Cell-Specific Expression of Glucose-Dependent-Insulinotropic Polypeptide Is Regulated by the Transcription Factor PDX-1
http://www.100md.com
内分泌学杂志 2005年第1期
Section of Gastroenterology (L.I.J., M.O.B., C.N.W., M.M.W.), Boston University School of Medicine and Boston Medical Center, Boston, Massachusetts 02118; and Department of Cell and Developmental Biology (Y.F., C.V.W.), Vanderbilt University Medical Center, Nashville, Tennessee 37232
Address all correspondence and requests for reprints to: M. Michael Wolfe, M.D., Section of Gastroenterology, Boston Medical Center, 650 Albany Street, Boston, Massachusetts 02118. E-mail: michael.wolfe@bmc.org.
Abstract
Glucose-dependent insulinotropic polypeptide (GIP) is a potent stimulator of insulin secretion and comprises an important component of the enteroinsular axis. GIP is synthesized in enteroendocrine K-cells located principally in the upper small intestine. The homeobox-containing gene PDX-1 is also expressed in the small intestine and plays a critical role in pancreatic development and in the expression of pancreatic-specific genes. Previous studies determined that the transcription factors GATA-4 and ISL-1 are important for GIP expression. In this study, we demonstrate that PDX-1 is also involved in regulating GIP expression in K-cells. Using immunohistochemistry, we verified the expression of PDX-1 protein in the nucleus of GIP-expressing mouse K-cells and evaluated the expression of PDX-1, serotonin, and GIP in wild-type and PDX-1–/– mice at 18.5 d after conception. Although we demonstrated a 97.8% reduction in the number of GIP-expressing cells in PDX-1–/– mice; there was no statistical difference in the number of serotonin-positive cells. Additionally, PDX-1 transcripts and protein were detected in a GIP-expressing neuroendocrine cell line, STC-1. Electromobility shift assays using STC-1 nuclear extracts demonstrated the specific binding of PDX-1 protein to a specific regulatory region in the GIP promoter. Using chromatin immunoprecipitation analysis, we demonstrated binding of PDX-1 to this same region of the GIP promoter in intact cells. Lastly, overexpression of PDX-1 in transient transfection assays led to a specific increase in the activity of GIP/Luc reporter constructs. The results of these studies indicate that the transcription factor PDX-1 plays a critical role in the cell-specific expression of the GIP gene.
Introduction
GLUCOSE-DEPENDENT INSULINOTROPIC polypeptide (GIP) is a 42-amino-acid peptide that is a member of the secretin-vasoactive intestinal polypeptide family of gastrointestinal regulatory peptides (1). The primary site of endogenous GIP release is the endocrine K-cells dispersed throughout the mucosa of the duodenum and proximal jejunum (2). GIP is released into circulation principally in response to the ingestion of two major nutrient stimuli, carbohydrate and fat (3, 4). Studies have established that the primary function of GIP is to stimulate insulin secretion from the pancreatic ?-cells in the presence of elevated blood glucose (4, 5, 6). Thus, GIP, along with glucagon-like peptide-1 (GLP-1), function as incretins, the proposed mediators of the enteroinsular axis that help maintain glucose homeostasis under physiological conditions. Although GLP-1 has been established as a more potent pharmacological stimulator of insulin release (7, 8), it has recently been demonstrated, using GIP and GLP-1-specific receptor antagonists, that GIP is the major physiological incretin, accounting for approximately 80% of nutrient-induced enteroinsular pancreatic ?-cell stimulation (9, 10).
Recent studies have demonstrated the importance of defining both the regulation of GIP expression and the process of K-cell differentiation. Miyawaki et al. (11) created a GIP-receptor-deficient (GIPR–/–) mouse to investigate the role of GIP in the development of obesity. These investigators demonstrated that although normal littermates became obese and developed insulin resistance and type 2 diabetes mellitus in response to a high-fat diet, GIPR–/– mice were protected and remained normal while consuming an identical diet. The results of this study provide evidence for GIP as an important mediator of nutrient deposition, and it has been proposed that attenuation of the GIP response could be used as a treatment for obesity. It is thus essential to determine the mechanisms controlling GIP expression, secretion, and function.
In a second important study, Cheung et al. (6) engineered a transgenic mouse in which the human insulin gene was expressed under the control of the rat GIP promoter. In these mice, human insulin mRNA was detected only in the duodenum and the stomach, corresponding to the known tissue distribution of GIP. Furthermore, immunohistochemical analysis using antisera specific for GIP and human insulin confirmed that insulin biosynthesis was targeted specifically to the enteroendocrine K-cells of the stomach and duodenum. When native islet ?-cells of these transgenic mice were eliminated by the use of streptozocin, human insulin secreted from intestinal K-cells rendered the mice euglycemic and protected them against the development of diabetes mellitus. These findings demonstrate that K-cells, like pancreatic ?-cells, are capable of processing and storing insulin and of releasing it in a manner that maintains normal glucose homeostasis. Furthermore, they establish the K-cell as a potential cell to exploit in gene therapy for the treatment of diabetes mellitus and demonstrate the effectiveness of the GIP promoter in directing transgene expression to the K-cell.
The important regulatory elements in the GIP promoter that direct cell-specific expression are beginning to be identified and understood. The human and rat GIP genes have been isolated and their promoter sequences partially characterized (12, 13). Although two potential promoter regions were identified upstream of the translation initiation codon, primer extension and RNase protection analysis indicated that GIP transcription initiates from the upstream promoter in the duodenum of adult rats (14). Deletion analysis of the GIP promoter revealed that the first 193 bp upstream of the transcription initiation site were sufficient to direct specific expression of the gene in STC-1 cells (15). Analysis of this region led to the identification of two transcription factors whose binding is necessary for GIP expression in STC-1 cells (15, 16). A binding motif located at base pairs –190 to –184 with respect to the transcription initiation site (+1) of the GIP promoter and conforming to the consensus motif for the GATA family of DNA binding proteins, AGATAA (17, 18, 19), was found to bind the transcription factor GATA-4 (16). A second cis-regulatory region downstream of the GATA-4 binding site was identified that binds the transcription factor ISL-1 and appears to function in conjunction with GATA-4 to promote GIP gene expression (16).
PDX-1, a homeobox-containing gene that can recognize the same DNA consensus motif as ISL-1, has been shown to be important in both the development of the pancreas and the expression of pancreatic-specific genes. In this present study, we demonstrate the presence of PDX-1 in native K-cells and provide evidence to support the integral involvement of PDX-1 in the regulation of GIP expression.
Materials and Methods
Animals
The PDX-1XBko mice used in these studies are described in detail elsewhere (20). Briefly, in this mouse line, the protein-encoding sequences in exon 2 of the PDX-1 gene are deleted, including those encoding the DNA-binding homeodomain, thus creating a functional null PDX-1 allele. Studies for this report were performed on 18.5 d postconception (dpc) PDX-1XBko–/– mice and their age-matched, wild-type littermates.
Cell culture
STC-1 cells, a mouse neuroendocrine cell line (Dr. D. Hanahan, San Francisco, CA), and IEC-6 cells, an immature intestinal stem cell line derived from normal rat small intestine (American Type Culture Collection, Manassas, VA), were grown in DMEM containing 10% fetal bovine serum at 37 C in an atmosphere of 5% CO2 in the presence of 100 U/ml penicillin G, 100 μg/ml streptomycin, and 0.25 μg/ml amphotericin B.
Immunohistochemistry
Embryos were dissected at 18.5 dpc, and tissues were fixed in ice-cold 4% paraformaldehyde at 4 C for 60 min. Tissues were then dehydrated in an increasing ethanol series, which was followed by two Histo-Clear (National Diagnostics, Atlanta, GA) washes, infiltration in Histo-Clear/paraffin (1:1 vol/vol), and three changes of paraffin under vacuum at 56 C. Fixed tissues were embedded and cut into 5-μm sections. Deparaffinized and rehydrated slides were subjected to microwave antigen retrieval in a 10 mM citric acid buffer (pH 6.0) and allowed to cool for 30 min at room temperature (RT). Slides were washed in PBS and then blocked with 5% normal donkey serum for 60 min at RT. The primary antibodies were diluted in PBS containing 1% BSA and incubated with the sections overnight at 4 C. Slides were washed in PBS and incubated with the appropriate secondary antibodies diluted in PBS containing 1% BSA for 1 h at RT. Slides were washed in PBS and then counterstained with the nuclear dye YO-RPO-1 (Molecular Probes, Eugene, OR) (1:1000 dilution in PBS) for 10 min. Slides were washed again in PBS, mounted, and examined using confocal microscopy (Zeiss LSM-510). TIFF images were processed in Adobe Photoshop.
The following primary antibodies were used at the indicated dilutions for immunohistochemistry: guinea pig antibody against PDX-1, 1:5000 (21); and rabbit antibody against GIP (Research Diagnostics, Inc., Flanders, NJ), 1:1000. For immunofluorescence, antirabbit immunoglobulin conjugated to Cy3 (for GIP) and anti-guinea pig immunoglobulin-conjugated Cy5 (for PDX-1) were used as secondary antibodies.
Quantification of GIP- and serotonin-positive cells
Sections of wild-type and PDX-1-deficient mice at 18.5 dpc were stained with rabbit antiserum to GIP or mouse monoclonal antibody against serotonin (Dako, Carpinteria, CA) (22). For the counting of immunoreactive cells, sections were subjected to immunoperoxidase staining. GIP and serotonin stained sections were lightly counterstained with hematoxylin to reveal nuclei and general morphology. Ten random fields of each section were digitally photographed using a Spot camera and printed at x450 magnification to count the number of GIP- or serotonin-positive cells, as well as the total number of epithelial cells. For each animal examined, cell counting was performed on five nonadjacent longitudinal sections of rostral duodenum, thus avoiding the double scoring of cells. In all, three wild-type and three PDX-1–/– mice were examined. The number of immunopositive nucleated cells per 1000 epithelial cells was expressed as the mean ± SE (n = 3) and analyzed statistically using Student’s t test. Statistical significance was assigned if P 0.05.
Western blot analysis
Cells grown on 10-cm plates to 80% confluence were washed once with 1x PBS and then suspended by scraping in 1 ml of RIPA buffer (Boston Bioproducts, Ashland, MA) containing Complete protease inhibitor cocktail (Roche Molecular Biochemicals, Indianapolis, IN) followed by incubation on ice for 15 min. Cell lysates were centrifuged at 14,000 rpm at 4 C for 5 min, and the supernatant was aspirated. Protein concentrations were measured by using the BCA protein assay kit (Pierce, Rockford, IL), and protein lysates (50 μg) were subjected to SDS-PAGE and transferred to Protran nitrocellulose membranes (Perkin Elmer Life Sciences, Boston, MA). Membranes were blocked in 5% nonfat milk, 1x PBS, and 0.5% Tween 20 at RT for 2 h and then incubated with the appropriate primary antibody [rabbit anti-PDX-1, 1 1:500 (23), or mouse anti-?-actin, 1:20,000 (BD Biosciences, Palo Alto, CA)] diluted in 5% nonfat milk in 1x PBS. After a 1-h incubation, membranes were washed three times for 5 min at RT with wash buffer (1x PBS, 0.5% Tween 20) and incubated with the appropriate horseradish peroxidase-conjugated secondary antiserum diluted in 5% nonfat milk in 1x PBS. Membranes were then rinsed three times for 5 min in wash buffer, soaked in chemiluminescence reagent, as instructed by the manufacturer (Super Signal West Pico Chemiluminescent substrate, DuPont, Wilmington, DE), and exposed to x-ray film for 0.1–5 min.
Immunofluorescence cell staining
Cells were plated onto Lab-Tek chamber slides (Nunc International, Naperville, IL) to approximately 25% confluence and grown overnight in normal media. On the following day, the cells were washed once with PBS, fixed by a 10-min incubation in Histochoice tissue fixative (Sigma Chemical Co., St. Louis, IL), and then washed again with PBS for 10 min and permeabilized by a 5-min incubation in 0.1% Triton, 1x PBS. After an additional wash in PBS, cells were blocked in PBS containing 10% donkey serum (Santa Cruz Biotechnology Inc., Santa Cruz, CA) for 20 min, washed with PBS, and incubated for 1 h with primary serum diluted in PBS with 1.5% donkey serum, followed by three washes with PBS for 5 min each. Next, the cells were incubated for 45 min with a Cy3-conjugated secondary antibody diluted 1:200 in PBS, containing 1.5% normal blocking serum. Cells were again washed three times for 5 min with PBS, and a coverslip was mounted using ProLong Antifade (Molecular Probes). Slides were examined and photographed using a fluorescence microscope with appropriate filters. Antibodies to the following proteins were used at the listed concentrations: GIP (24) at 1:400 dilution and PDX-1, rabbit polyclonal, at 1:1000 (23). The secondary antibody used was Cy3-conjugated donkey antirabbit (Jackson ImmunoResearch Lab Inc., West Grove, PA).
