the addition of the expression vector. Cells were harvested after transfection and analyzed for luciferase activity. Data were normalized with Renilla values. Results obtained in triplicate were reported as a percentage of the controls (means ± SD) and are representa- tive of three independent experiments. (E) The percentage of repression activity of CUX1 on both wild-type and mutant pGL3basic ⁄ PLZF 776-bp promoter luciferase constructs was monitored as described in (A). ***P < 0.001. I. Fre ´ chette et al. CUX1 represses PLZF transcriptional activity FEBS Journal 277 (2010) 4241–4253 ª 2010 The Authors Journal compilation ª 2010 FEBS 4247 expression, this correlation was not maintained in the HEK293T cell line (Fig. 5A). Thus, PLZF-positive expression correlated well with the relatively low level of CUX1 repressor detection in colonic cancer cells, but other regulatory mechanisms are likely to be involved in the context of transformed human cells. We then tested whether PLZF expression could be responsive to CUX1 variations in colon cancer cells. The Caco-2 ⁄ 15 cell line was used to generate stable populations of cells expressing a short hairpin RNA (shRNA) against CUX1. The efficiency of CUX1 mRNA reduction was more than 70% (P < 0.0001), as assessed by quantitative RT-PCR (Fig. 5C), and correlated with an efficient loss of CUX1 pro- tein abundance, as determined by western blotting (Fig. 5D). Coincidently, PLZF mRNA expression was induced more than 2.8-fold in shCUX1 Caco-2 ⁄ 15 cell populations as opposed to shCUX1 mutated controls (P < 0.05) (Fig. 5E). CUX1 and PLZF as a novel molecular link in human diseases Human CUX1 is located on chromosome 7q22, a region that is often deleted in 7q-related acute myeloid leukemia and myeloid dysplasia [40–44]. These studies suggest that CUX1 is a tumor suppressor and that its loss is a significant event in the generation or progres- sion (or both) of myeloid disorders. The demonstration that PLZF is a direct target for CUX1 repressive action represents an interesting avenue for the specific investigation of this molecular interaction during leu- kemia progression. However, the PLZF ortholog was demonstrated to promote the expression of Ras and AB C E D Fig. 5. CUX1 and PLZF expression in human colon cancer cell lines. (A) Western blot analysis was performed with CUX1, PLZF and actin antibodies on total protein extracts from various human cancer cell lines [colorectal cancer cell lines, liver cancer cell line (HepG2), pancreas cancer cell line (MIA PaCa-2) and the transformed kidney cell line (HEK293T)] and from epithelial cells isolated from human fetal colon. (B) Total RNA was isolated from Colo205, Caco-2 ⁄ 15, DLD-1 and normal colonocytes derived from human fetus and subjected to quantitative real-time PCR. PLZF expression was quantified and normalized using Tata-box binding protein (TBP) reference gene. (C) Total RNA was iso- lated from stable Caco-2 ⁄ 15 cell populations that contained an integrated shRNA against the CUX1 mRNA or a mutated shRNA as a control. Real-time PCR was performed and CUX1 expression was normalized using human b2-microglobulin (b2-mic) reference gene. (D) Western blot experiment using CUX1 and actin antibodies was performed to monitor CUX1 expression in the Caco-2 ⁄ 15 cell populations described in (C). (E) Total RNA isolated as described in (C) was used to monitor PLZF mRNA expression. *P < 0.05. CUX1 represses PLZF transcriptional activity I. Fre ´ chette et al. 4248 FEBS Journal 277 (2010) 4241–4253 ª 2010 The Authors Journal compilation ª 2010 FEBS Wnt responsive genes in Caenorhabditis elegans [25]. Sustained activation of Wnt and Ras signaling is very frequent in human CRC [26,45] and results in intesti- nal polyposis and adenocarcinoma in transgenic mice [46,47]. Taken together, our findings demonstrate that CUX1 is a negative