The intestine-specific homeobox gene Cdx2 induces expression of the basic helix–loop–helix transcription factor Math1

The intestine-specific homeobox gene Cdx2 induces expression of the basic helix–loop–helix transcription factor Math1

r 2006, Copyright the Authors Differentiation (2006) 74:313–321 DOI: 10.1111/j.1432-0436.2006.00074.x Journal compilation r 2006, International Society...

3MB Sizes 2 Downloads 38 Views

r 2006, Copyright the Authors Differentiation (2006) 74:313–321 DOI: 10.1111/j.1432-0436.2006.00074.x Journal compilation r 2006, International Society of Differentiation

O RI G INA L AR T I C L E

Hiroyuki Mutoh . Hirotsugu Sakamoto . Hiroko Hayakawa . Yukitomo Arao . Kiichi Satoh . Mitsuhiro Nokubi . Kentaro Sugano

The intestine-specific homeobox gene Cdx2 induces expression of the basic helix–loop–helix transcription factor Math1

Received November 13, 2005; accepted in revised form February 26, 2006

Abstract The basic helix–loop–helix transcription factor Math1, which is transiently expressed in proliferating neural precursors in multiple domains of the developing nervous system, is also related to the cell fate decision of enteroendocrine, goblet, and Paneth cells in the intestine. On the other hand, the transcription factor Cdx2, which is normally confined to intestinal epithelial cells, is related to the differentiation of these cells. Therefore, we investigated the relationship between Math1 and Cdx2 in intestinal epithelial cells. The Math1 and Cdx2 expressions in normal intestinal mucosa and intestinal metaplastic mucosa from mouse and human stomachs, as well as an intestinal crypt-derived cell line, were analyzed by immunohistochemistry, reverse transcription-polymerase chain reaction and Northern blotting, and Math1 enhancer element was analyzed by luciferase reporter assays. Math1-positive epithelial cells co-expressing Cdx2 were found in normal intestinal mucosa from humans and mice. Furthermore, . )  Hirotsugu Sakamoto  Hiroko Hiroyuki Mutoh (* Hayakawa  Kiichi Satoh  Kentaro Sugano Department of Gastroenterology Jichi Medical School Yakushiji 3311-1, Minamikawachimachi, Kawachigun, Tochigi 329-0498 Japan Tel: 181 285 58 7348 Fax: 181 285 44 8297 E-mail: [email protected] Yukitomo Arao Department of Environmental Medicine Center for Community Medicine Tochigi, Japan Mitsuhiro Nokubi Department of Pathology Jichi Medical School Tochigi, Japan U.S. Copyright Clearance Center Code Statement:

Math1-producing epithelial cells that showed positive immunostaining for Cdx2 were also observed in intestinal metaplastic mucosa from human and Cdx2 transgenic mouse stomachs, although they were not detected in normal gastric mucosa of humans and mice. Expression of Cdx2 stimulated endogenous Math1 mRNA expression in the intestinal crypt-derived cell line IEC-6, corroborating observations in Cdx2-expressing intestinal metaplastic mucosa. Furthermore, expression of Cdx2 in IEC-6 cells conferred the ability to express a Math1 reporter gene containing a Math1 enhancer. Based on these results, we hypothesize that Cdx2 is involved in activating Math1 expression in intestinal epithelial cells. Key words

Cdx2  Math1  intestine

Introduction Cdx2, a caudal-related homeobox transcription factor, is selectively localized in the nuclei of fetal and adult mucosal epithelial cells in the small and large intestines in both humans and mice (Meyer and Gruss, 1993). Cdx2 haploinsufficiency results in abnormal differentiation of the midgut endoderm (Chawengsaksophak et al., 1997). Inactivation of Cdx2 by homologous recombination results in the development of multiple intestinal polyp-like lesions in pericecal areas of the midgut of heterozygous mice (Beck et al., 1999; Tamai et al., 1999; Beck et al., 2003). These polyps do not express Cdx2 and consist of hamartoma-containing areas that resemble the keratinizing stratified squamous epithelium found in the mouse forestomach, columnar mucous-secreting cells found in the gastric cardia and parietal cells found in the gastric corpus (Beck et al., 1999; Tamai et al., 1999; Beck et al., 2003). Thus, under

0301–4681/2006/7406–313 $ 15.00/0

314

conditions of Cdx2 deficiency, cells that would normally differentiate into the proximal colon follow a default pathway and form the characteristic epithelium of the stomach. Although normal gastric mucosa does not express Cdx2, strong nuclear immunoreactivity for Cdx2 is detected in human gastric intestinal metaplastic mucosa (Mizoshita et al., 2001). We previously showed that specific expression of Cdx2 in the stomach of Cdx2 transgenic mice induces intestinal metaplasia consisting of absorptive enterocytes, enteroendocrine cells, and goblet cells (Mutoh et al., 2002). The intestinal metaplasia of the Cdx2 transgenic mouse stomach and the polyps observed in the heterozygous Cdx2 knock-out mouse intestine indicate that Cdx2 plays a key role in intestinal epithelial cell fate determination and differentiation. The basic helix–loop–helix transcription factor Math1 is transiently expressed in proliferating neural precursors in multiple domains of the developing murine nervous system, including the dorsal hindbrain and neural tube, hair cells of the vestibular and auditory systems, mechanoreceptor (Merkel) cells of hairy skin, and cells of the external granule layer of the developing cerebellum (Akazawa et al., 1995; Ben-Arie et al., 1997; Bermingham et al., 1999). In addition to the nervous system, Math1 is also expressed in developing and mature mouse intestinal epithelium. Math1 expression within the gut is restricted to the intestinal epithelium starting at E16.5 and is sustained until adulthood (Yang et al., 2001). Math1 expression is not detected in the normal stomach (Yang et al., 2001). Loss of Math1 leads to depletion of enteroendocrine, goblet, and Paneth cells, indicating that it is important for intestinal epithelial cell fate decisions (Yang et al., 2001). These observations suggest that the transcription factors Cdx2 and Math1 are both closely related to intestinal epithelial differentiation. However, the relationship between these two transcription factors during intestinal epithelial differentiation has not yet been clarified. In the present study, we analyzed the relationship between the intestine-specific homeobox gene Cdx2 and the basic helix–loop–helix transcription factor Math1 in intestinal epithelial cells, and found that Cdx2 induces the expression of Math1.