Northern blot analysis
Total RNA was extracted from cells grown to 80% confluence using a RNeasy kit, as recommended by the manufacturer (QIAGEN, Valencia, CA). Total RNA (20 μg, quantified by measuring absorbance at 260 nm) was size-fractionated on a 0.8% agarose gel containing 2.2 M formaldehyde and then transferred to a GeneScreen Plus nylon membrane (NEN Life Science Products, Boston, MA) by capillary blotting in 10x saline-sodium citrate buffer (SSC) (pH 7.4). The RNA was cross-linked to the membrane using a UV-Stratalinker (Stratagene, La Jolla, CA), and the blot was prehybridized for 2 h at 42 C in 5x SSC, 10x Denhardt’s solution, 50% formamide, 1% SDS, and 10 μg/ml herring sperm DNA. The filter was then hybridized overnight at 42 C in 5x SSC, 1x Denhardt’s solution, 50% formamide, 20 mM NaPO4, 0.5% SDS, 20 μg/ml herring sperm DNA, and approximately 107 cpm of probe per filter. After hybridization, the blots were washed once at RT in 1x SSC, 1% SDS for 15 min, once at RT in 0.5x SSC, 0.5% SDS for 15 min, twice at RT in 0.1x SSC, 0.1% SDS for 15 min, and once at 50 C in 0.1x SSC, 0.1% SDS for 30 min. Autoradiographs were developed after exposure to x-ray film for 24 h at –70 C, using a Cronex intensifying screen (DuPont).
Electrophoretic mobility shift assays (EMSAs)
Nuclear extracts were prepared according to the method of Dignam et al. (25). The protein concentration was subsequently estimated using a BCA protein assay kit. EMSAs were performed using 4% polyacrylamide gels (44:1 acrylamide/bisacrylamide) in 40 mM Tris/HCl and 195 mM glycine (pH 8.5) at 4 C. The following probe sequence was used: wild type, 5'-GTCACCCATTAGCACAGGCC-3 (presumptive binding site underlined). Each reaction contained 10,000 cpm (7 fmol) of double-stranded oligonucleotide probe, end-labeled by T4 kinase in the presence of [-32P]ATP and purified on a Quik Spin G25 column (Roche Molecular Biochemicals). The reactions were performed in 20 μl of a mixture containing 20 mM HEPES (pH 7.3), 50 mM KCl, 5 mM ?-mercaptoethanol, 20% glycerol, 1 μg poly dI-dC, 1 μg BSA, and 2.5 μg STC-1 cell nuclear extract. The reaction mixtures were incubated for 30 min at RT, followed by 5 min on ice. The samples were separated by electrophoresis for 1.5–2.5 h at 250 V. For supershift experiments, antibodies were preincubated with nuclear extract for 10 min at RT before the addition of the probe. The affinity-purified polyclonal rabbit anti-ISL-1 was purchased from Chemicon (Temecula, CA), and rabbit anti-GATA-4 was purchased from Santa Cruz Biotechnology.
Chromatin immunoprecipitation assays (ChIPs)
ChIPs were performed using the protocol of Gerrish et al. (26) and Cissell et al. (27). STC-1 cells grown on 10-cm plates to 90% confluence were exposed to 1% formaldehyde in DMEM for 5 min at 23 C. Glycine was added to a final concentration of 0.125 M to quench the formaldehyde, and the cultures were incubated for 2 min. The cells were collected in ice-cold PBS, pelleted by centrifugation, and incubated for 10 min on ice in 0.6 ml of SDS lysis buffer [1% SDS, 10 mM EDTA, 50 mM Tris/HCl (pH 8.1), and 1 mM phenylmethylsulfonyl fluoride]. Lysed samples were transferred to prechilled microcentrifuge tubes and subjected to sonication consisting of 12 10-sec pulses. The reactions were centrifuged for 10 min at 4 C and stored at –70 C. For each condition, a 100-μl aliquot was diluted with 0.9 ml of buffer [0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl (pH 8.1), 167 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, and 0.1% protease inhibitor mixture for mammalian cells (Sigma)] and precleared with 60 μl of BSA-blocked protein A/G-agarose (Santa Cruz Biotechnology) for 1 h at 4 C. After removal of the agarose beads by centrifugation, either 1 μl of antiserum, 10 μg of normal rabbit IgG (Santa Cruz Biotechnology), or no antibody was added to the supernatant, and the mixture was incubated for 1 h at 4 C. Specific rabbit polyclonal antibodies directed against either PDX-1 (23) or ISL-1 were used in this study (Abcam, Cambridge, UK). Antibody-protein-DNA complexes were isolated by incubation with 60 μl of blocked protein A/G-agarose for 3 h at 4 C. After extensive washing, bound DNA fragments were eluted with 300 μl of elution buffer (50 mM NaHCO3, 1% SDS) and analyzed by PCR using PCR Master (Roche, Mannheim, Germany), 15 pmol of each primer, and 10 μl of the immunoprecipitated DNA per reaction. Cycling parameters were as follows: one cycle of 95 C for 2 min and 28 cycles of 95 C for 30 sec, 61 C for 30 sec, and 72 C for 30 sec. The primers used for amplification of the GIP promoter fragment bound by ISL-1 and PDX-1 were 5'-CCTTCTGTTCCCTCAGAAG-3' and 5'-GTTGGTTCTCAGGATCTTGC-3'. Amplified products were electrophoresed through a 1.4% agarose gel in Tris acetate/EDTA buffer and visualized after ethidium bromide staining.
Plasmids
The GIP/Luc reporter constructs used in this paper are described in detail elsewhere (15, 16). To make the plasmid pGL-193D, the wild-type PDX-1 sequence between base pairs –156 and –151 of the GIP promoter was mutated from CATTAG to GAAAAG. The PDX-1 expression plasmid was generated by cloning the PDX-1 cDNA into the Kpn1-Not1 site of the eukaryotic expression vector pcDNA3.1/Zeo (Invitrogen Corp., Carlsbad, CA).
Transient transfection assays
One day before transfection, cells were plated in the appropriate growth media onto six-well plates at a density of approximately 1–2 x 105 cells per well. A mixture containing 0.5 μg of the derivative of a pGL2 reporter plasmid containing the GIP sequences, 4.5 μl lipofectamine (Invitrogen), 16 ng pRL-cytomegalovirus (included to control for transfection efficiency) (Promega, Madison, WI), 0.5 μg of PDX-1/pcDNA3.1 or pcDNA3.1 vector, and 600 μl serum-free media were incubated at RT. After 15 min, 0.8 ml of medium was added, and the mixture was added to cells previously washed twice with serum-free medium. After 5 h, 1 ml of medium containing 20% serum was added, and the incubation was continued for 48 h. For analysis, the cells were washed twice with PBS, lysed in 400 μl lysis buffer, and subjected to two freeze/thaw cycles, using the manufacturer’s instructions for the Dual-luciferase reporter assay system (Promega). To measure the luciferase and renilla activities, 20 μl of each sample were measured in duplicate using an Optocomp 1 luminometer (MJ Research Inc., Waltham, MA). Luciferase activity is expressed in light units and was corrected for renilla activity to compensate for variations in transfection efficiency. All constructs were analyzed in triplicate in at least three separate experiments ± SE. Data were analyzed using Student’s t test for unpaired samples. Statistical significance was assigned for P < 0.05.
Results
Expression of PDX-1 in native K-cells
Interspersed among the mucosal enterocytes, the mammalian epithelium contains enteroendocrine cells, including GIP-producing K-cells. The K-cells are sparsely distributed within the mucosa, and in the dog, K-cells account for less than 1 in 1000 of the epithelial cell population (28). To demonstrate the expression of PDX-1 in GIP-producing K-cells, we analyzed, by immunohistochemistry, intestinal mucosal tissue from 18.5-dpc embryos (Fig. 1, A–D) and 10-wk-old mice (Fig. 1, E–H) for the expression of both GIP and PDX-1. Late-gestation embryos were examined because it has previously been determined that enteroendocrine cells can be detected more readily at this stage of development (20). Between 16 and 18 dpc, the intestine differentiates into a columnar epithelium with villi, and neuroendocrine expression is up-regulated in anticipation of feeding (20, 29). In addition, because PDX-1 is required for normal rostral patterning, it is highly expressed by all epithelial cells in the rostral duodenum region at embryonic d 16.5 (E16.5), after which it becomes progressively down-regulated in cells of nonendocrine lineage. Therefore, at E18.5 we would anticipate seeing PDX-1 expression in a large percentage of epithelial cells. To determine whether PDX-1 continues to be expressed in K-cells after birth, we also analyzed the expression pattern of GIP and PDX-1 in 10-wk-old mice.
FIG. 1. Double staining for GIP-1 and PDX-1 in mouse duodenum demonstrates PDX-1 expression in intestinal K-cells. Tissue sections from wild-type 18.5-dpc mice (A–D), and 10-wk-old mice (E–H) were probed with rabbit anti-GIP antibody (red, C and G) and guinea pig anti-PDX-1 antibody (blue, B and F). Nuclei (green, A and E) were counterstained with YoPro-1 (Molecular Probes). Merged pictures are shown in D (for A–C) and H (for E–G). White arrows indicate the location of a PDX and GIP double-positive cell. Green arrows indicate a PDX-1-positive, GIP-negative cell. Note that in F the blue spots located outside the epithelial layer are auto-fluorescent signals from erythrocytes. White arrowheads indicate three examples.
As illustrated in Fig. 1, PDX-1-positive cells (blue) were found both in sections obtained from wild-type E18.5 embryos (B) and in sections from 10-wk-old mice (F) and were restricted to the epithelial cells with the enteroendocrine cells staining more intensely positive. Cells in the submucosal and muscle layers were negative for PDX-1. In Fig. 1F, the blue spots seen outside the epithelial layer are the result of autofluorescent signals from erythrocytes and are indicated by white arrowheads. Nuclei were counterstained with YoPro-1 and are shown in green (Fig. 1, A and E), whereas GIP-positive cells (C and G) are indicated in red. Merged images for E18.5 and wk 10 are shown in Fig. 1, D (for A–C) and H (for E–G), respectively. The white arrows point to PDX-1-positive cells that also stain for GIP immunoreactivity. The green arrow indicates a PDX-1-positive cell that is negative for GIP. Examples of GIP-positive cells that were negative for PDX-1 were not detected.
GIP expression in PDX-1 null mice
Previously, analysis of PDX-1 homozygous null embryos demonstrated that an absence of PDX-1 expression during development resulted in the formation of an abnormal rostral duodenum and a reduction in the number of enteroendocrine cells present in the rostral duodenum (20, 30). We analyzed tissues from 18.5-dpc wild-type (+/+) and PDX-1XBko homozygous null (–/–) embryos for the distribution and number of GIP-producing K-cells by immunohistochemistry. Duodenal sections from three mice from each group were examined by confocal microscopy. In Fig. 2, two representative sections demonstrate the presence of GIP immunoreactivity (red) in wild-type mice (A) but not in PDX-1 null mice (B). In addition, mice were examined for the presence of serotonin-expressing cells. For each animal, five nonadjacent sections from 18.5 dpc were evaluated for GIP- and serotonin-positive cells. Figure 2C depicts the results obtained from the examination of three wild-type (black bar) and three PDX-1–/– (white bar) mice and are represented as the number of GIP- or serotonin-positive cells per 1000 epithelial nuclei. Our analysis demonstrated a 97.8% reduction in GIP-positive cells in the PDX-1 null embryos, compared with wild-type embryos (P = 0.005). A few GIP-positive cells were detected in the relatively distal portion of duodenum of the PDX-1 null embryos, with two thirds located within the crypt region and the other one third located within the lower half of the villi. GIP-positive cells were not detected in the proximal portion of the duodenum in the PDX-1 null embryos. In contrast, differences between the number of serotonin-positive cells detected in wild-type and PDX-1–/– mice (a 28% reduction in the latter) did not reach statistical significance (P > 0.05; data not shown).
FIG. 2. GIP-positive cell number is reduced by 97.8% in the duodenum of PDX-1 null mutation mice. Mouse duodenal sections from 18.5-dpc wild-type (A) and PDX1–/– mice (B) were stained for the presence of GIP immunoreactivity (red) using a rabbit polyclonal antibody. GIP immunoreactivity was detected in wild-type mice but not in PDX-1 null, mice. C, Quantification of GIP-positive and serotonin-positive cells. Five nonadjacent sections from 18.5-dpc animals were counted for GIP- and serotonin-positive cells. The figure depicts the results obtained from the examination of three wild-type (black bar) and three PDX-1–/– (white bar) mice. The vertical axis represents the number of GIP- and serotonin-positive cells per 1000 epithelial nuclei. Data are expressed as the mean ± SE (n = 3). The reduction in the number of GIP-positive cells, but not in serotonin-positive cells, in PDX-1–/– mice was significant; *, P = 0.005.