regulator of the PLZF gene. Future investigations into the molecular status of CUX1 in human CRC, as well as the comprehensive role of PLZF in intestinal homeostasis, represent future challenges that will enable us to expand our knowledge about the mechanisms related to intestinal disease. Experimental procedures Animals The generation of Cux1 mutant mice has been described elsewhere [16]. C57BL ⁄ 6J mice heterozygous for the tar- geted allele were subsequently bred with normal CD1 mice. Homozygous Cux1 mice were identified by PCR, and their identity was confirmed on the basis of their typically small size and curly hair. Mice were treated in accordance with a protocol approved by the Institutional Animal Research Review Committee. Isolation of epithelial cells from mouse and human intestine Intestinal epithelial cells were isolated from the intestinal epithelium of mouse youngsters and human fetuses as described previously [48]. Human intestine from fetuses ranging from 17 to 20 weeks of age (post-fertilization) were obtained after legal abortion. The project was in accor- dance with a protocol approved by the Institutional Human Research Review Committee for the use of human material. Briefly, the intestine was separated in sections and the colon was opened longitudinally and rinsed with cold NaCl ⁄ P i . The colon sections were further cut into 5-mm pieces and incubated in 5 mL of cold MatriSperse (Becton-Dickinson Canada, Oakville, ON, Canada) in 15-mL tubes at 4 °C for 18–24 h. The epithelial layer was dissociated by gentle manual shaking. The epithelial sus- pension was collected, centrifuged, washed with cold NaCl ⁄ P i and processed for RNA and protein isolation as described previously [48]. Cell culture and shRNA knockdown The HepG2 human liver carcinoma cell line, MIA PaCa-2 human pancreatic carcinoma cell line and the human ade- nocarcinoma colorectal cell lines DLD1, T84 and Colo205 were all obtained from the American Type Culture Collec- tion (ATCC, Rockville, MD, USA). The Caco-2 ⁄ 15 cell line [49] was kindly provided by Dr J. F. Beaulieu. DLD1 cells were cultured in RPMI1640 medium and T84 cells in Ham’s:DMEM. Caco-2 ⁄ 15, Colo205, HepG2, MIA PaCa-2 and HEK293T cells were cultured in DMEM. All culture media were supplemented with 10% fetal bovine serum (ICN Biomedicals, Aurora, OH, USA), 2 mm glutamine (Gibco, Burlington, ON, USA), 0.01 m Hepes (Gibco) and 100 lgÆmL )1 penicillin ⁄ streptomycin (Gibco). Fetal bovine serum was heat inactivated as recommended for DLD1 cell culture (ATCC). The cell lines were maintained at 37 °Cin a humidified atmosphere with 5% CO 2 . An shRNA cloned in pSuper.retro.puro (Oligoengine, Seattle, WA, USA) that targeted a conserved sequence in rat, mouse and human CUX1 was kindly provided by Dr Julian Downward [50]. Two bases (in capitals) were further mutated (5¢-aagaaga acaGAccagaggattt-3¢) to be used as a control. The Caco- 2 ⁄ 15 cell line was infected with the retroviral shRNA con- structs and selected with 5 lgÆmL )1 of puromycine for 1 week (Sigma Chemical Co., St Louis, MO, USA). Microarray analysis Total RNA was isolated from the colon of newborn Cux1 hypomorphic mutant and control mice as described previ- ously [9]. Briefly, 5 lg of total RNA from three indepen- dent individuals of both normal and mutant mice were pooled and submitted to the Penn Microarray Facility (University of Pennsylvania, Philadelphia, PA, USA) for target preparation and hybridization to murine MG_U74Av2 GeneChips (Affymetrix, Santa Clara, CA, USA), followed by microarray analysis as described elsewhere [51]. Micro- array Analysis Suite 5.0 (MAS, Affymetrix) was used to quantify microarray signals [51]. EMSA EMSAs were performed essentially as described previously [9,52] with some modifications. The reactions were per- formed in a volume of 24 lL of binding buffer D (10 mm Hepes, pH 