Materials and methods Cdx2 transgenic mice Cdx2 transgenic mice with stomach-specific expression of Cdx2 under the control of the rat H1/K1-ATPase b-subunit gene promoter were used (Mutoh et al., 2002). The gastric mucosa of these mice is completely changed to intestinal metaplastic mucosa (Mutoh et al., 2002). Mice had free access to standard food and drinking water and were maintained on a 12 hr light/dark cycle. All experiments in this study were performed in accordance with the Jichi Medical School Guide for Laboratory Animals.

Human tissue samples We examined surgically resected intestinal metaplastic mucosa samples taken from around gastric adenocarcinomas in surgically resected small and large intestines obtained from the surgical pathology files for 2003 of the Department of Pathology, Jichi Medical School. All procedures were performed in accordance with local ethical guidelines. Surgically resected intestinal metaplastic mucosa as well as small and large intestinal mucosas were evaluated for their expressions of Math1 and Cdx2. Immunohistochemistry Human and murine tissues were fixed in 4% formaldehyde in phosphate-buffered saline (PBS) overnight at room temperature, embedded in paraffin and sectioned at a thickness of 3 mm. Sections were mounted on glass slides, deparaffinized in two changes of xylene for 10 min each and rehydrated in distilled water through a series of graded alcohols. For histological evaluation, sections were stained with hematoxylin and eosin (Mutoh et al., 2002). For immunohistochemical experiments, antigenicity was enhanced by boiling the sections in 10 mM citrate buffer (pH 6.0) in a microwave oven for 15 min, and the endogenous peroxidase activity was blocked by incubation in methanol containing 0.3% H2O2 for 30 min. After two washes with PBS containing 0.1% Triton X-100, the sections were preincubated with blocking buffer (DakoCytomation, Carpinteria, CA) in a humidified chamber for 15 min at room temperature, and then incubated in a primary antiserum diluted in PBS overnight at 41C. Next, the sections were washed in PBS and incubated with horseradish peroxidase (HRP)-labeled polymers conjugated to secondary antibodies for primary mouse or rabbit antibodies (DAKO EnVision System; DakoCytomation) without dilution at 371C for 30 min. Color development was carried out by incubating the sections with 3,3-diaminobenzidine tetrahydrochloride (DAB; Wako Pure Chemical Industries, Osaka, Japan) as the chromogenic substrate. Finally, the sections were lightly counterstained with hematoxylin, mounted and viewed under a light microscope. For double-immunostaining of Cdx2 and Math1, residual peroxidase activity after DAB color development for Cdx2 or Math1 was removed by incubation in methanol containing 0.3% H2O2 for 30 min. Thereafter, the sections were sequentially incubated with blocking buffer, an anti-Math1 or anti-Cdx2 antibody and HRP-labeled polymers conjugated to secondary antibodies for primary mouse or rabbit antibodies. Color development was carried out by incubating the slides with TrueBlue (KPL, Gaithersburg, MD) as the chromogenic substrate, and the sections were mounted without counterstaining. The primary antibodies used for the immunohistochemistry were: mouse monoclonal anti-Cdx2 (1:100; BioGenex, San Ramon, CA) (Mutoh et al., 2002, 2004a, 2004b, 2005; Qualtrough et al., 2002; Bonhomme et al., 2003; Hinoi et al., 2003; Osawa et al., 2004; Tot, 2004), rabbit polyclonal anti-Math1 (1:200; CeMines, Evergreen, CO) and rabbit polyclonal anti-chromogranin A (Nichirei, Tokyo, Japan). RNA isolation, reverse transcription-polymerase chain reaction (RT-PCR) and plasmid construction Total RNA was extracted from the stomach (normal mice), small intestine (normal mice), and intestinal metaplastic mucosa (Cdx2 transgenic mice) using the guanidinium isothiocyanate/phenol method (Isogen; Nippon Gene, Tokyo, Japan) according to the manufacturer’s instructions. Total RNA (1 mg) was reverse-transcribed at 371C for 1 hr in a final volume of 20 ml of reverse transcription buffer (50 mM TrisHCl pH 8.3, 50 mM KCl, 10 mM MgCl2, 0.5 mM spermidine and 10 mM DTT) containing reverse transcriptase (ReverTraAce; TOYOBO, Osaka, Japan), 16 U RNase inhibitors, 200 pmol random