PDX-1 expression in STC-1 cells
To determine whether the loss of GIP expression in the PDX-1 null mice occurs as a result of a direct interaction between PDX-1 and the GIP promoter region, we examined GIP/PDX-1 interactions in STC-1 cells. Initially, Northern blot analysis was used to confirm the presence of PDX-1 transcripts in STC-1 cells (Fig. 3A). Using a PDX-1 cDNA probe, we demonstrated the presence of a 2.3-kb transcript in STC-1 cells but not in control IEC-6 cells. In addition, we detected the presence of a larger 7-kb transcript in the STC-1 cell RNA extract. The presence of this larger transcript has previously been detected in HIT cells, ?TC6 cells, and isolated mouse and rat islets (31). Northern analysis using a GIP-specific probe confirmed the presence of GIP transcripts in STC-1 cells.
FIG. 3. PDX-1 is expressed in STC-1 cells. A, Northern blot analysis detects the presence of PDX-1 transcripts in STC-1 cells. Total RNA was extracted from STC-1 and IEC-6 cells, as described in Materials and Methods. After electrophoresis through a 1.0% agarose gel, the RNA was transferred to a nylon membrane, hybridized to a 1.4-kb PDX-1 cDNA probe, and exposed overnight to x-ray film (top). The blot was washed and then probed with a GIP-specific probe. As a positive control for RNA loading, the blot was reprobed with a ?-actin-specific cDNA probe. The blot shown is representative of three independent experiments. B, Western blot analysis demonstrates PDX-1 immunoreactivity in STC-1 cells. Total cell lysates from STC-1 and IEC-6 cells were resolved by SDS-PAGE and transferred to a nylon membrane before incubation with antiserum directed against PDX-1 (top). After washing, the blot was incubated with a secondary antibody conjugated with horseradish peroxidase, subsequently washed, and incubated with chemiluminescence reagents before being exposed to x-ray film. The blot was washed and then probed with an antibody directed against ?-actin (bottom) as a control for loading. Similar results were obtained in four independent experiments. C, Immunohistochemical analysis demonstrates the presence of PDX-1 in the nucleus. STC-1 cells, grown on chamber slides, were fixed as described in Materials and Methods and incubated with antiserum directed against either PDX-1 or GIP. After washing, cells were incubated with donkey antirabbit fluorescein isothiocyanate-conjugated secondary antibody, mounted, and photographed using standard fluorescence microscopy.
IEC-6 cells, an immature intestinal stem cell line derived from normal rat small intestine that displays characteristics of immature intestinal crypt cells, were used as a negative control. IEC-6 cells exhibit undifferentiated morphology, have limited expression of intestinal cell-specific genes, and have been shown to possess the ability to differentiate into numerous cell types, including other enteroendocrine cells.
Western blot analysis was next used to demonstrate the presence of PDX-1 protein in STC-1 cells. When whole-cell lysate from IEC-6 and STC-1 cells were probed with antibodies directed against the N terminus of PDX-1, a 46-kDa band was detected in STC-1 cells but not in IEC-6 cells (Fig. 3B). Incubation of the blot with an antibody directed against ?-actin confirmed the presence of protein in both lanes. The cellular location of PDX-1 protein in STC-1 cells was determined by immunohistochemical analysis (Fig. 3C). When formalin-fixed STC-1 cells were incubated with antibodies directed against PDX-1 and GIP, PDX-1 immunoreactivity was detected primarily in the nucleus, whereas GIP immunoreactivity was identified in the cytoplasm.
PDX-1 binds to the CATTA region in the GIP promoter
The region located between –156 and –151 of the GIP promoter was previously shown to be important for GIP expression in STC-1 cells (16). Mutation of the sequence CATTA to gAaaA resulted in an 85% reduction of promoter activity, as assayed by transient transfection assays in STC-1 cells. In addition, EMSAs demonstrated specific binding to this region and identified the transcription factor ISL-1 as a component of the DNA-protein complex. In the present study, supershift assays using a PDX-1-specific antibody demonstrated that PDX-1 also binds to this region (Fig. 4). Nuclear extract from STC-1 cells was mixed with a 32P-labeled double-stranded oligonucleotide corresponding to the region spanning from –160 to –141 of the GIP gene in the absence of specific antiserum (lane 1) or in the presence of antiserum directed against GATA-4 (lane 2), ISL-1 (lane 3), PDX-1 (lane 4), or a combination of ISL1 and PDX-1 (lane 5). Both ISL-1 and PDX-1 antisera were able to supershift the complex, indicating that both proteins interact with the corresponding regulatory element in vitro.
FIG. 4. PDX-1 binds to the CATTA element in the GIP promoter. EMSAs were performed with labeled 20-bp double-stranded oligonucleotides containing sequences spanning –140 to –120 of the GIP gene. The probe was incubated with 2.5 μg of STC-1 cell nuclear protein in the absence of antiserum (lane 1) or presence of 1 μl of GATA-4 antibody (lane 2), 1 μl of ISL-1 antibody (lane 3), 1 μl PDX-1 antibody (lane 4), or 1 μl both PDX-1 and ISL-1 (lane 5). The black arrow denotes the supershifted complexes.
PDX-1 binds to control region of the GIP promoter in an intact cell
To demonstrate that PDX-1 binds to the endogenous GIP gene, a ChIP analysis was performed using formaldehyde cross-linked chromatin isolated from STC-1 cells. After precipitation of chromatin in the presence of 1 or 2 μl anti-PDX-1 antiserum or 10 μg normal rabbit antiserum or in the absence of antiserum, the GIP 5'-flanking sequence spanning nucleotides –461 to –11 was selectively amplified by PCR. As shown in Fig. 5, only the samples treated with PDX-1 antiserum yielded a PCR product. As expected, negative results were obtained in immunoprecipitation reactions in which chromatin was omitted. In addition, the PDX-1 antiserum did not immunoprecipitate regulatory sequences from the phosphoenolpyruvate carboxykinase (PEPCK) gene, a gene that contains a functional PDX-1 binding site but is not transcriptionally active in STC-1 cells. These results demonstrate that PDX-1 binds to the GIP promoter region in intact STC-1 cells. Similar results were obtained using the ISL-1 antiserum (data not shown).
FIG. 5. PDX-1 binds to the GIP gene control sequences in intact neuroendocrine cells. Formaldehyde cross-linked chromatin from STC-1 cells was incubated with 1 μl (lane 3) or 2 μl (lane 4) of an antiserum raised against the N terminus of PDX-1. Immunoprecipitated DNA was analyzed by PCR with primers to transcriptional regulatory sequences of the mouse GIP and PEPCK genes. As controls, PCRs were performed with DNA immunoprecipitated with normal rabbit immune sera (lane 2), input DNA (1:100 dilution; lane 5), DNA immunoprecipitated in the absence of antiserum (lane 6), chromatin minus sample immunoprecipitated with 2 μl PDX-1 antiserum (lane 7), and no template (lane 1). The blot shown is representative of three independent experiments.
PDX-1 overexpression increases the activity of the GIP promoter
To demonstrate the functional relevance of our findings, we next examined the effects of PDX-1 protein overexpression on GIP promoter activity in STC-1 cells (Fig. 6). Transient transfection analyses of various GIP-Luc reporter constructs with either a PDX1/pcDNA3.1 expression plasmid or pcDNA3.1 empty vector were conducted. PDX-1 overexpression caused an increase in the luciferase activity of all GIP promoter constructs containing the PDX-1 binding site (pGL-2569, pGL-193, and pGL-173). In contrast, PDX-1 expression did not affect the activity of a GIP-Luc construct containing a mutation in the PDX-1 binding site (pGL-193D) or the promoterless construct, pGL2 basic. These observations further support the importance of the transcription factor PDX-1 in the regulation of the GIP gene.
FIG. 6. Overexpression of PDX-1 leads to a specific increase in activity of GIP/Luc reporter constructs. The effect that the overexpression of PDX-1 on the functional activity of GIP/Luc fusion constructs was assessed in STC-1 cells. Cotransfections were performed using pRL-cytomegalovirus plus GIP/Luc promoter constructs containing various lengths of the GIP promoter (pGL-2569, pGL-193, and pGL-173) or a clone containing a mutation in the PDX-1 binding site (pGL-193D) and either the expression plasmid PDX-1/pcDNA3.1 or the empty vector (pcDNA3.1). After 48 h, the cells were harvested and analyzed for luciferase and renilla activity, as described in Materials and Methods. The data represent mean activity ± SE of six independent transfections normalized to the activity of pGL2 basic, after correcting for differences in transfection efficiencies determined by associated renilla activity.
Discussion
Previous studies from our laboratory have established the importance of the sequence between base pairs –156 and –151 of the GIP promoter for transcription of the GIP gene in STC-1 cells and have demonstrated the in vitro binding of ISL-1 to this region (16). ISL-1, a member of a family of LIM-homeodomain genes (32), has been implicated in the transcriptional regulation of such genes as proglucagon (33), somatostatin (34), and amylin (35) and has been determined to be critical for normal pancreatic development and endocrine cell formation (36).
In this report, we show that PDX-1, a transcription factor that is capable of binding to the same DNA consensus motif as ISL-1, is present in STC-1 cells and also interacts with the region between –156 and –151 of the GIP promoter in vitro. Gel shift assays using antisera directed against ISL-1 and PDX-1 demonstrated that both transcription factors were capable of binding to the cis-regulatory element. The ability of both PDX-1 and a LIM-homeodomain family member to bind to and activate transcription from the same cis-regulatory region has been described previously. Both PDX-1 and the LIM-homeodomain family member Lmx1.1 have been reported to possess the capacity to bind to the A3 and A4 sites of the rat insulin promoter and activate transcription of the insulin gene (37, 38, 39, 40).
Although in vitro binding assays and transient transfection assays with reporter constructs provide evidence for the involvement of a factor in the transcription of a gene, what occurs in these artificial systems may not necessarily reflect what is occurring in an intact cell. In the nucleus, transcriptional activation is affected by endogenous chromatin content (41). Closed chromatin structure or competition from another DNA binding protein with similar nuclear specificity may exclude transcription factors from a potential binding site. This possibility is particularly relevant for transcription factors such as PDX-1, which recognize fairly nonselective consensus sequences within promoter regions. Thus, the ability of PDX-1 to interact with these potential binding sites in vitro may not directly equate to the involvement of this transcription factor in the regulation of the gene. For example, when examining PDX-1 binding by in vitro assays using nuclear extract from ?TC-3 cells, Cissell et al. (27) found that PDX-1 was able to bind to TAAT motifs in both the PEPCK and insulin promoters. In contrast, when these binding interactions were examined using ChIP analysis, PDX-1 was shown to interact with the endogenous insulin, but not the PEPCK, gene (27). This result suggests that PDX-1 interacts with the insulin gene and not the PEPCK gene in the context of chromatin. In the present study, using ChIP analysis, we were likewise able to confirm our in vitro findings and establish a direct interaction between the transcription factor PDX-1 and the GIP gene.
PDX-1, a transcription factor originally discovered as an activator of insulin and somatostatin, has proven to be a key factor in both pancreatic development and adult islet ?-cell function (20, 42, 43, 44, 45, 46, 47). Analysis of PDX-1 knockout mice demonstrated that animals lacking PDX-1 exhibit pancreatic agenesis (20, 48), resulting in extreme hyperglycemia and perinatal death. In addition, Ahlgren et al. (49), using the Cre-LoxP system to produce cell-type gene inactivation in ?-cells, determined that continued PDX-1 expression in the adult mouse is required for the expression of various ?-cell genes, including insulin, GLUT2, and IAPP. Furthermore, functional activity of PDX-1 is not limited to the pancreas. Offield et al. (20) reported that in addition to pancreatic abnormalities, null mutation of the PDX-1 gene resulted in an altered architecture of the rostral duodenum, an absence of Brunner’s glands, and a reduction in enteroendocrine cells in the villous epithelium.
Guz et al. (45) examined the developmental expression pattern of PDX-1 in the gut and reported the presence of PDX-1-positive cells in the endodermal layer of the dorsal region of the gut beginning at E8.5. They also found that by E9.5, PDX-1 was expressed in most of the epithelial cells within the mucosal layer but not in nearby connective tissues. When older embryos and adults were examined, PDX-1 expression persisted in almost all the epithelial cells lining the villi but was not detected within the crypts. In addition, PDX-1 expression was found neither in other mucosal cells nor in cells comprising the submucosa, muscularis, or adventitial layers of the duodenal wall. As a result of their observations, the authors speculated that PDX-1 may play a critical role in duodenal epithelial cell function.