7.9, 10% glycerol, 0.1 mm EDTA and 0.25 mm phenylmethanesulfonyl fluoride) containing 5 lg of nuclear protein extracts from HEK293T cells, transfected or not with the peGFP-CUX1 expression vector, 50 mm KCl, 50 ng of poly(dI–dC) and 25 000 cpm of 32 P-labeled DNA probes for 10 min. For the supershift analysis, 200 ng of CUX1 M-222X antibody or rabbit IgG (Santa Cruz Biotechnology, Santa Cruz, CA, USA) were added and the binding reactions were pursued for 10 min at room temperature. Competition assays were performed with the addition of an excess (10-fold) of nonlabelled double-stranded oligonucleotides. Retarded complexes were then separated on a 5% polyacryl- amide gel at 4 °C for 4 h, dried for 1 h at 80 °C and exposed overnight on a Molecular Imager FX screen (Biorad, Missis- sauga, ON, Canada). The running buffer used was a Tris– glycine 0.5· buffer (0.2 m glycine, 0.025 m Tris and 1 mm I. Fre ´ chette et al. CUX1 represses PLZF transcriptional activity FEBS Journal 277 (2010) 4241–4253 ª 2010 The Authors Journal compilation ª 2010 FEBS 4249 EDTA). The DNA probes consisted of double-stranded oli- gonucleotides of 15 potential CUX1 binding sites within the 5¢-UTR and 776-bp promoter regions of the human PLZF gene (Fig. 1) (matinspector software tool; http://www.gen- omatix.de) [30]. Mutated oligonucleotides for CUX1 inter- acting sites were designed as follows (upper strand, mutation in bold): site 3, 5¢-gctgtaggggactcgatttaactcgagtctctctcca-3¢; site 8, 5¢-gcgccaagcttctcgtcttttggagcttccctccct-3¢; site 10, 5¢-tcg cttcgacatcactgccccgcggacct-3¢; site 12, 5¢-tgggccctcgagtccttat- caaac-3¢; site 13, 5¢-ggccagcttcgctattcctctgtc-3¢; site 14, 5¢-ac- cctcctgttgtccttcgtgagctctgaaag-3¢; site 15, 5¢-tctgatgttttcga gtcctacagt-3¢. Genomic DNA amplification, reporter plasmid constructs and mutagenesis The 3-kb 5¢-UTR region of the human PLZF gene was amplified by PCR from purified genomic DNA isolated from the human normal intestinal epithelial cell line HIEC [53] and Herculase DNA polymerase (Stratagene Cloning Sys- tems, La Jolla, CA, USA). The 776-bp promoter region was amplified with iProof DNA polymerase (Biorad) from the vector pGL3basic already containing a 2-kb region of the human PLZF gene. The primer sequences used for the 3-kb 5¢-UTR were 3¢HPLZF (5¢-gaggggaagaagcaaaagaga-3¢) and (+)2878 HPLZF (5¢-gatccggaggctttgtacc-3¢). The primer sequences for the 776-bp region were pPLZF742SmaIA (5¢-tcactacccgggaagcccttgcttccttcatc-3¢) and pPLZF742SmaIB (5¢-agtgatcccgggagataaagcagcagcagctg-3¢). The 3-kb 5¢-UTR ⁄ PLZF amplified fragment was subcloned into the pBluescript KS(–) vector (EcoRV). The integrity of the subcloned PCR products was confirmed by sequence analysis. The 3-kb 5¢-UTR ⁄ PLZF fragment was then released from pBluescript KS(–) with BamHI and KpnI (Roche Diagnostics, Laval, QC, Canada) and subcloned into the KpnIandBglII restriction sites of the pGL3basic luciferase reporter plasmid. The 776-bp PLZF promoter fragment was released with SmaI (Roche Diagnostics) restriction enzyme and subcloned into the SmaI restriction site of pGL3basic. Mutagenesis of the CUX1 sites 13, 14 and 15 of the 776-bp PLZF promoter was performed by overlap extension with mutated oligonucleo- tides. The integrity of the mutated promoter was confirmed in each case by sequence analysis. Transient transfections and luciferase assays HEK293T cells were seeded in 24-well plates for 24 h. Transfections were performed with Lipofectamine 2000 (Invitrogen, Burlington, ON, Canada) according to the manufacturer’s recommendations. Cells at 90% confluence were cotransfected with 200 ng of luciferase reporter con- struct (pGL3basic ⁄ PLZF 3-kb 5¢-UTR, pGL3basic ⁄ PLZF 776-bp or pGL3basic ⁄ PLZF 776-bp mut13, pGL3basic ⁄ PLZF 776-bp mut14, pGL3basic ⁄ PLZF 776-bp mut15, pGL3basic ⁄ PLZF 776-bp mut13,15, pGL3basic ⁄ PLZF 776-bp mut13,14,15) and a fixed