315 Table 1 Oligonucleotides used as primers for polymerase chain reaction amplifications Name

Nucleotide sequence (5 0 –3 0 )

Primers for cDNA amplification (probes for northern blot analyses and expression vectors) Cdx2cDNAfwEcoRI GAATTCGCCACCATGTACGTGAGC Cdx2cDNArvKpnI GGTACCTCACTGGGTGACAGTGGAGT Math1cDNAfw ATGTCCCGCCTGCTGCATGC Math1cDNArv CTAACTGGCCTCATCAGAGT Primers for genomic DNA amplification (reporter gene assays) Math1 KpnI/1–20 GGTACCTCCAAGGTCCGGCAATGAAG Math1 BglII/688–669 AGATCTTTTGGGGCTCAGTTTTAAAT Math1 KpnI/688–707 GGTACCAGTTGTAATGTTATTGAAGT Math1 BglII/1,266–1,247 AGATCTTGTGGCCGCTCAGCTCCCCG Math1 KpnI/1,266–1,285 GGTACCATTTAACACCGTCGTCACCC Math1 KpnI/1,359–1,378 GGTACCAAGCCAGAGCCTCTCGCCGT Math1 BglII/1,517–1,488 AGATCTCCTCCCCTAGGCTTTGCTTG Math1 SacI/1,455–1,436 GAGCTCTGCCGGGAGGTGCAGTGGGC Math1 SacI/1,461–1,480 GAGCTCCAGCCCCGCCAGAAAAGGAG

primer, and 1.0 mM dNTPs (Sigma, St. Louis, MO). The reaction was terminated by incubating the mixture at 951C for 10 min. To compare Math1 expressions in the stomach (normal mice), small intestine (normal mice), and intestinal metaplastic mucosa (Cdx2 transgenic mice), PCR amplification was performed using the primers Math1cDNAfw and Math1cDNArv (Table 1) by incubation at 941C for 2 min, followed by 35 cycles of 941C for 30 s, 601C for 30 s, and 721C for 30 s, and a final extension of 721C for 10 min. The PCR products were separated in 1% agarose gels. Full-length Math1 and Cdx2 cDNAs were amplified by PCR from one-tenth of the first-strand cDNA (from the small intestinal RNA) using the primer sets Math1cDNAfw and Math1cDNArv, and Cdx2cDNAfwEcoRI and Cdx2cDNArvKpnI, respectively (Table 1). The Math1 cDNA was verified by corresponding double-restriction enzyme digestion and DNA sequencing. The amplified mouse Cdx2 cDNA was directly cloned into the TA cloning vector pCRII (Invitrogen, Carlsbad, CA) to yield the plasmid pCRII/Cdx2. pCRII/Cdx2 was digested with EcoRI and KpnI (sites underlined in the primers in Table 1), and the resulting fragment was subcloned into the EcoRI and KpnI restriction sites of the pcDNA3.1(  ) vector (Promega, Madison, WI) and confirmed by sequence analysis. A Math1 enhancer was previously identified 3 kb downstream of the Math1 coding sequence (Helms et al., 2000; Ebert et al., 2003). To construct the luciferase reporter vector pGL3-Basic1Math1 enhancer, a 1,517 bp sequence 3 0 of the mouse Math1 coding sequence (GenBank Accession Number, AF218258) and its deletion constructs were amplified by PCR with specific primers (Table 1) from 500 ng of mouse genomic DNA isolated from mouse tail according to standard phenol–chloroform extraction procedures. The amplified fragments for the Math1 enhancer were directly cloned into the TA cloning vector pCRII (Invitrogen) to yield the plasmid pCRII/Math1Enhancer. Each pCRII/Math1Enhancer was digested with BamHI and SalI (sites underlined in the primers in Table 1), and the resulting fragments were subcloned into the BamHI and SalI restriction sites of the pGL3-Basic vector (Promega) and confirmed by sequence analysis. Sequence of presumptive Cdx2-binding site (AACTAAC) was changed to AGAGCTC underlined in the primers by using Math1 SacI/1455-1436 and Math1 SacI/14611480 primers (Table 1).

IEC-6 cells IEC-6 cells were purchased from the American Type Culture Collection (Manassas, VA). These cells originated from normal rat small intestinal epithelium and were developed and characterized by Quaroni et al. (1979). IEC-6 cells lack specific immunological

determinants for differentiated villus cells and represent cultured epithelioid cells that have features of undifferentiated small intestinal crypt cells. IEC-6 cells were maintained in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal calf serum and 1% penicillin-streptomycin (Life Science-Gibco BRL, Gaithersburg, MD) at 371C under a humidified atmosphere of 5% CO2 and 95% air.

Northern blot analysis To determine the induction of Math1 by Cdx2 at the transcriptional level, IEC-6 cells were transfected with a Cdx2 expression plasmid or empty pcDNA3.1(  ) plasmid. Total RNA was extracted from the cells at 24 hr after transfection by the guanidinium isothiocyanate/phenol method (Isogen; Nippon Gene) according to the manufacturer’s instructions. Samples containing 10 mg of total RNA were electrophoresed in a 1% agarose–formaldehyde gel, transferred to a Hybond-N1 nylon membrane (Amersham Biosciences, Fairfield, CT) and baked at 801C for 2 hr. Next, the membrane was probed with 32P-labeled open reading frame probes for mouse Cdx2 and Math1. The probes were generated using a BcaBEST labeling kit (TaKaRa, Kyoto, Japan). Hybridization was performed at 681C overnight, and the membrane was subsequently exposed to X-ray film (Kodak, Rochester, NY). Northern blot was performed three times and the representative result was shown in Fig. 6.