In the present study, we provide evidence for the involvement of PDX-1 in the regulation of GIP expression within the enteroendocrine K-cells of the proximal duodenum, and we have demonstrated that GIP and PDX-1 are both expressed in the K-cells of the duodenum. In addition, after a careful analysis of GIP expression in both PDX-1 wild-type and null mice, we found that the loss of PDX-1 expression nearly abolished GIP-producing cells in the gut. Previously, homozygosity for mutations in the PDX-1 gene in mice was reported to affect the enteroendocrine cells in the duodenum (20, 30, 48). Specifically, Offield et al. (20) reported a general reduction in the number of enteroendocrine cells in PDX-1 null mice but did not specifically measure GIP-expressing cells. In addition, Larsson et al. (30) reported finding a 72% reduction in K-cells. The reason for the discrepancy between our findings and their report is unclear but may be because of methodological differences, including antibody specificity. Another important variable might be the technique used to determine the distribution and extent of GIP-expressing cells. Although we did detect a very small number of GIP-expressing cells (six GIP-positive cells of >70,000 epithelial cells) in the PDX-1 null mice, they were located predominately in the very distal duodenum, outside the boundaries of PDX-1 expression. Because these cells appear to be restricted to the crypt region or the lower half of the villi, it is possible they represent undifferentiated progenitor cells. However, because so few cells were detected, we cannot definitively conclude that they are restricted to the lower half of the villi. Nonetheless, it is important to note that no GIP-positive cells were detected in the proximal portion of the duodenum in the PDX-1 null embryos. This finding, along with the 97.8% reduction in GIP-positive cells in PDX-1 null mutants, clearly indicates a primary and nonredundant role for PDX-1 in cell differentiation for most, if not all, GIP-producing cells.
It is not surprising that the GIP and insulin genes are regulated by common transcription factors. GIP and insulin are both hormones involved in nutrient storage, and GIP-producing intestinal K-cells and insulin-producing pancreatic ?-cells share a common developmental lineage. Intestinal K-cells are derived from stem cells located in the intestinal crypts, where, like other neuroendocrine tissue such as the pancreatic islets, they are thought to originate developmentally from the primitive endoderm. Interestingly, PDX-1 has been shown to be required for both the regionalization of the primitive gut endoderm and the maturation of the pancreatic ?-cells (47, 50).
It is not known whether PDX-1 plays a role in the process of stem cell differentiation into GIP-producing K-cells. However, it seems likely that the use of a transcription factor to control both the spatial and temporal differentiation of a specific endocrine cell type and the main hormone that defines the cell would be advantageous from an evolutionary standpoint. As discussed above, PDX-1 seems to support this contention with respect to ?-cell differentiation and insulin gene regulation. Therefore, if PDX-1 plays a similar role in the K-cell, it is possible that the decreased presence of GIP-producing cells we observed in both embryonic and adult PDX-1 null mice may result not only from a decreased rate of transcription of the GIP gene but also from a decreased number of K-cells. Because GIP presently represents the only known K-cell marker, future experiments will be necessary to enable us to identify additional K-cell markers whose transcription might occur independently of PDX-1 to examine this possibility.
References
Brown JC, Dryburgh JR 1971 A gastric inhibitory polypeptide. II. The complete amino acid sequence. Can J Biochem 49:867–872
Buchan AM, Polak JM, Capella C, Solcia E, Pearse AG 1978 Electronimmunocytochemical evidence for the K cell localization of gastric inhibitory polypeptide (GIP) in man. Histochemistry 56:37–44
Pederson RA, Schubert HE, Brown JC 1975 Gastric inhibitory polypeptide: its physiologic release and insulinotropic action in the dog. Diabetes 24:1050–1056
Cataland S, Crockett SE, Brown JC, Mazzaferri EL 1974 Gastric inhibitory polypeptide (GIP) stimulation by oral glucose in man. J Clin Endocrinol Metab 39:223–228
Buffa R, Polak JM, Pearse AG, Solcia E, Grimelius L, Capella C 1975 Identification of the intestinal cell storing gastric inhibitory peptide. Histochemistry 43:249–255
Cheung AT, Dayanandan B, Lewis JT, Korbutt GS, Rajotte RV, Bryer-Ash M, Boylan MO, Wolfe MM, Kieffer TJ 2000 Glucose-dependent insulin release from genetically engineered K cells. Science 290:1959–1962
Nauck MA, Holst JJ, Willms B, Schmiegel W 1997 Glucagon-like peptide 1 (GLP-1) as a new therapeutic approach for type 2-diabetes. Exp Clin Endocrinol Diabetes 105:187–195
Gutniak M, Orskov C, Holst JJ, Ahren B, Efendic S 1992 Antidiabetogenic effect of glucagon-like peptide-1 (7–36)amide in normal subjects and patients with diabetes mellitus. N Engl J Med 326:1316–1322
Gault VA, O’Harte FP, Harriott P, Mooney MH, Green BD, Flatt PR 2003 Effects of the novel (Pro(3))GIP antagonist and exendin(9–39)amide on GIP- and GLP-1-induced cyclic AMP generation, insulin secretion and postprandial insulin release in obese diabetic (ob/ob) mice: evidence that GIP is the major physiological incretin. Diabetologia 46:222–230
Tseng CC, Kieffer TJ, Jarboe LA, Usdin TB, Wolfe MM 1996 Postprandial stimulation of insulin release by glucose-dependent insulinotropic polypeptide (GIP): effect of a specific glucose-dependent insulinotropic polypeptide receptor antagonist in the rat. J Clin Invest 98:2440–2445
Miyawaki K, Yamada Y, Ban N, Ihara Y, Tsukiyama K, Zhou H, Fujimoto S, Oku A, Tsuda K, Toyokuni S, Hiai H, Mizunoya W, Fushiki T, Holst JJ, Makino M, Tashita A, Kobara Y, Tsubamoto Y, Jinnouchi T, Jomori T, Seino Y 2002 Inhibition of gastric inhibitory polypeptide signaling prevents obesity. Nat Med 8:738–742
Takeda J, Seino Y, Tanaka K, Fukumoto H, Kayano T, Takahashi H, Mitani T, Kurono M, Suzuki T, Tobe T 1987 Sequence of an intestinal cDNA encoding human gastric inhibitory polypeptide precursor. Proc Natl Acad Sci USA 84:7005–7008
Tseng CC, Jarboe LA, Landau SB, Williams EK, Wolfe MM 1993 Glucose-dependent insulinotropic peptide: structure of the precursor and tissue-specific expression in rat. Proc Natl Acad Sci USA 90:1992–1996
Higashimoto Y, Liddle RA 1993 Isolation and characterization of the gene encoding rat glucose-dependent insulinotropic peptide. Biochem Biophys Res Commun 193:182–190
Boylan MO, Jepeal LI, Jarboe LA, Wolfe MM 1997 Cell-specific expression of the glucose-dependent insulinotropic polypeptide gene in a mouse neuroendocrine tumor cell line. J Biol Chem 272:17438–17443
Jepeal LI, Boylan MO, Wolfe MM 2003 Cell-specific expression of the glucose-dependent insulinotropic polypeptide gene functions through a GATA and an ISL-1 motif in a mouse neuroendocrine tumor cell line. Regul Pept 113:139–147
Orkin SH 1992 GATA-binding transcription factors in hematopoietic cells. Blood 80:575–581
Evans T, Reitman M, Felsenfeld G 1988 An erythrocyte-specific DNA-binding factor recognizes a regulatory sequence common to all chicken globin genes. Proc Natl Acad Sci USA 85:5976–5980
Inagaki N, Seino Y, Takeda J, Yano H, Yamada Y, Bell GI, Eddy RL, Fukushima Y, Byers MG, Shows TB 1989 Gastric inhibitory polypeptide: structure and chromosomal localization of the human gene. Mol Endocrinol 3:1014–1021
Offield MF, Jetton TL, Labosky PA, Ray M, Stein RW, Magnuson MA, Hogan BL, Wright CV 1996 PDX-1 is required for pancreatic outgrowth and differentiation of the rostral duodenum. Development 122:983–995
Kitamura T, Nakae J, Kitamura Y, Kido Y, Biggs 3rd WH, Wright CV, White MF, Arden KC, Accili D 2002 The forkhead transcription factor Foxo1 links insulin signaling to Pdx1 regulation of pancreatic ?-cell growth. J Clin Invest 110:1839–1847
Jenny M, Uhl C, Roche C, Duluc I, Guillermin V, Guillemot F, Jensen J, Kedinger M, Gradwohl G 2002 Neurogenin3 is differentially required for endocrine cell fate specification in the intestinal and gastric epithelium. EMBO J 21:6338–6347
Brissova M, Shiota M, Nicholson WE, Gannon M, Knobel SM, Piston DW, Wright CV, Powers AC 2002 Reduction in pancreatic transcription factor PDX-1 impairs glucose-stimulated insulin secretion. J Biol Chem 277:11225–11232
Wolfe MM, Hocking MP, Maico DG, McGuigan JE 1983 Effects of antibodies to gastric inhibitory peptide on gastric acid secretion and gastrin release in the dog. Gastroenterology 84:941–948
Dignam JD, Lebovitz RM, Roeder RG 1983 Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic Acids Res 11:1475–1489
Gerrish K, Cissell MA, Stein R 2001 The role of hepatic nuclear factor 1 and PDX-1 in transcriptional regulation of the pdx-1 gene. J Biol Chem 276:47775–47784
Cissell MA, Zhao L, Sussel L, Henderson E, Stein R 2003 Transcription factor occupancy of the insulin gene in vivo: evidence for direct regulation by Nkx2.2. J Biol Chem 278:751–756
Kieffer TJ, Buchan AM, Barker H, Brown JC, Pederson RA 1994 Release of gastric inhibitory polypeptide from cultured canine endocrine cells. Am J Physiol 267:E489–E496
Roth KA, Rubin DC, Birkenmeier EH, Gordon JI 1991 Expression of liver fatty acid-binding protein/human growth hormone fusion genes within the enterocyte and enteroendocrine cell populations of fetal transgenic mice. J Biol Chem 266:5949–5954
Larsson LI, Madsen OD, Serup P, Jonsson J, Edlund H 1996 Pancreatic-duodenal homeobox 1: role in gastric endocrine patterning. Mech Dev 60:175–184
Olson LK, Qian J, Poitout V 1998 Glucose rapidly and reversibly decreases INS-1 cell insulin gene transcription via decrements in STF-1 and C1 activator transcription factor activity. Mol Endocrinol 12:207–219
Schmeichel KL, Beckerle MC 1994 The LIM domain is a modular protein-binding interface. Cell 79:211–219
Wang M, Drucker DJ 1995 The LIM domain homeobox gene isl-1 is a positive regulator of islet cell-specific proglucagon gene transcription. J Biol Chem 270:12646–12652
Vallejo M, Penchuk L, Habener JF 1992 Somatostatin gene upstream enhancer element activated by a protein complex consisting of CREB, Isl-1-like, and -CBF-like transcription factors. J Biol Chem 267:12876–12884
Wang M, Drucker DJ 1996 Activation of amylin gene transcription by LIM domain homeobox gene isl-1. Mol Endocrinol 10:243–251
Ahlgren U, Pfaff SL, Jessell TM, Edlund T, Edlund H 1997 Independent requirement for ISL1 in formation of pancreatic mesenchyme and islet cells. Nature 385:257–260
Ohneda K, Mirmira RG, Wang J, Johnson JD, German MS 2000 The homeodomain of PDX-1 mediates multiple protein-protein interactions in the formation of a transcriptional activation complex on the insulin promoter. Mol Cell Biol 20:900–911
Peers B, Leonard J, Sharma S, Teitelman G, Montminy MR 1994 Insulin expression in pancreatic islet cells relies on cooperative interactions between the helix loop helix factor E47 and the homeobox factor STF-1. Mol Endocrinol 8:1798–1806
Peshavaria M, Henderson E, Sharma A, Wright CV, Stein R 1997 Functional characterization of the transactivation properties of the PDX-1 homeodomain protein. Mol Cell Biol 17:3987–3996
German MS, Wang J, Chadwick RB, Rutter WJ 1992 Synergistic activation of the insulin gene by a LIM-homeo domain protein and a basic helix-loop-helix protein: building a functional insulin minienhancer complex. Genes Dev 6:2165–2176
Ozcan S 2003 Diabetes mellitus methods and protocols. Totowa, NJ: Humana Press
Habener JF, Stoffers DA 1998 A newly discovered role of transcription factors involved in pancreas development and the pathogenesis of diabetes mellitus. Proc Assoc Am Physicians 110:12–21
Edlund H 1998 Transcribing pancreas. Diabetes 47:1817–1823
Jonsson J, Ahlgren U, Edlund T, Edlund H 1995 IPF1, a homeodomain protein with a dual function in pancreas development. Int J Dev Biol 39:789–798
Guz Y, Montminy MR, Stein R, Leonard J, Gamer LW, Wright CV, Teitelman G 1995 Expression of murine STF-1, a putative insulin gene transcription factor, in ?-cells of pancreas, duodenal epithelium and pancreatic exocrine and endocrine progenitors during ontogeny. Development 121:11–18
Leonard J, Peers B, Johnson T, Ferreri K, Lee S, Montminy MR 1993 Characterization of somatostatin transactivating factor-1, a novel homeobox factor that stimulates somatostatin expression in pancreatic islet cells. Mol Endocrinol 7:1275–1283
Ahlgren U, Jonsson J, Edlund H 1996 The morphogenesis of the pancreatic mesenchyme is uncoupled from that of the pancreatic epithelium in IPF1/PDX1-deficient mice. Development 122:1409–1416
Jonsson J, Carlsson L, Edlund T, Edlund H 1994 Insulin-promoter-factor 1 is required for pancreas development in mice. Nature 371:606–609
Ahlgren U, Jonsson J, Jonsson L, Simu K, Edlund H 1998 ?-Cell-specific inactivation of the mouse Ipf1/Pdx1 gene results in loss of the ?-cell phenotype and maturity onset diabetes. Genes Dev 12:1763–1768
Ohlsson H, Karlsson K, Edlund T 1993 IPF1, a homeodomain-containing transactivator of the insulin gene. EMBO J 12:4251–4259(Lisa I. Jepeal, Yoshio Fu)
Address all correspondence and requests for reprints to: M. Michael Wolfe, M.D., Section of Gastroenterology, Boston Medical Center, 650 Albany Street, Boston, Massachusetts 02118. E-mail: michael.wolfe@bmc.org.