amount of either CMV- Cux1 (murine) or peGFP-CUX1 (human) expression vec- tors with a constant total DNA amount of 800 ng per transfection in OptiMEM medium (Invitrogen); 5 ng of the pRL SV40 Renilla luciferase vector (Promega, Madison, WI, USA) was also included in each reaction as a control for transfection efficiency. After 4 h, transfection medium was replaced by DMEM containing 10% fetal bovine serum. Luciferase and Renilla activities were determined with the dual luciferase assay kit (Promega). Each experi- ment was repeated at least three times in triplicate. ChIP assays ChIP assays were performed using the ChIP assay kit, according to the manufacturer’s instructions (Upstate, Mil- lipore, Billerica, MA, USA). Subconfluent (75%) Caco- 2 ⁄ 15 cells were cross-linked with 1% formaldehyde for 10 min at 37 °C and sonicated to obtain a DNA average size of 500 bp in length. Chromatin was immunoprecipitat- ed with a rabbit polyclonal antibody against CUX1 (Santa Cruz Biotechnology). Rabbit IgG (Santa Cruz Biotechnol- ogy) was also used as a negative control. Ten percent of the lysate was kept to verify the amount of DNA used for each immunoprecipitation. Immunoprecipitated DNA was purified with a phenol–chloroform extraction and resus- pended in 20 lL of ultrapure water before PCR amplifica- tion with Chip1up primer (5¢-aagctccagagggtctgcac-3¢) and Chip1dw primer (5¢-gaaaggcatcccgaacgcat-3¢); Chip2up primer (5¢-aaatgtcttgaccagccgtc-3¢) and Chip2dw primer (5¢-gaaacaaaggcctctcccag-3¢); Chip3up primer (5¢-gctttgcagt- cagaatggtc-3¢) and Chip3dw primer (5¢-ctgagcactgactac- gaaac-3¢). The Chip1up and Chip1dw oligonucleotides amplified a 139-bp PLZF gene region that contained no predicted CUX1 binding site; the Chip2up and Chip2dw oligonucleotides amplified a 289-bp PLZF gene region containing CUX1 binding sites 5, 6 and 7 (Chip2, Fig. 3A); the Chip3up and Chip3dw oligonucleotides amplified a 207-bp PLZF gene region containing the CUX1 binding site 15 (Chip3, Fig. 3A). The program used for amplification was a first cycle of 95 °C for 2 min (Hot- Start), followed by 10 cycles of 95 °C for 30 s, 58.7 °C for 30 s and 72 °C for 2 min, followed by 25 cycles of 95 °C for 30 s, 58.7 °C for 30 s and 72 °C for 2 min, with an increasing elongation time of 10 s every cycle, and a final incubation at 72 °C for 7 min. Amplified PCR products were separated on a 1.5% agarose gel and visualized by ethidium bromide staining. The immunoprecipitated DNA was also amplified by real-time PCR using the QuantiTect SYBR Green PCR kit (Qiagen, Valencia, CA, USA) with the same primers as described above. Data are expressed as the fold increase over background (negative control, IgG) normalized to input, as proposed by SuperArray Bio- sciences and adapted as described previously (http:// www.workingthebench.com). CUX1 represses PLZF transcriptional activity I. Fre ´ chette et al. 4250 FEBS Journal 277 (2010) 4241–4253 ª 2010 The Authors Journal compilation ª 2010 FEBS [...]... 5¢-tcgcgctactctctctttctg-3¢; hB2MIC227dw, 5¢-tcaatgtcggatg gatgaaa-3¢ Acknowledgements We thank Dr Peter G Traber for the generous gift of the microarray data analysis that was originally performed at the Penn Microarray Facility (University of Pennsylvania, Philadelphia, PA, USA) We also thank Drs Angus M Sinclair and Richard H Scheuermann (UT Southwestern Graduate School of Biomedical Sciences, Dallas,... Sucrase–isomaltase: a marker of foetal and malignant epithelial cells of the human colon Int J Cancer 32, 407–412 4 Beaulieu JF, Weiser MM, Herrera L & Quaroni A (199 0) Detection and characterization of sucrase–isomaltase in adult human colon and in colonic polyps Gastroenterology 98, 1467–1477 5 Yoshida K, Nakamura W, Hirano K, Yuasa H, Tsukamoto T & Tatematsu M (199 8) Expression of sucrase and intestinal-type... poly-oligo(dT) ´ (Amersham Biosciences, Baie