Luciferase assays IEC-6 cells were seeded at 2  105 cells/well in Nunc 24-well dishes at 18–24 hr before transfection. Transient transfections were performed using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. Next, 100 ng of a Math1 enhancer reporter plasmid and 0.8 mg of a Cdx2 expression plasmid or empty expression vector pcDNA3.1(  ) were added to each plate, together with 50 ng of the Renilla luciferase control reporter plasmid pRLTK (Promega) as a control for the transfection efficiency. At 20 hr after the transfection, the cells were lysed in lysis buffer (Promega) and the firefly and Renilla luciferase activities were measured using the Dual-Luciferase Reporter Assay System (Promega) in a luminometer, according to the manufacturer’s instructions. The relative firefly luciferase activities were calculated by normalizing the transfection efficiency according to the Renilla luciferase activities produced by the internal control plasmid pRL-TK. The results from three separate experiments carried out in triplicate were analyzed by ANOVA. A value of po0.05 was considered significant.

316

Fig. 1 Math1 expression in mouse intestines. (A–C) Math1 expression is observed in normal small (B) and large (C) intestines, but not in normal gastric mucosa (A). Math1 is localized in the cytoplasm of some cells (arrowhead) and in the nucleus of other cells (arrow) in normal small intestine (D). Cdx2 (brown) and Math1 (blue) co-localize in the same cells in normal intestine (E). Math1

(brown) and chromogranin A (blue) co-localize in the same cells of intestinal mucosa from normal mice (F). Neither Cdx2 nor Math1 in normal small intestine was stained without the antibodies for Cdx2 and Math1 (G). Magnification:  200 (A, B, C, G),  400 (D, E, F). Scale bars: 50 mm (A, B, C, G), 20 mm (D, E, F).

Results

for absorptive enterocytes, lysozyme for Paneth cells and chromogranin A for enteroendocrine cells were examined. Antibody for chromogranin A stained 35% (35/100) of Math1-expressing cells (Fig. 1F) while antibodies for MUC2, CD10, and lysozyme did not stain any Math1-expressing cells (data not shown). The reason why we recognized Math1-expressing cells that were not stained by any antibodies for MUC2, CD10, lysozyme and chromogranin A might be that Math1 might be transiently expressed during the differentiation of intestinal epithelial cells and switch off before terminal differentiation. On the contrary, Math1, Neurogenin3, and NeuroD1 function in cascades, in which one factor activates a later factor in a sequential order, to control both the cell fate determination and differentiation of enteroendocrine cells. Therefore, chromogranin A might have expressed in 35% (35/100) of Math1expressing cells. Next, we examined Math1 expression in human gastric and intestinal epithelial cells. Math1 was not found in normal gastric mucosa (Fig. 2A), but was expressed in normal small and large intestines (Figs. 2B,2C). To clarify the correlation between Cdx2-expressing cells and Math1-positive cells among the intestinal epithelial cells, we investigated the expressions of Cdx2 and Math1 in normal intestinal epithelial cells by doubleimmunostaining for Cdx2 and Math1. All the cells that expressed Math1 were also co-stained for Cdx2 among the normal intestinal epithelial cells (Fig. 2D). These results indicate that Cdx2 is closely correlated with

Expression of Math1 in normal intestinal epithelial cells The basic helix–loop–helix transcription factor Math1 is related to the cell fate decision of enteroendocrine, goblet and Paneth cells in the intestine (Yang et al., 2001). We investigated Math1 expression in the stomach and small and large intestines of normal adult mice using immunohistochemistry. Math1 was not found in normal gastric mucosa (Fig. 1A), but was expressed in normal small and large intestines (Figs. 1B,1C). Many of Math1-expressing cells situated between the crypts and villus (Fig. 1B). Math1 was localized in the cytoplasm of some cells and in the nucleus of other cells in the normal small intestine (Fig. 1D). The intestine-specific homeobox transcription factor Cdx2 is also expressed in intestinal epithelial cells and related to their differentiation (Meyer and Gruss, 1993). To clarify the correlation between Cdx2-expressing cells and Math1positive cells among intestinal epithelial cells, we investigated the expressions of Cdx2 and Math1 in normal intestinal epithelial cells by double-immunostaining for Cdx2 and Math1. The cells that expressed Math1 were all co-stained for Cdx2 among the normal intestinal epithelial cells (Fig. 1E). In total, 4% (8/200) of Cdx-2 positive cells expressed Math1 among the normal intestinal epithelial cells. We performed double immunostaining to identify cell type of Math1-expressing cells in the mouse small intestine. MUC2 for goblet cells, CD10

317

Fig. 2 Math1 expression in human intestines. (A–C) Math1 expression is observed in normal small (B) and large (C) intestines, but not in normal gastric mucosa (A). Cdx2 (brown) and Math1 (blue) co-localize in the same cells in normal intestine (D). Neither

Cdx2 nor Math1 in normal small intestine was stained without the antibodies for Cdx2 and Math1 (E). Magnification:  200 (A, B, C, E),  400 (D). Scale bars: 50 mm (A, B, C, E), 20 mm (D).