Abstract
Glucose-dependent insulinotropic polypeptide (GIP) is a potent stimulator of insulin secretion and comprises an important component of the enteroinsular axis. GIP is synthesized in enteroendocrine K-cells located principally in the upper small intestine. The homeobox-containing gene PDX-1 is also expressed in the small intestine and plays a critical role in pancreatic development and in the expression of pancreatic-specific genes. Previous studies determined that the transcription factors GATA-4 and ISL-1 are important for GIP expression. In this study, we demonstrate that PDX-1 is also involved in regulating GIP expression in K-cells. Using immunohistochemistry, we verified the expression of PDX-1 protein in the nucleus of GIP-expressing mouse K-cells and evaluated the expression of PDX-1, serotonin, and GIP in wild-type and PDX-1–/– mice at 18.5 d after conception. Although we demonstrated a 97.8% reduction in the number of GIP-expressing cells in PDX-1–/– mice; there was no statistical difference in the number of serotonin-positive cells. Additionally, PDX-1 transcripts and protein were detected in a GIP-expressing neuroendocrine cell line, STC-1. Electromobility shift assays using STC-1 nuclear extracts demonstrated the specific binding of PDX-1 protein to a specific regulatory region in the GIP promoter. Using chromatin immunoprecipitation analysis, we demonstrated binding of PDX-1 to this same region of the GIP promoter in intact cells. Lastly, overexpression of PDX-1 in transient transfection assays led to a specific increase in the activity of GIP/Luc reporter constructs. The results of these studies indicate that the transcription factor PDX-1 plays a critical role in the cell-specific expression of the GIP gene.
Introduction
GLUCOSE-DEPENDENT INSULINOTROPIC polypeptide (GIP) is a 42-amino-acid peptide that is a member of the secretin-vasoactive intestinal polypeptide family of gastrointestinal regulatory peptides (1). The primary site of endogenous GIP release is the endocrine K-cells dispersed throughout the mucosa of the duodenum and proximal jejunum (2). GIP is released into circulation principally in response to the ingestion of two major nutrient stimuli, carbohydrate and fat (3, 4). Studies have established that the primary function of GIP is to stimulate insulin secretion from the pancreatic ?-cells in the presence of elevated blood glucose (4, 5, 6). Thus, GIP, along with glucagon-like peptide-1 (GLP-1), function as incretins, the proposed mediators of the enteroinsular axis that help maintain glucose homeostasis under physiological conditions. Although GLP-1 has been established as a more potent pharmacological stimulator of insulin release (7, 8), it has recently been demonstrated, using GIP and GLP-1-specific receptor antagonists, that GIP is the major physiological incretin, accounting for approximately 80% of nutrient-induced enteroinsular pancreatic ?-cell stimulation (9, 10).
Recent studies have demonstrated the importance of defining both the regulation of GIP expression and the process of K-cell differentiation. Miyawaki et al. (11) created a GIP-receptor-deficient (GIPR–/–) mouse to investigate the role of GIP in the development of obesity. These investigators demonstrated that although normal littermates became obese and developed insulin resistance and type 2 diabetes mellitus in response to a high-fat diet, GIPR–/– mice were protected and remained normal while consuming an identical diet. The results of this study provide evidence for GIP as an important mediator of nutrient deposition, and it has been proposed that attenuation of the GIP response could be used as a treatment for obesity. It is thus essential to determine the mechanisms controlling GIP expression, secretion, and function.
In a second important study, Cheung et al. (6) engineered a transgenic mouse in which the human insulin gene was expressed under the control of the rat GIP promoter. In these mice, human insulin mRNA was detected only in the duodenum and the stomach, corresponding to the known tissue distribution of GIP. Furthermore, immunohistochemical analysis using antisera specific for GIP and human insulin confirmed that insulin biosynthesis was targeted specifically to the enteroendocrine K-cells of the stomach and duodenum. When native islet ?-cells of these transgenic mice were eliminated by the use of streptozocin, human insulin secreted from intestinal K-cells rendered the mice euglycemic and protected them against the development of diabetes mellitus. These findings demonstrate that K-cells, like pancreatic ?-cells, are capable of processing and storing insulin and of releasing it in a manner that maintains normal glucose homeostasis. Furthermore, they establish the K-cell as a potential cell to exploit in gene therapy for the treatment of diabetes mellitus and demonstrate the effectiveness of the GIP promoter in directing transgene expression to the K-cell.
The important regulatory elements in the GIP promoter that direct cell-specific expression are beginning to be identified and understood. The human and rat GIP genes have been isolated and their promoter sequences partially characterized (12, 13). Although two potential promoter regions were identified upstream of the translation initiation codon, primer extension and RNase protection analysis indicated that GIP transcription initiates from the upstream promoter in the duodenum of adult rats (14). Deletion analysis of the GIP promoter revealed that the first 193 bp upstream of the transcription initiation site were sufficient to direct specific expression of the gene in STC-1 cells (15). Analysis of this region led to the identification of two transcription factors whose binding is necessary for GIP expression in STC-1 cells (15, 16). A binding motif located at base pairs –190 to –184 with respect to the transcription initiation site (+1) of the GIP promoter and conforming to the consensus motif for the GATA family of DNA binding proteins, AGATAA (17, 18, 19), was found to bind the transcription factor GATA-4 (16). A second cis-regulatory region downstream of the GATA-4 binding site was identified that binds the transcription factor ISL-1 and appears to function in conjunction with GATA-4 to promote GIP gene expression (16).
PDX-1, a homeobox-containing gene that can recognize the same DNA consensus motif as ISL-1, has been shown to be important in both the development of the pancreas and the expression of pancreatic-specific genes. In this present study, we demonstrate the presence of PDX-1 in native K-cells and provide evidence to support the integral involvement of PDX-1 in the regulation of GIP expression.
Materials and Methods
Animals
The PDX-1XBko mice used in these studies are described in detail elsewhere (20). Briefly, in this mouse line, the protein-encoding sequences in exon 2 of the PDX-1 gene are deleted, including those encoding the DNA-binding homeodomain, thus creating a functional null PDX-1 allele. Studies for this report were performed on 18.5 d postconception (dpc) PDX-1XBko–/– mice and their age-matched, wild-type littermates.
Cell culture
STC-1 cells, a mouse neuroendocrine cell line (Dr. D. Hanahan, San Francisco, CA), and IEC-6 cells, an immature intestinal stem cell line derived from normal rat small intestine (American Type Culture Collection, Manassas, VA), were grown in DMEM containing 10% fetal bovine serum at 37 C in an atmosphere of 5% CO2 in the presence of 100 U/ml penicillin G, 100 μg/ml streptomycin, and 0.25 μg/ml amphotericin B.
Immunohistochemistry
Embryos were dissected at 18.5 dpc, and tissues were fixed in ice-cold 4% paraformaldehyde at 4 C for 60 min. Tissues were then dehydrated in an increasing ethanol series, which was followed by two Histo-Clear (National Diagnostics, Atlanta, GA) washes, infiltration in Histo-Clear/paraffin (1:1 vol/vol), and three changes of paraffin under vacuum at 56 C. Fixed tissues were embedded and cut into 5-μm sections. Deparaffinized and rehydrated slides were subjected to microwave antigen retrieval in a 10 mM citric acid buffer (pH 6.0) and allowed to cool for 30 min at room temperature (RT). Slides were washed in PBS and then blocked with 5% normal donkey serum for 60 min at RT. The primary antibodies were diluted in PBS containing 1% BSA and incubated with the sections overnight at 4 C. Slides were washed in PBS and incubated with the appropriate secondary antibodies diluted in PBS containing 1% BSA for 1 h at RT. Slides were washed in PBS and then counterstained with the nuclear dye YO-RPO-1 (Molecular Probes, Eugene, OR) (1:1000 dilution in PBS) for 10 min. Slides were washed again in PBS, mounted, and examined using confocal microscopy (Zeiss LSM-510). TIFF images were processed in Adobe Photoshop.
The following primary antibodies were used at the indicated dilutions for immunohistochemistry: guinea pig antibody against PDX-1, 1:5000 (21); and rabbit antibody against GIP (Research Diagnostics, Inc., Flanders, NJ), 1:1000. For immunofluorescence, antirabbit immunoglobulin conjugated to Cy3 (for GIP) and anti-guinea pig immunoglobulin-conjugated Cy5 (for PDX-1) were used as secondary antibodies.
Quantification of GIP- and serotonin-positive cells
Sections of wild-type and PDX-1-deficient mice at 18.5 dpc were stained with rabbit antiserum to GIP or mouse monoclonal antibody against serotonin (Dako, Carpinteria, CA) (22). For the counting of immunoreactive cells, sections were subjected to immunoperoxidase staining. GIP and serotonin stained sections were lightly counterstained with hematoxylin to reveal nuclei and general morphology. Ten random fields of each section were digitally photographed using a Spot camera and printed at x450 magnification to count the number of GIP- or serotonin-positive cells, as well as the total number of epithelial cells. For each animal examined, cell counting was performed on five nonadjacent longitudinal sections of rostral duodenum, thus avoiding the double scoring of cells. In all, three wild-type and three PDX-1–/– mice were examined. The number of immunopositive nucleated cells per 1000 epithelial cells was expressed as the mean ± SE (n = 3) and analyzed statistically using Student’s t test. Statistical significance was assigned if P 0.05.
Western blot analysis
Cells grown on 10-cm plates to 80% confluence were washed once with 1x PBS and then suspended by scraping in 1 ml of RIPA buffer (Boston Bioproducts, Ashland, MA) containing Complete protease inhibitor cocktail (Roche Molecular Biochemicals, Indianapolis, IN) followed by incubation on ice for 15 min. Cell lysates were centrifuged at 14,000 rpm at 4 C for 5 min, and the supernatant was aspirated. Protein concentrations were measured by using the BCA protein assay kit (Pierce, Rockford, IL), and protein lysates (50 μg) were subjected to SDS-PAGE and transferred to Protran nitrocellulose membranes (Perkin Elmer Life Sciences, Boston, MA). Membranes were blocked in 5% nonfat milk, 1x PBS, and 0.5% Tween 20 at RT for 2 h and then incubated with the appropriate primary antibody [rabbit anti-PDX-1, 1 1:500 (23), or mouse anti-?-actin, 1:20,000 (BD Biosciences, Palo Alto, CA)] diluted in 5% nonfat milk in 1x PBS. After a 1-h incubation, membranes were washed three times for 5 min at RT with wash buffer (1x PBS, 0.5% Tween 20) and incubated with the appropriate horseradish peroxidase-conjugated secondary antiserum diluted in 5% nonfat milk in 1x PBS. Membranes were then rinsed three times for 5 min in wash buffer, soaked in chemiluminescence reagent, as instructed by the manufacturer (Super Signal West Pico Chemiluminescent substrate, DuPont, Wilmington, DE), and exposed to x-ray film for 0.1–5 min.
Immunofluorescence cell staining
Cells were plated onto Lab-Tek chamber slides (Nunc International, Naperville, IL) to approximately 25% confluence and grown overnight in normal media. On the following day, the cells were washed once with PBS, fixed by a 10-min incubation in Histochoice tissue fixative (Sigma Chemical Co., St. Louis, IL), and then washed again with PBS for 10 min and permeabilized by a 5-min incubation in 0.1% Triton, 1x PBS. After an additional wash in PBS, cells were blocked in PBS containing 10% donkey serum (Santa Cruz Biotechnology Inc., Santa Cruz, CA) for 20 min, washed with PBS, and incubated for 1 h with primary serum diluted in PBS with 1.5% donkey serum, followed by three washes with PBS for 5 min each. Next, the cells were incubated for 45 min with a Cy3-conjugated secondary antibody diluted 1:200 in PBS, containing 1.5% normal blocking serum. Cells were again washed three times for 5 min with PBS, and a coverslip was mounted using ProLong Antifade (Molecular Probes). Slides were examined and photographed using a fluorescence microscope with appropriate filters. Antibodies to the following proteins were used at the listed concentrations: GIP (24) at 1:400 dilution and PDX-1, rabbit polyclonal, at 1:1000 (23). The secondary antibody used was Cy3-conjugated donkey antirabbit (Jackson ImmunoResearch Lab Inc., West Grove, PA).