d’Urfe, QC, Canada) and 40 units of reverse transcriptase (Roche Diagnostics, Laval, QC, Canada) Quantitative PCR was performed on a LightCycler PCR apparatus V2.0 (Roche Diagnostics) and relative mRNA expression was measured using lightcycler software 4.0 according to the manufacturer’s protocol (Roche Diagnostics) Tata-box binding protein (TBP) and human b2-microglobulin... Mistry AR, Pedersen EW, Solomon E & Grimwade D (200 3) The molecular pathogenesis of acute promyelocytic leukaemia: implications for the clinical management of the disease Blood Rev 17, 71–97 28 Wiltz O, O’Hara CJ, Steele GD & Mercurio AM (199 0) Sucrase–isomaltase: a marker associated with the progression of adenomatous polyps to adenocarcinomas Surgery 108, 269–275 29 Zhang T, Xiong H, Kan LX, Zhang CK,... Beaulieu JF & Quaroni A (199 1) Clonal analysis of sucrase–isomaltase expression in the human colon adenocarcinoma Caco-2 cells Biochem J 280 (Pt 3, 599–608 50 Michl P, Ramjaun AR, Pardo OE, Warne PH, Wagner M, Poulsom R, D’Arrigo C, Ryder K, Menke A, Gress T et al (200 5) CUTL1 is a target of TGF(beta) signaling that enhances cancer cell motility and invasiveness Cancer Cell 7, 521–532 51 Zeng F, Baldwin... USA) for the gift of the eGFP ⁄ CUX1 retroviral construct, Dr Jean-Francois Beaulieu ¸ CUX1 represses PLZF transcriptional activity ´ (Universite de Sherbrooke, QC, Canada) for providing the Caco-2 ⁄ 15 cell lines and Elizabeth Herring for critical reading of the manuscript This work was supported by a grant from the Canadian Institutes of Health Research (MOP-8977 0) FB is a senior scholar ´ ´ from the. .. receptor-alpha locus due to a variant t(11;1 7) translocation associated with acute promyelocytic leukaemia EMBO J 12, 1161–1167 Reid A, Gould A, Brand N, Cook M, Strutt P, Li J, Licht J, Waxman S, Krumlauf R & Zelent A (199 5) Leukemia translocation gene, PLZF, is expressed with a speckled nuclear pattern in early hematopoietic progenitors Blood 86, 4544–4552 Melnick A, Ahmad KF, Arai S, Polinger A, Ball... b2-microglobulin (hb2mic) mRNA expression were analysed as a reference Double-stranded DNA amplification during PCR was monitored using SYBR Green I (QuantiTect SYBR Green PCR kit; Qiagen) The primers hybridized at 59 °C were as follows: 5¢-hrmPLZF1923up, 5¢-agcacactcaagagccacaa-3¢; hrmPLZF2056dw, 5¢-tcaaag ggcttctcacctgt-3¢; hrmTBP1009up, 5¢-ggggagctgtgatgtgaagt-3¢; hrmTBP1139dw, 5¢-ggagaacaattctgggtttga-3¢; hB2MIC89up,... Gambardella L, Horcher M, Tschanz S, Capol J, Bertram P, Jochum W, Barrandon Y & Busslinger M (200 1) The transcriptional repressor CDP (Cutl 1) is essential for epithelial cell differentiation of the lung and the hair follicle Genes Dev 15, 2307–2319 Chen Z, Brand NJ, Chen A, Chen SJ, Tong JH, Wang ZY, Waxman S & Zelent A (199 3) Fusion between a novel Kruppel-like zinc finger gene and the retinoic acid... tumorigenesis in mice Gastroenterology 123, 492–504 48 Boudreau F, Rings EH, van Wering HM, Kim RK, Swain GP, Krasinski SD, Moffett J, Grand RJ, Suh ER & Traber PG (200 2) Hepatocyte nuclear factor-1 alpha, GATA-4, and caudal related homeodomain protein Cdx2 interact functionally to modulate intestinal gene transcription Implication for the developmental regulation of the sucrase–isomaltase gene J Biol . The Promyelocytic Leukemia Zinc Finger (PLZF ) gene is a novel transcriptional target of the CCAAT-Displacement- Protein (CUX 1) repressor Isabelle Fre ´ chette, Mathieu Darsigny, Karine. (5¢-gaaaggcatcccgaacgcat-3 ); Chip2up primer (5¢-aaatgtcttgaccagccgtc-3 ) and Chip2dw primer (5¢-gaaacaaaggcctctcccag-3 ); Chip3up primer (5¢-gctttgcagt- cagaatggtc-3 ) and Chip3dw primer (5¢-ctgagcactgactac- gaaac-3 ). . hB2MIC227dw, 5¢-tcaatgtcggatg gatgaaa-3¢. Acknowledgements We thank Dr Peter G. Traber for the generous gift of the microarray data analysis that was originally per- formed at the Penn Microarray Facility