Math1 in normal intestinal mucosa in humans and mice.

under the control of the rat H1/K1-ATPase b-subunit gene promoter (Mutoh et al., 2002). These mice develop normally into superficially healthy adults but show progressive intestinal metaplasia of the stomach until 12 weeks of age. Subsequently, the gastric mucosa of these Cdx2 transgenic mice is completely replaced by intestinal metaplastic mucosa, which consists of terminally differentiated intestinal epithelial cells including absorptive enterocytes, enteroendocrine cells and goblet

Expression of Math1 in Cdx2-induced intestinal metaplastic epithelial cells Previously, we generated Cdx2 transgenic mice that exclusively expressed the Cdx2 gene in gastric mucosa

Fig. 3 Math1 expression in intestinal metaplastic mucosa from Cdx2 transgenic mice. Math1 expression is observed in intestinal metaplastic mucosa from Cdx2 transgenic mouse stomach (A). Cdx2 (blue) and Math1 (brown) co-localize in the same cells of intestinal metaplastic mucosa from Cdx2 transgenic mouse stomach (B). Neither Cdx2 nor Math1 in intestinal metaplastic mucosa was stained without the antibodies for Cdx2 and Math1 (C). Magnification:  400 (A, B, C). Scale bars: 50 mm (A, B, C).

318

Fig. 4 Math1 expression in human intestinal metaplastic mucosa. Math1 expression is observed in intestinal metaplastic mucosa from human stomach (A). Cdx2 (brown) and Math1 (blue) co-localize in the same cells of intestinal metaplastic mucosa from human stomach (B). Magnification: 200 (A), 400 (B). Scale bars: 50 mm (A), 20 mm (B).

cells. To clarify whether Cdx2 is related to Math1 expression in vivo, we examined the expression of Math1 in Cdx2-induced intestinal metaplastic mucosa from Cdx2 transgenic mice. Math1-expressing epithelial cells were observed in Cdx2-expressing intestinal metaplastic mucosa of Cdx2-transgenic mice (Fig. 3A). Next, we investigated whether Cdx2 and Math1 were co-expressed in the same cells of Cdx2-induced intestinal metaplastic mucosa by double-immunostaining for Cdx2 and Math1 to further clarify their relationship. All the cells that expressed Math1 were co-stained for Cdx2 in intestinal metaplastic mucosa (Fig. 3B). In total, 3% (6/200) of Cdx2-positive cells expressed Math1 in Cdx2-induced intestinal metaplastic mucosa. These results indicate that Cdx2 is closely related to Math1 in Cdx2-induced intestinal metaplastic mucosa. Furthermore, we examined Math1 expression in human intestinal metaplastic mucosa. Math1-expressing epithelial cells were observed in human intestinal metaplastic mucosa (Fig. 4A). Next, we investigated whether Cdx2 and Math1 were co-expressed in the same cells of intestinal metaplastic mucosa by double-immunostaining for Cdx2 and Math1 to further clarify their relationship. All the cells that expressed Math1 were co-stained for Cdx2 in intestinal metaplastic mucosa (Fig. 4B).

Cdx2 up-regulates Math1 gene transcription in IEC-6 cells To determine whether expression of Cdx2 can alter the level of endogenous Math1 transcription, we expressed Cdx2 by transient transfection of a Cdx2 cDNA expression vector into IEC-6 cells (Fig. 6). As a control, IEC-6 cells were transiently transfected with the empty expression vector pcDNA3.1(  ). Northern blotting was performed with total RNA from the transfected cells, and the blots were hybridized with probes for Cdx2 and Math1. Expression of Cdx2 in IEC-6 cells (Fig. 6A) led to an increase in the Math1 mRNA signal compared with control cells (Fig. 6C). Math1 is expressed in the normal mouse intestine while it is not expressed in the normal mouse stomach (Fig. 6C). These results suggest that Cdx2 up-regulates the level of Math1 gene expression in IEC-6 cells, and are consistent with recently published data (Uesaka et al., 2004). Cdx2 regulates the Math1 enhancer activity Next, we examined the ability of Cdx2 to activate Math1 gene transcription by transiently cotransfecting

Expression of Math1 mRNA transcripts We examined whether Math1 is expressed in the stomach (normal mice), intestine (normal mice), and intestinal metaplastic mucosa (Cdx2 transgenic mice) at the transcriptional level. Math1 mRNA expression was observed in the small intestine (normal mice), and intestinal metaplastic mucosa (Cdx2 transgenic mice), but not in the stomach of normal mice (Fig. 5). These results coincided with the immunohistochemical staining for the normal stomach, normal small intestine and Cdx2 transgenic mouse stomach. As Cdx2 is expressed in the normal small intestine and Cdx2 transgenic mouse stomach, but not in the normal stomach, these results indicate that Cdx2 may induce Math1 expression.

Fig. 5 Reverse transcription-polymerase chain reaction (RT-PCR) analysis of Math1 expression. RT-PCR analyses of Math1 mRNA transcripts in normal mouse stomach (1), Cdx2 transgenic mouse stomach (2) and normal mouse small intestine (3) are shown. (M: marker).