Northern blot analysis
Total RNA was extracted from cells grown to 80% confluence using a RNeasy kit, as recommended by the manufacturer (QIAGEN, Valencia, CA). Total RNA (20 μg, quantified by measuring absorbance at 260 nm) was size-fractionated on a 0.8% agarose gel containing 2.2 M formaldehyde and then transferred to a GeneScreen Plus nylon membrane (NEN Life Science Products, Boston, MA) by capillary blotting in 10x saline-sodium citrate buffer (SSC) (pH 7.4). The RNA was cross-linked to the membrane using a UV-Stratalinker (Stratagene, La Jolla, CA), and the blot was prehybridized for 2 h at 42 C in 5x SSC, 10x Denhardt’s solution, 50% formamide, 1% SDS, and 10 μg/ml herring sperm DNA. The filter was then hybridized overnight at 42 C in 5x SSC, 1x Denhardt’s solution, 50% formamide, 20 mM NaPO4, 0.5% SDS, 20 μg/ml herring sperm DNA, and approximately 107 cpm of probe per filter. After hybridization, the blots were washed once at RT in 1x SSC, 1% SDS for 15 min, once at RT in 0.5x SSC, 0.5% SDS for 15 min, twice at RT in 0.1x SSC, 0.1% SDS for 15 min, and once at 50 C in 0.1x SSC, 0.1% SDS for 30 min. Autoradiographs were developed after exposure to x-ray film for 24 h at –70 C, using a Cronex intensifying screen (DuPont).
Electrophoretic mobility shift assays (EMSAs)
Nuclear extracts were prepared according to the method of Dignam et al. (25). The protein concentration was subsequently estimated using a BCA protein assay kit. EMSAs were performed using 4% polyacrylamide gels (44:1 acrylamide/bisacrylamide) in 40 mM Tris/HCl and 195 mM glycine (pH 8.5) at 4 C. The following probe sequence was used: wild type, 5'-GTCACCCATTAGCACAGGCC-3 (presumptive binding site underlined). Each reaction contained 10,000 cpm (7 fmol) of double-stranded oligonucleotide probe, end-labeled by T4 kinase in the presence of [-32P]ATP and purified on a Quik Spin G25 column (Roche Molecular Biochemicals). The reactions were performed in 20 μl of a mixture containing 20 mM HEPES (pH 7.3), 50 mM KCl, 5 mM ?-mercaptoethanol, 20% glycerol, 1 μg poly dI-dC, 1 μg BSA, and 2.5 μg STC-1 cell nuclear extract. The reaction mixtures were incubated for 30 min at RT, followed by 5 min on ice. The samples were separated by electrophoresis for 1.5–2.5 h at 250 V. For supershift experiments, antibodies were preincubated with nuclear extract for 10 min at RT before the addition of the probe. The affinity-purified polyclonal rabbit anti-ISL-1 was purchased from Chemicon (Temecula, CA), and rabbit anti-GATA-4 was purchased from Santa Cruz Biotechnology.
Chromatin immunoprecipitation assays (ChIPs)
ChIPs were performed using the protocol of Gerrish et al. (26) and Cissell et al. (27). STC-1 cells grown on 10-cm plates to 90% confluence were exposed to 1% formaldehyde in DMEM for 5 min at 23 C. Glycine was added to a final concentration of 0.125 M to quench the formaldehyde, and the cultures were incubated for 2 min. The cells were collected in ice-cold PBS, pelleted by centrifugation, and incubated for 10 min on ice in 0.6 ml of SDS lysis buffer [1% SDS, 10 mM EDTA, 50 mM Tris/HCl (pH 8.1), and 1 mM phenylmethylsulfonyl fluoride]. Lysed samples were transferred to prechilled microcentrifuge tubes and subjected to sonication consisting of 12 10-sec pulses. The reactions were centrifuged for 10 min at 4 C and stored at –70 C. For each condition, a 100-μl aliquot was diluted with 0.9 ml of buffer [0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl (pH 8.1), 167 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, and 0.1% protease inhibitor mixture for mammalian cells (Sigma)] and precleared with 60 μl of BSA-blocked protein A/G-agarose (Santa Cruz Biotechnology) for 1 h at 4 C. After removal of the agarose beads by centrifugation, either 1 μl of antiserum, 10 μg of normal rabbit IgG (Santa Cruz Biotechnology), or no antibody was added to the supernatant, and the mixture was incubated for 1 h at 4 C. Specific rabbit polyclonal antibodies directed against either PDX-1 (23) or ISL-1 were used in this study (Abcam, Cambridge, UK). Antibody-protein-DNA complexes were isolated by incubation with 60 μl of blocked protein A/G-agarose for 3 h at 4 C. After extensive washing, bound DNA fragments were eluted with 300 μl of elution buffer (50 mM NaHCO3, 1% SDS) and analyzed by PCR using PCR Master (Roche, Mannheim, Germany), 15 pmol of each primer, and 10 μl of the immunoprecipitated DNA per reaction. Cycling parameters were as follows: one cycle of 95 C for 2 min and 28 cycles of 95 C for 30 sec, 61 C for 30 sec, and 72 C for 30 sec. The primers used for amplification of the GIP promoter fragment bound by ISL-1 and PDX-1 were 5'-CCTTCTGTTCCCTCAGAAG-3' and 5'-GTTGGTTCTCAGGATCTTGC-3'. Amplified products were electrophoresed through a 1.4% agarose gel in Tris acetate/EDTA buffer and visualized after ethidium bromide staining.
Plasmids
The GIP/Luc reporter constructs used in this paper are described in detail elsewhere (15, 16). To make the plasmid pGL-193D, the wild-type PDX-1 sequence between base pairs –156 and –151 of the GIP promoter was mutated from CATTAG to GAAAAG. The PDX-1 expression plasmid was generated by cloning the PDX-1 cDNA into the Kpn1-Not1 site of the eukaryotic expression vector pcDNA3.1/Zeo (Invitrogen Corp., Carlsbad, CA).
Transient transfection assays
One day before transfection, cells were plated in the appropriate growth media onto six-well plates at a density of approximately 1–2 x 105 cells per well. A mixture containing 0.5 μg of the derivative of a pGL2 reporter plasmid containing the GIP sequences, 4.5 μl lipofectamine (Invitrogen), 16 ng pRL-cytomegalovirus (included to control for transfection efficiency) (Promega, Madison, WI), 0.5 μg of PDX-1/pcDNA3.1 or pcDNA3.1 vector, and 600 μl serum-free media were incubated at RT. After 15 min, 0.8 ml of medium was added, and the mixture was added to cells previously washed twice with serum-free medium. After 5 h, 1 ml of medium containing 20% serum was added, and the incubation was continued for 48 h. For analysis, the cells were washed twice with PBS, lysed in 400 μl lysis buffer, and subjected to two freeze/thaw cycles, using the manufacturer’s instructions for the Dual-luciferase reporter assay system (Promega). To measure the luciferase and renilla activities, 20 μl of each sample were measured in duplicate using an Optocomp 1 luminometer (MJ Research Inc., Waltham, MA). Luciferase activity is expressed in light units and was corrected for renilla activity to compensate for variations in transfection efficiency. All constructs were analyzed in triplicate in at least three separate experiments ± SE. Data were analyzed using Student’s t test for unpaired samples. Statistical significance was assigned for P < 0.05.
Results
Expression of PDX-1 in native K-cells
Interspersed among the mucosal enterocytes, the mammalian epithelium contains enteroendocrine cells, including GIP-producing K-cells. The K-cells are sparsely distributed within the mucosa, and in the dog, K-cells account for less than 1 in 1000 of the epithelial cell population (28). To demonstrate the expression of PDX-1 in GIP-producing K-cells, we analyzed, by immunohistochemistry, intestinal mucosal tissue from 18.5-dpc embryos (Fig. 1, A–D) and 10-wk-old mice (Fig. 1, E–H) for the expression of both GIP and PDX-1. Late-gestation embryos were examined because it has previously been determined that enteroendocrine cells can be detected more readily at this stage of development (20). Between 16 and 18 dpc, the intestine differentiates into a columnar epithelium with villi, and neuroendocrine expression is up-regulated in anticipation of feeding (20, 29). In addition, because PDX-1 is required for normal rostral patterning, it is highly expressed by all epithelial cells in the rostral duodenum region at embryonic d 16.5 (E16.5), after which it becomes progressively down-regulated in cells of nonendocrine lineage. Therefore, at E18.5 we would anticipate seeing PDX-1 expression in a large percentage of epithelial cells. To determine whether PDX-1 continues to be expressed in K-cells after birth, we also analyzed the expression pattern of GIP and PDX-1 in 10-wk-old mice.
FIG. 1. Double staining for GIP-1 and PDX-1 in mouse duodenum demonstrates PDX-1 expression in intestinal K-cells. Tissue sections from wild-type 18.5-dpc mice (A–D), and 10-wk-old mice (E–H) were probed with rabbit anti-GIP antibody (red, C and G) and guinea pig anti-PDX-1 antibody (blue, B and F). Nuclei (green, A and E) were counterstained with YoPro-1 (Molecular Probes). Merged pictures are shown in D (for A–C) and H (for E–G). White arrows indicate the location of a PDX and GIP double-positive cell. Green arrows indicate a PDX-1-positive, GIP-negative cell. Note that in F the blue spots located outside the epithelial layer are auto-fluorescent signals from erythrocytes. White arrowheads indicate three examples.
As illustrated in Fig. 1, PDX-1-positive cells (blue) were found both in sections obtained from wild-type E18.5 embryos (B) and in sections from 10-wk-old mice (F) and were restricted to the epithelial cells with the enteroendocrine cells staining more intensely positive. Cells in the submucosal and muscle layers were negative for PDX-1. In Fig. 1F, the blue spots seen outside the epithelial layer are the result of autofluorescent signals from erythrocytes and are indicated by white arrowheads. Nuclei were counterstained with YoPro-1 and are shown in green (Fig. 1, A and E), whereas GIP-positive cells (C and G) are indicated in red. Merged images for E18.5 and wk 10 are shown in Fig. 1, D (for A–C) and H (for E–G), respectively. The white arrows point to PDX-1-positive cells that also stain for GIP immunoreactivity. The green arrow indicates a PDX-1-positive cell that is negative for GIP. Examples of GIP-positive cells that were negative for PDX-1 were not detected.
GIP expression in PDX-1 null mice
Previously, analysis of PDX-1 homozygous null embryos demonstrated that an absence of PDX-1 expression during development resulted in the formation of an abnormal rostral duodenum and a reduction in the number of enteroendocrine cells present in the rostral duodenum (20, 30). We analyzed tissues from 18.5-dpc wild-type (+/+) and PDX-1XBko homozygous null (–/–) embryos for the distribution and number of GIP-producing K-cells by immunohistochemistry. Duodenal sections from three mice from each group were examined by confocal microscopy. In Fig. 2, two representative sections demonstrate the presence of GIP immunoreactivity (red) in wild-type mice (A) but not in PDX-1 null mice (B). In addition, mice were examined for the presence of serotonin-expressing cells. For each animal, five nonadjacent sections from 18.5 dpc were evaluated for GIP- and serotonin-positive cells. Figure 2C depicts the results obtained from the examination of three wild-type (black bar) and three PDX-1–/– (white bar) mice and are represented as the number of GIP- or serotonin-positive cells per 1000 epithelial nuclei. Our analysis demonstrated a 97.8% reduction in GIP-positive cells in the PDX-1 null embryos, compared with wild-type embryos (P = 0.005). A few GIP-positive cells were detected in the relatively distal portion of duodenum of the PDX-1 null embryos, with two thirds located within the crypt region and the other one third located within the lower half of the villi. GIP-positive cells were not detected in the proximal portion of the duodenum in the PDX-1 null embryos. In contrast, differences between the number of serotonin-positive cells detected in wild-type and PDX-1–/– mice (a 28% reduction in the latter) did not reach statistical significance (P > 0.05; data not shown).
FIG. 2. GIP-positive cell number is reduced by 97.8% in the duodenum of PDX-1 null mutation mice. Mouse duodenal sections from 18.5-dpc wild-type (A) and PDX1–/– mice (B) were stained for the presence of GIP immunoreactivity (red) using a rabbit polyclonal antibody. GIP immunoreactivity was detected in wild-type mice but not in PDX-1 null, mice. C, Quantification of GIP-positive and serotonin-positive cells. Five nonadjacent sections from 18.5-dpc animals were counted for GIP- and serotonin-positive cells. The figure depicts the results obtained from the examination of three wild-type (black bar) and three PDX-1–/– (white bar) mice. The vertical axis represents the number of GIP- and serotonin-positive cells per 1000 epithelial nuclei. Data are expressed as the mean ± SE (n = 3). The reduction in the number of GIP-positive cells, but not in serotonin-positive cells, in PDX-1–/– mice was significant; *, P = 0.005.
PDX-1 expression in STC-1 cells
To determine whether the loss of GIP expression in the PDX-1 null mice occurs as a result of a direct interaction between PDX-1 and the GIP promoter region, we examined GIP/PDX-1 interactions in STC-1 cells. Initially, Northern blot analysis was used to confirm the presence of PDX-1 transcripts in STC-1 cells (Fig. 3A). Using a PDX-1 cDNA probe, we demonstrated the presence of a 2.3-kb transcript in STC-1 cells but not in control IEC-6 cells. In addition, we detected the presence of a larger 7-kb transcript in the STC-1 cell RNA extract. The presence of this larger transcript has previously been detected in HIT cells, ?TC6 cells, and isolated mouse and rat islets (31). Northern analysis using a GIP-specific probe confirmed the presence of GIP transcripts in STC-1 cells.
FIG. 3. PDX-1 is expressed in STC-1 cells. A, Northern blot analysis detects the presence of PDX-1 transcripts in STC-1 cells. Total RNA was extracted from STC-1 and IEC-6 cells, as described in Materials and Methods. After electrophoresis through a 1.0% agarose gel, the RNA was transferred to a nylon membrane, hybridized to a 1.4-kb PDX-1 cDNA probe, and exposed overnight to x-ray film (top). The blot was washed and then probed with a GIP-specific probe. As a positive control for RNA loading, the blot was reprobed with a ?-actin-specific cDNA probe. The blot shown is representative of three independent experiments. B, Western blot analysis demonstrates PDX-1 immunoreactivity in STC-1 cells. Total cell lysates from STC-1 and IEC-6 cells were resolved by SDS-PAGE and transferred to a nylon membrane before incubation with antiserum directed against PDX-1 (top). After washing, the blot was incubated with a secondary antibody conjugated with horseradish peroxidase, subsequently washed, and incubated with chemiluminescence reagents before being exposed to x-ray film. The blot was washed and then probed with an antibody directed against ?-actin (bottom) as a control for loading. Similar results were obtained in four independent experiments. C, Immunohistochemical analysis demonstrates the presence of PDX-1 in the nucleus. STC-1 cells, grown on chamber slides, were fixed as described in Materials and Methods and incubated with antiserum directed against either PDX-1 or GIP. After washing, cells were incubated with donkey antirabbit fluorescein isothiocyanate-conjugated secondary antibody, mounted, and photographed using standard fluorescence microscopy.
IEC-6 cells, an immature intestinal stem cell line derived from normal rat small intestine that displays characteristics of immature intestinal crypt cells, were used as a negative control. IEC-6 cells exhibit undifferentiated morphology, have limited expression of intestinal cell-specific genes, and have been shown to possess the ability to differentiate into numerous cell types, including other enteroendocrine cells.
Western blot analysis was next used to demonstrate the presence of PDX-1 protein in STC-1 cells. When whole-cell lysate from IEC-6 and STC-1 cells were probed with antibodies directed against the N terminus of PDX-1, a 46-kDa band was detected in STC-1 cells but not in IEC-6 cells (Fig. 3B). Incubation of the blot with an antibody directed against ?-actin confirmed the presence of protein in both lanes. The cellular location of PDX-1 protein in STC-1 cells was determined by immunohistochemical analysis (Fig. 3C). When formalin-fixed STC-1 cells were incubated with antibodies directed against PDX-1 and GIP, PDX-1 immunoreactivity was detected primarily in the nucleus, whereas GIP immunoreactivity was identified in the cytoplasm.
PDX-1 binds to the CATTA region in the GIP promoter
The region located between –156 and –151 of the GIP promoter was previously shown to be important for GIP expression in STC-1 cells (16). Mutation of the sequence CATTA to gAaaA resulted in an 85% reduction of promoter activity, as assayed by transient transfection assays in STC-1 cells. In addition, EMSAs demonstrated specific binding to this region and identified the transcription factor ISL-1 as a component of the DNA-protein complex. In the present study, supershift assays using a PDX-1-specific antibody demonstrated that PDX-1 also binds to this region (Fig. 4). Nuclear extract from STC-1 cells was mixed with a 32P-labeled double-stranded oligonucleotide corresponding to the region spanning from –160 to –141 of the GIP gene in the absence of specific antiserum (lane 1) or in the presence of antiserum directed against GATA-4 (lane 2), ISL-1 (lane 3), PDX-1 (lane 4), or a combination of ISL1 and PDX-1 (lane 5). Both ISL-1 and PDX-1 antisera were able to supershift the complex, indicating that both proteins interact with the corresponding regulatory element in vitro.
FIG. 4. PDX-1 binds to the CATTA element in the GIP promoter. EMSAs were performed with labeled 20-bp double-stranded oligonucleotides containing sequences spanning –140 to –120 of the GIP gene. The probe was incubated with 2.5 μg of STC-1 cell nuclear protein in the absence of antiserum (lane 1) or presence of 1 μl of GATA-4 antibody (lane 2), 1 μl of ISL-1 antibody (lane 3), 1 μl PDX-1 antibody (lane 4), or 1 μl both PDX-1 and ISL-1 (lane 5). The black arrow denotes the supershifted complexes.
PDX-1 binds to control region of the GIP promoter in an intact cell
To demonstrate that PDX-1 binds to the endogenous GIP gene, a ChIP analysis was performed using formaldehyde cross-linked chromatin isolated from STC-1 cells. After precipitation of chromatin in the presence of 1 or 2 μl anti-PDX-1 antiserum or 10 μg normal rabbit antiserum or in the absence of antiserum, the GIP 5'-flanking sequence spanning nucleotides –461 to –11 was selectively amplified by PCR. As shown in Fig. 5, only the samples treated with PDX-1 antiserum yielded a PCR product. As expected, negative results were obtained in immunoprecipitation reactions in which chromatin was omitted. In addition, the PDX-1 antiserum did not immunoprecipitate regulatory sequences from the phosphoenolpyruvate carboxykinase (PEPCK) gene, a gene that contains a functional PDX-1 binding site but is not transcriptionally active in STC-1 cells. These results demonstrate that PDX-1 binds to the GIP promoter region in intact STC-1 cells. Similar results were obtained using the ISL-1 antiserum (data not shown).
FIG. 5. PDX-1 binds to the GIP gene control sequences in intact neuroendocrine cells. Formaldehyde cross-linked chromatin from STC-1 cells was incubated with 1 μl (lane 3) or 2 μl (lane 4) of an antiserum raised against the N terminus of PDX-1. Immunoprecipitated DNA was analyzed by PCR with primers to transcriptional regulatory sequences of the mouse GIP and PEPCK genes. As controls, PCRs were performed with DNA immunoprecipitated with normal rabbit immune sera (lane 2), input DNA (1:100 dilution; lane 5), DNA immunoprecipitated in the absence of antiserum (lane 6), chromatin minus sample immunoprecipitated with 2 μl PDX-1 antiserum (lane 7), and no template (lane 1). The blot shown is representative of three independent experiments.
PDX-1 overexpression increases the activity of the GIP promoter
To demonstrate the functional relevance of our findings, we next examined the effects of PDX-1 protein overexpression on GIP promoter activity in STC-1 cells (Fig. 6). Transient transfection analyses of various GIP-Luc reporter constructs with either a PDX1/pcDNA3.1 expression plasmid or pcDNA3.1 empty vector were conducted. PDX-1 overexpression caused an increase in the luciferase activity of all GIP promoter constructs containing the PDX-1 binding site (pGL-2569, pGL-193, and pGL-173). In contrast, PDX-1 expression did not affect the activity of a GIP-Luc construct containing a mutation in the PDX-1 binding site (pGL-193D) or the promoterless construct, pGL2 basic. These observations further support the importance of the transcription factor PDX-1 in the regulation of the GIP gene.
FIG. 6. Overexpression of PDX-1 leads to a specific increase in activity of GIP/Luc reporter constructs. The effect that the overexpression of PDX-1 on the functional activity of GIP/Luc fusion constructs was assessed in STC-1 cells. Cotransfections were performed using pRL-cytomegalovirus plus GIP/Luc promoter constructs containing various lengths of the GIP promoter (pGL-2569, pGL-193, and pGL-173) or a clone containing a mutation in the PDX-1 binding site (pGL-193D) and either the expression plasmid PDX-1/pcDNA3.1 or the empty vector (pcDNA3.1). After 48 h, the cells were harvested and analyzed for luciferase and renilla activity, as described in Materials and Methods. The data represent mean activity ± SE of six independent transfections normalized to the activity of pGL2 basic, after correcting for differences in transfection efficiencies determined by associated renilla activity.
Discussion
Previous studies from our laboratory have established the importance of the sequence between base pairs –156 and –151 of the GIP promoter for transcription of the GIP gene in STC-1 cells and have demonstrated the in vitro binding of ISL-1 to this region (16). ISL-1, a member of a family of LIM-homeodomain genes (32), has been implicated in the transcriptional regulation of such genes as proglucagon (33), somatostatin (34), and amylin (35) and has been determined to be critical for normal pancreatic development and endocrine cell formation (36).
In this report, we show that PDX-1, a transcription factor that is capable of binding to the same DNA consensus motif as ISL-1, is present in STC-1 cells and also interacts with the region between –156 and –151 of the GIP promoter in vitro. Gel shift assays using antisera directed against ISL-1 and PDX-1 demonstrated that both transcription factors were capable of binding to the cis-regulatory element. The ability of both PDX-1 and a LIM-homeodomain family member to bind to and activate transcription from the same cis-regulatory region has been described previously. Both PDX-1 and the LIM-homeodomain family member Lmx1.1 have been reported to possess the capacity to bind to the A3 and A4 sites of the rat insulin promoter and activate transcription of the insulin gene (37, 38, 39, 40).
Although in vitro binding assays and transient transfection assays with reporter constructs provide evidence for the involvement of a factor in the transcription of a gene, what occurs in these artificial systems may not necessarily reflect what is occurring in an intact cell. In the nucleus, transcriptional activation is affected by endogenous chromatin content (41). Closed chromatin structure or competition from another DNA binding protein with similar nuclear specificity may exclude transcription factors from a potential binding site. This possibility is particularly relevant for transcription factors such as PDX-1, which recognize fairly nonselective consensus sequences within promoter regions. Thus, the ability of PDX-1 to interact with these potential binding sites in vitro may not directly equate to the involvement of this transcription factor in the regulation of the gene. For example, when examining PDX-1 binding by in vitro assays using nuclear extract from ?TC-3 cells, Cissell et al. (27) found that PDX-1 was able to bind to TAAT motifs in both the PEPCK and insulin promoters. In contrast, when these binding interactions were examined using ChIP analysis, PDX-1 was shown to interact with the endogenous insulin, but not the PEPCK, gene (27). This result suggests that PDX-1 interacts with the insulin gene and not the PEPCK gene in the context of chromatin. In the present study, using ChIP analysis, we were likewise able to confirm our in vitro findings and establish a direct interaction between the transcription factor PDX-1 and the GIP gene.
PDX-1, a transcription factor originally discovered as an activator of insulin and somatostatin, has proven to be a key factor in both pancreatic development and adult islet ?-cell function (20, 42, 43, 44, 45, 46, 47). Analysis of PDX-1 knockout mice demonstrated that animals lacking PDX-1 exhibit pancreatic agenesis (20, 48), resulting in extreme hyperglycemia and perinatal death. In addition, Ahlgren et al. (49), using the Cre-LoxP system to produce cell-type gene inactivation in ?-cells, determined that continued PDX-1 expression in the adult mouse is required for the expression of various ?-cell genes, including insulin, GLUT2, and IAPP. Furthermore, functional activity of PDX-1 is not limited to the pancreas. Offield et al. (20) reported that in addition to pancreatic abnormalities, null mutation of the PDX-1 gene resulted in an altered architecture of the rostral duodenum, an absence of Brunner’s glands, and a reduction in enteroendocrine cells in the villous epithelium.
Guz et al. (45) examined the developmental expression pattern of PDX-1 in the gut and reported the presence of PDX-1-positive cells in the endodermal layer of the dorsal region of the gut beginning at E8.5. They also found that by E9.5, PDX-1 was expressed in most of the epithelial cells within the mucosal layer but not in nearby connective tissues. When older embryos and adults were examined, PDX-1 expression persisted in almost all the epithelial cells lining the villi but was not detected within the crypts. In addition, PDX-1 expression was found neither in other mucosal cells nor in cells comprising the submucosa, muscularis, or adventitial layers of the duodenal wall. As a result of their observations, the authors speculated that PDX-1 may play a critical role in duodenal epithelial cell function.