319 Luciferase activity (Ratio of firefly to renilla luciferase) 0 10 20 30 40 1

Luciferase

688 688

vector Cdx2

1266 1517

* 1266

1517

* Fig. 6 Induction of Math1 transcription by expression of Cdx2 in intestinal crypt-derived cells. (A–D) IEC-6 cells were transiently transfected with pcDNA3.1(  )  Cdx2 or the pcDNA3.1(  ) empty vector as a control. Northern blotting was performed with 10 mg of total RNA from the transfected cells and from stomach and small intestine of normal mice, and the blots were separately hybridized with probes for Cdx2 (A) and Math1 (C). Loading was assessed by ethidium bromide staining of rRNAs (B and D).

IEC-6 cells with a mammalian expression vector for Cdx2 and a Math1 promoter-luciferase reporter plasmid. However, the promoter-luciferase reporter containing 2,000 bp sequence from transcriptional initiation site was not activated by the cotransfection of a mammalian expression vector for Cdx2 (data not shown). Ebert et al. previously reported that a Math1 enhancer (1,517 bp sequence) resides at  3.4 kb on the 3 0 side of the mouse Math1 coding sequence (Helms et al., 2000; Ebert et al., 2003). Therefore, we examined whether Cdx2 could activate the transcriptional activities of different deletion constructs of the Math1 enhancer in IEC-6 cells (Fig. 7). Cotransfection of the Cdx2 expression plasmid stimulated expression of the Math1 enhancer reporter gene containing an element between residues 1,359 and 1,517 (Fig. 7). This result suggests that the element between residues 1,359 and 1,517 in the Math1 enhancer may be important for Cdx2 activation of Math1 gene transcription in IEC-6 cells. In an effort to identify potential Cdx2-binding sites in the sequence between 1,359 and 1,517 in the Math1 enhancer, using a consensus binding-element for the CdxA chicken caudal-related protein (5 0 -A, A/T, T, A/ T, A, T, A/G-3 0 ) (Margalit et al., 1993), a candidate Cdx2-binding site was found in the region from 1,455 to 1,461 (AACTAAC). Analysis of reporter construct with mutation of the Cdx2 consensus-binding element in the residues 1,359 and 1,517 revealed that site (1,455–1,461) was critical for transcription activity of the Math1 reporter gene construct in IEC6 cells (Fig. 7).

Discussion In the present study, we have shown that the intestinespecific homeobox gene Cdx2 induces expression of the basic helix–loop–helix transcription factor Math1.

1359

1517

* 1266

1517 AACTAAC AGAGCTC

Fig. 7 Transient transfection assays investigating the effect of Cdx2 on a Math1 enhancer. A schematic diagram of the Math1 enhancer (1,517 bp sequence) that resides  3.4 kb on the 3 0 side of the mouse Math1 coding sequence is shown (GenBank Accession Number, AF218258). Various lengths of the Math1 enhancer were cloned into the pGL3-basic plasmid, and cells were cotransfected with a Math1-luciferase reporter gene and a Cdx2 expression plasmid. The results are expressed as the relative luciferase activity compared with that of the reporter vector (pGL3-basic vector) and the Cdx2 expression vector. Each bar represents the mean  SD. Transfections were performed in triplicate and repeated three times. po0.01 versus Cdx2-transfected cells.

The stomach and intestine are both derived from the primitive undifferentiated gut tube formed during gastrulation, but Cdx2 expression is only activated in regions distal to the gastric-duodenal junction (James et al., 1994; Silberg et al., 2000). In addition to its expression during gut tube development, Cdx2 is also expressed in the adult intestine where it plays a pivotal role in intestinal epithelial differentiation. The generation of intestinal epithelium following misexpression of Cdx2 in the stomach in our Cdx2 transgenic mice also clearly indicates that Cdx2 has an essential role in intestinal epithelial differentiation. These results suggest that Cdx2 may be a key regulator of the generation of intestinal epithelial cells for intestinal development as well as of intestinal differentiation. In addition to Cdx2, Math1 is also expressed in the gut. The epithelium of the small intestine consists of four principal cell types, namely enterocytes, goblet cells, enteroendocrine cells, and Paneth cells, which are all derived from one multipotent stem cell type. Studies of Math1 knock-out mice have revealed that loss of Math1 leads to depletion of the secretory cell lineages (goblet, enteroendocrine, and Paneth cells) in the gut, without affecting enterocytes (Yang et al., 2001). However, very little is known about the genetic programs that control the commitment of the multipotent stem cells to the Math1-positive

320

secretory lineages. In the present study, we observed that Math1 expression within the gut was restricted to the epithelia of the small and large intestines. Although Math1 expression was not detected in normal gastric epithelial cells, it was observed in the Cdx2-expressing epithelial cells of intestinal metaplastic mucosa from the Cdx2 transgenic mouse stomach. Math1 was also detected in human intestinal metaplastic mucosa as well as in human intestinal epithelium. Suh and Traber (1996) reported that expression of Cdx2 in stably transfected IEC-6 cells induced morphological changes toward enterocyte-like cells and goblet cell-like cells, which represent two of the four intestinal epithelial cell lineages. We further demonstrated that expression of Cdx2 stimulated endogenous Math1 mRNA expression in Cdx2transfected IEC-6 cells. Moreover, transfection of a Cdx2 expression plasmid stimulated the transcriptional activity of a Math1 reporter gene in IEC-6 cells. These results indicate that Cdx2 regulates the expression of Math1. Adult intestinal epithelium is replaced every 3–4 days by the generation of multiple cell lineages from multipotent epithelial stem cells. The different intestinal cell types are not directly derived from these stem cells, but produced via an intermediate population of committed progenitor cells, probably Math1-expressing cells. In addition to Math1, other neurogenic basic helix–loop– helix transcription factors, namely Neurogenin3 and NeuroD1, have been reported to regulate cell fate determination in the intestine. For example, Neurogenin3 and NeuroD1 regulate the cell fate specification of in-

testinal endocrine cells. Math1, Neurogenin3, and NeuroD1 function in cascades, in which one factor activates a later factor in a sequential order, to control both the cell fate determination and differentiation of specific cell types, and appear to be expressed sequentially during the differentiation of enteroendocrine cells. Cdx2 may play a pivotal role at an early stage of the transcriptional cascade by inducing Math1. Only a few cells within the Cdx2-positive intestinal epithelium expressed Math1, indicating that other factors may be needed for its expression in addition to Cdx2. Furthermore, Cdx2 induced absorptive enterocytes in addition to secretory cell lineages. Math1 is also regulated by Notch signaling (Jensen et al., 2000; Zheng et al., 2000). Deletion of Hes1, a factor downstream of activated Notch that represses basic helix– loop–helix transcriptional activators, was reported to elevate Math1 expression in the small intestine and lead to increased numbers of enteroendocrine and goblet cells but fewer enterocytes (Jensen et al., 2000). Hes1 also negatively regulates inner ear hair cell differentiation by suppressing Math1 (Zheng et al., 2000). These studies support the hypothesis that Math1 controls cell fate determination and that Hes1 negatively regulates Math1-mediated differentiation in the gut. However, positive regulators that induce the differentiation of progenitor stem cells into Math1-expressing cells in the crypt have not yet been clarified. The expression of Math1 in Cdx2-induced intestinal metaplastic mucosa suggests that Math1 is a downstream target of Cdx2. It is reasonable to suggest that the secretory and absorptive lineages arise from two types of progenitors specified on the basis of their Math1 expression. Cdx2 may induce the differentiation of epithelial stem cells into Math1-positive multipotent intermediate cells, thereby leading to the production of the secretory cell lineages. It can be hypothesized that the positive regulation of Math1 by Cdx2 is antagonized by Hes1 and that Hes1positive and Math1-negative intermediate cells lead to the absorptive cell lineage. Therefore, regulation of Math1 by Hes1 and Cdx2 may regulate the earliest stage of cell determination in the intestinal epithelium. In conclusion, Cdx2 induces expression of the basic helix–loop–helix transcription factor Math1 in intestinal epithelial cells. Acknowledgments This work was supported in part by a Grant-inAid for Scientific Research (B) (12470126 to HM) from the Japan Society for the Promotion of Science.

References Fig. 8 Co-expression of Math1 and chromogranin A in the intestinal mucosa from normal mice. Math1 (brown) and chromogranin A (blue) co-localize in the same cells of intestinal mucosa from normal mice. Magnification:  400. Scale bars: 50 mm.

Akazawa, C., Ishibashi, M., Shimizu, C., Nakanishi, S. and Kageyama, R. (1995) A mammalian helix–loop–helix factor structurally related to the product of Drosophila proneural gene atonal is a positive transcriptional regulator expressed in the developing nervous system. J Biol Chem 270:8730–8738.

321 Beck, F., Chawengsaksophak, K., Luckett, J., Giblett, S., Tucci, J., Brown, J., Poulsom, R., Jeffery, R. and Wright, N.A. (2003) A study of regional gut endoderm potency by analysis of Cdx2 null mutant chimaeric mice. Dev Biol 255:399–406. Beck, F., Chawengsaksophak, K., Waring, P., Playford, R.J. and Furness, J.B. (1999) Reprogramming of intestinal differentiation and intercalary regeneration in Cdx2 mutant mice. Proc Natl Acad Sci USA 96:7318–7323. Ben-Arie, N., Bellen, H.J., Armstrong, D.L., McCall, A.E., Gordadze, P.R., Guo, Q., Matzuk, M.M. and Zoghbi, H.Y. (1997) Math1 is essential for genesis of cerebellar granule neurons. Nature 390:169–172. Bermingham, N.A., Hassan, B.A., Price, S.D., Vollrath, M.A., Ben-Arie, N., Eatock, R.A., Bellen, H.J., Lysakowski, A. and Zoghbi, H.Y. (1999) Math1: an essential gene for the generation of inner ear hair cells. Science 284:1837–1841. Bonhomme, C., Duluc, I., Martin, E., Chawengsaksophak, K., Chenard, M.P., Kedinger, M., Beck, F., Freund, J.N. and Domon-Dell, C. (2003) The Cdx2 homeobox gene has a tumour suppressor function in the distal colon in addition to a homeotic role during gut development. Gut 52:1465–1471. Chawengsaksophak, K., James, R., Hammond, V.E., Kontgen, F. and Beck, F. (1997) Homeosis and intestinal tumours in Cdx2 mutant mice. Nature 386:84–87. Ebert, P.J., Timmer, J.R., Nakada, Y., Helms, A.W., Parab, P.B., Liu, Y., Hunsaker, T.L. and Johnson, J.E. (2003) Zic1 represses Math1 expression via interactions with the Math1 enhancer and modulation of Math1 autoregulation. Development 130:1949–1959. Helms, A.W., Abney, A.L., Ben-Arie, N., Zoghbi, H.Y. and Johnson, J.E. (2000) Autoregulation and multiple enhancers control Math1 expression in the developing nervous system. Development 127:1185–1196. Hinoi, T., Loda, M. and Fearon, E.R. (2003) Silencing of CDX2 expression in colon cancer via a dominant repression pathway. J Biol Chem 278:44608–44616. James, R., Erler, T. and Kazenwadel, J. (1994) Structure of the murine homeobox gene cdx-2. Expression in embryonic and adult intestinal epithelium. J Biol Chem 269:15229–15237. Jensen, J., Pedersen, E.E., Galante, P., Hald, J., Heller, R.S., Ishibashi, M., Kageyama, R., Guillemot, F., Serup, P. and Madsen, O.D. (2000) Control of endodermal endocrine development by Hes-1. Nat Genet 24:36–44. Margalit, Y., Yarus, S., Shapira, E., Gruenbaum, Y. and Fainsod, A. (1993) Isolation and characterization of target sequences of the chicken cdxa homeobox gene. Nucleic Acids Res 21:4915–4922. Meyer, B.I. and Gruss, P. (1993) Mouse cdx-1 expression during gastrulation. Development 117:191–203. Mizoshita, T., Inada, K., Tsukamoto, T., Kodera, Y., Yamamura, Y., Hirai, T., Kato, T., Joh, T., Itoh, M. and Tatematsu, M. (2001) Expression of Cdx1 and Cdx2 mrnas and relevance of this expression to differentiation in human gastrointestinal mucosa— with special emphasis on participation in intestinal metaplasia of the human stomach. Gastric Cancer 4:185–191.

Mutoh, H., Hakamata, Y., Sato, K., Eda, A., Yanaka, I., Honda, S., Osawa, H., Kaneko, Y. and Sugano, K. (2002) Conversion of gastric mucosa to intestinal metaplasia in Cdx2- expressing transgenic mice. Biochem Biophys Res Commun 294: 470–479. Mutoh, H., Sakurai, S., Satoh, K., Osawa, H., Hakamata, Y., Takeuchi, T. and Sugano, K. (2004a) Cdx1 induced intestinal metaplasia in the transgenic mouse stomach: comparative study with Cdx2 transgenic mice. Gut 53:1416–1423. Mutoh, H., Sakurai, S., Satoh, K., Osawa, H., Tomiyama, T., Kita, H., Yoshida, T., Tamada, K., Yamamoto, H., Isoda, N., Ido, K. and Sugano, K. (2005) Pericryptal fibroblast sheath in intestinal metaplasia and gastric carcinoma. Gut 54:33–39. Mutoh, H., Sakurai, S., Satoh, K., Tamada, K., Kita, H., Osawa, H., Tomiyama, T., Sato, Y., Yamamoto, H., Isoda, N., Yoshida, T., Ido, K. and Sugano, K. (2004b) Development of gastric carcinoma from intestinal metaplasia in Cdx2-transgenic mice. Cancer Res 64:7740–7747. Osawa, H., Kita, H., Satoh, K., Ohnishi, H., Kaneko, Y., Mutoh, H., Tamada, K., Ido, K. and Sugano, K. (2004) Aberrant expression of CDX2 in the metaplastic epithelium and inflammatory mucosa of the gallbladder. Am J Surg Pathol 28: 1253–1254. Qualtrough, D., Hinoi, T., Fearon, E. and Paraskeva, C. (2002) Expression of CDX2 in normal and neoplastic human colon tissue and during differentiation of an in vitro model system. Gut 51:184–190. Quaroni, A., Wands, J., Trelstad, R.L. and Isselbacher, K.J. (1979) Epithe-lioid cell cultures from rat small intestine: characterization by morphologic and immunologic criteria. J Cell Biol 80:248–265. Silberg, D.G., Swain, G.B., Suh, E.R. and Traber, P.G. (2000) Cdx1 and Cdx2 expression during intestinal development. Gastroenterology 119:961–971. Suh, E. and Traber, P.G. (1996) An intestine-specific homeobox gene regulates proliferation and differentiation. Mol Cell Biol 16:619–625. Tamai, Y., Nakajima, R., Ishikawa, T., Takaku, K., Seldin, M.F. and Taketo, M.M. (1999) Colonic hamartoma development by anomalous duplication in Cdx2 knockout mice. Cancer Res 59:2965–2970. Tot, T. (2004) Identifying colorectal metastases in liver biopsies: the novel CDX2 antibody is less specific than the cytokeratin 201/ 7  phenotype. Med Sci Monit 10:BR139–BR143. Uesaka, T., Kageyama, N. and Watanabe, H. (2004) Identifying target genes regulated downstream of Cdx2 by microarray analysis. J Mol Biol 337:647–660. Yang, Q., Bermingham, N.A., Finegold, M.J. and Zoghbi, H.Y. (2001) Requirement of Math1 for secretory cell lineage commitment in the mouse intestine. Science 294:2155–2158. Zheng, J.L., Shou, J., Guillemot, F., Kageyama, R. and Gao, W.Q. (2000) Hes1 is a negative regulator of inner ear hair cell differentiation. Development 127:4551–4560.