In the present study, we provide evidence for the involvement of PDX-1 in the regulation of GIP expression within the enteroendocrine K-cells of the proximal duodenum, and we have demonstrated that GIP and PDX-1 are both expressed in the K-cells of the duodenum. In addition, after a careful analysis of GIP expression in both PDX-1 wild-type and null mice, we found that the loss of PDX-1 expression nearly abolished GIP-producing cells in the gut. Previously, homozygosity for mutations in the PDX-1 gene in mice was reported to affect the enteroendocrine cells in the duodenum (20, 30, 48). Specifically, Offield et al. (20) reported a general reduction in the number of enteroendocrine cells in PDX-1 null mice but did not specifically measure GIP-expressing cells. In addition, Larsson et al. (30) reported finding a 72% reduction in K-cells. The reason for the discrepancy between our findings and their report is unclear but may be because of methodological differences, including antibody specificity. Another important variable might be the technique used to determine the distribution and extent of GIP-expressing cells. Although we did detect a very small number of GIP-expressing cells (six GIP-positive cells of >70,000 epithelial cells) in the PDX-1 null mice, they were located predominately in the very distal duodenum, outside the boundaries of PDX-1 expression. Because these cells appear to be restricted to the crypt region or the lower half of the villi, it is possible they represent undifferentiated progenitor cells. However, because so few cells were detected, we cannot definitively conclude that they are restricted to the lower half of the villi. Nonetheless, it is important to note that no GIP-positive cells were detected in the proximal portion of the duodenum in the PDX-1 null embryos. This finding, along with the 97.8% reduction in GIP-positive cells in PDX-1 null mutants, clearly indicates a primary and nonredundant role for PDX-1 in cell differentiation for most, if not all, GIP-producing cells.
It is not surprising that the GIP and insulin genes are regulated by common transcription factors. GIP and insulin are both hormones involved in nutrient storage, and GIP-producing intestinal K-cells and insulin-producing pancreatic ?-cells share a common developmental lineage. Intestinal K-cells are derived from stem cells located in the intestinal crypts, where, like other neuroendocrine tissue such as the pancreatic islets, they are thought to originate developmentally from the primitive endoderm. Interestingly, PDX-1 has been shown to be required for both the regionalization of the primitive gut endoderm and the maturation of the pancreatic ?-cells (47, 50).
It is not known whether PDX-1 plays a role in the process of stem cell differentiation into GIP-producing K-cells. However, it seems likely that the use of a transcription factor to control both the spatial and temporal differentiation of a specific endocrine cell type and the main hormone that defines the cell would be advantageous from an evolutionary standpoint. As discussed above, PDX-1 seems to support this contention with respect to ?-cell differentiation and insulin gene regulation. Therefore, if PDX-1 plays a similar role in the K-cell, it is possible that the decreased presence of GIP-producing cells we observed in both embryonic and adult PDX-1 null mice may result not only from a decreased rate of transcription of the GIP gene but also from a decreased number of K-cells. Because GIP presently represents the only known K-cell marker, future experiments will be necessary to enable us to identify additional K-cell markers whose transcription might occur independently of PDX-1 to examine this possibility.
References
Brown JC, Dryburgh JR 1971 A gastric inhibitory polypeptide. II. The complete amino acid sequence. Can J Biochem 49:867–872
Buchan AM, Polak JM, Capella C, Solcia E, Pearse AG 1978 Electronimmunocytochemical evidence for the K cell localization of gastric inhibitory polypeptide (GIP) in man. Histochemistry 56:37–44
Pederson RA, Schubert HE, Brown JC 1975 Gastric inhibitory polypeptide: its physiologic release and insulinotropic action in the dog. Diabetes 24:1050–1056
Cataland S, Crockett SE, Brown JC, Mazzaferri EL 1974 Gastric inhibitory polypeptide (GIP) stimulation by oral glucose in man. J Clin Endocrinol Metab 39:223–228
Buffa R, Polak JM, Pearse AG, Solcia E, Grimelius L, Capella C 1975 Identification of the intestinal cell storing gastric inhibitory peptide. Histochemistry 43:249–255
Cheung AT, Dayanandan B, Lewis JT, Korbutt GS, Rajotte RV, Bryer-Ash M, Boylan MO, Wolfe MM, Kieffer TJ 2000 Glucose-dependent insulin release from genetically engineered K cells. Science 290:1959–1962
Nauck MA, Holst JJ, Willms B, Schmiegel W 1997 Glucagon-like peptide 1 (GLP-1) as a new therapeutic approach for type 2-diabetes. Exp Clin Endocrinol Diabetes 105:187–195
Gutniak M, Orskov C, Holst JJ, Ahren B, Efendic S 1992 Antidiabetogenic effect of glucagon-like peptide-1 (7–36)amide in normal subjects and patients with diabetes mellitus. N Engl J Med 326:1316–1322
Gault VA, O’Harte FP, Harriott P, Mooney MH, Green BD, Flatt PR 2003 Effects of the novel (Pro(3))GIP antagonist and exendin(9–39)amide on GIP- and GLP-1-induced cyclic AMP generation, insulin secretion and postprandial insulin release in obese diabetic (ob/ob) mice: evidence that GIP is the major physiological incretin. Diabetologia 46:222–230
Tseng CC, Kieffer TJ, Jarboe LA, Usdin TB, Wolfe MM 1996 Postprandial stimulation of insulin release by glucose-dependent insulinotropic polypeptide (GIP): effect of a specific glucose-dependent insulinotropic polypeptide receptor antagonist in the rat. J Clin Invest 98:2440–2445
Miyawaki K, Yamada Y, Ban N, Ihara Y, Tsukiyama K, Zhou H, Fujimoto S, Oku A, Tsuda K, Toyokuni S, Hiai H, Mizunoya W, Fushiki T, Holst JJ, Makino M, Tashita A, Kobara Y, Tsubamoto Y, Jinnouchi T, Jomori T, Seino Y 2002 Inhibition of gastric inhibitory polypeptide signaling prevents obesity. Nat Med 8:738–742
Takeda J, Seino Y, Tanaka K, Fukumoto H, Kayano T, Takahashi H, Mitani T, Kurono M, Suzuki T, Tobe T 1987 Sequence of an intestinal cDNA encoding human gastric inhibitory polypeptide precursor. Proc Natl Acad Sci USA 84:7005–7008
Tseng CC, Jarboe LA, Landau SB, Williams EK, Wolfe MM 1993 Glucose-dependent insulinotropic peptide: structure of the precursor and tissue-specific expression in rat. Proc Natl Acad Sci USA 90:1992–1996
Higashimoto Y, Liddle RA 1993 Isolation and characterization of the gene encoding rat glucose-dependent insulinotropic peptide. Biochem Biophys Res Commun 193:182–190
Boylan MO, Jepeal LI, Jarboe LA, Wolfe MM 1997 Cell-specific expression of the glucose-dependent insulinotropic polypeptide gene in a mouse neuroendocrine tumor cell line. J Biol Chem 272:17438–17443
Jepeal LI, Boylan MO, Wolfe MM 2003 Cell-specific expression of the glucose-dependent insulinotropic polypeptide gene functions through a GATA and an ISL-1 motif in a mouse neuroendocrine tumor cell line. Regul Pept 113:139–147
Orkin SH 1992 GATA-binding transcription factors in hematopoietic cells. Blood 80:575–581
Evans T, Reitman M, Felsenfeld G 1988 An erythrocyte-specific DNA-binding factor recognizes a regulatory sequence common to all chicken globin genes. Proc Natl Acad Sci USA 85:5976–5980
Inagaki N, Seino Y, Takeda J, Yano H, Yamada Y, Bell GI, Eddy RL, Fukushima Y, Byers MG, Shows TB 1989 Gastric inhibitory polypeptide: structure and chromosomal localization of the human gene. Mol Endocrinol 3:1014–1021
Offield MF, Jetton TL, Labosky PA, Ray M, Stein RW, Magnuson MA, Hogan BL, Wright CV 1996 PDX-1 is required for pancreatic outgrowth and differentiation of the rostral duodenum. Development 122:983–995
Kitamura T, Nakae J, Kitamura Y, Kido Y, Biggs 3rd WH, Wright CV, White MF, Arden KC, Accili D 2002 The forkhead transcription factor Foxo1 links insulin signaling to Pdx1 regulation of pancreatic ?-cell growth. J Clin Invest 110:1839–1847
Jenny M, Uhl C, Roche C, Duluc I, Guillermin V, Guillemot F, Jensen J, Kedinger M, Gradwohl G 2002 Neurogenin3 is differentially required for endocrine cell fate specification in the intestinal and gastric epithelium. EMBO J 21:6338–6347
Brissova M, Shiota M, Nicholson WE, Gannon M, Knobel SM, Piston DW, Wright CV, Powers AC 2002 Reduction in pancreatic transcription factor PDX-1 impairs glucose-stimulated insulin secretion. J Biol Chem 277:11225–11232
Wolfe MM, Hocking MP, Maico DG, McGuigan JE 1983 Effects of antibodies to gastric inhibitory peptide on gastric acid secretion and gastrin release in the dog. Gastroenterology 84:941–948
Dignam JD, Lebovitz RM, Roeder RG 1983 Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic Acids Res 11:1475–1489
Gerrish K, Cissell MA, Stein R 2001 The role of hepatic nuclear factor 1 and PDX-1 in transcriptional regulation of the pdx-1 gene. J Biol Chem 276:47775–47784
Cissell MA, Zhao L, Sussel L, Henderson E, Stein R 2003 Transcription factor occupancy of the insulin gene in vivo: evidence for direct regulation by Nkx2.2. J Biol Chem 278:751–756
Kieffer TJ, Buchan AM, Barker H, Brown JC, Pederson RA 1994 Release of gastric inhibitory polypeptide from cultured canine endocrine cells. Am J Physiol 267:E489–E496
Roth KA, Rubin DC, Birkenmeier EH, Gordon JI 1991 Expression of liver fatty acid-binding protein/human growth hormone fusion genes within the enterocyte and enteroendocrine cell populations of fetal transgenic mice. J Biol Chem 266:5949–5954
Larsson LI, Madsen OD, Serup P, Jonsson J, Edlund H 1996 Pancreatic-duodenal homeobox 1: role in gastric endocrine patterning. Mech Dev 60:175–184
Olson LK, Qian J, Poitout V 1998 Glucose rapidly and reversibly decreases INS-1 cell insulin gene transcription via decrements in STF-1 and C1 activator transcription factor activity. Mol Endocrinol 12:207–219
Schmeichel KL, Beckerle MC 1994 The LIM domain is a modular protein-binding interface. Cell 79:211–219
Wang M, Drucker DJ 1995 The LIM domain homeobox gene isl-1 is a positive regulator of islet cell-specific proglucagon gene transcription. J Biol Chem 270:12646–12652
Vallejo M, Penchuk L, Habener JF 1992 Somatostatin gene upstream enhancer element activated by a protein complex consisting of CREB, Isl-1-like, and -CBF-like transcription factors. J Biol Chem 267:12876–12884
Wang M, Drucker DJ 1996 Activation of amylin gene transcription by LIM domain homeobox gene isl-1. Mol Endocrinol 10:243–251
Ahlgren U, Pfaff SL, Jessell TM, Edlund T, Edlund H 1997 Independent requirement for ISL1 in formation of pancreatic mesenchyme and islet cells. Nature 385:257–260
Ohneda K, Mirmira RG, Wang J, Johnson JD, German MS 2000 The homeodomain of PDX-1 mediates multiple protein-protein interactions in the formation of a transcriptional activation complex on the insulin promoter. Mol Cell Biol 20:900–911
Peers B, Leonard J, Sharma S, Teitelman G, Montminy MR 1994 Insulin expression in pancreatic islet cells relies on cooperative interactions between the helix loop helix factor E47 and the homeobox factor STF-1. Mol Endocrinol 8:1798–1806
Peshavaria M, Henderson E, Sharma A, Wright CV, Stein R 1997 Functional characterization of the transactivation properties of the PDX-1 homeodomain protein. Mol Cell Biol 17:3987–3996
German MS, Wang J, Chadwick RB, Rutter WJ 1992 Synergistic activation of the insulin gene by a LIM-homeo domain protein and a basic helix-loop-helix protein: building a functional insulin minienhancer complex. Genes Dev 6:2165–2176
Ozcan S 2003 Diabetes mellitus methods and protocols. Totowa, NJ: Humana Press
Habener JF, Stoffers DA 1998 A newly discovered role of transcription factors involved in pancreas development and the pathogenesis of diabetes mellitus. Proc Assoc Am Physicians 110:12–21
Edlund H 1998 Transcribing pancreas. Diabetes 47:1817–1823
Jonsson J, Ahlgren U, Edlund T, Edlund H 1995 IPF1, a homeodomain protein with a dual function in pancreas development. Int J Dev Biol 39:789–798
Guz Y, Montminy MR, Stein R, Leonard J, Gamer LW, Wright CV, Teitelman G 1995 Expression of murine STF-1, a putative insulin gene transcription factor, in ?-cells of pancreas, duodenal epithelium and pancreatic exocrine and endocrine progenitors during ontogeny. Development 121:11–18
Leonard J, Peers B, Johnson T, Ferreri K, Lee S, Montminy MR 1993 Characterization of somatostatin transactivating factor-1, a novel homeobox factor that stimulates somatostatin expression in pancreatic islet cells. Mol Endocrinol 7:1275–1283
Ahlgren U, Jonsson J, Edlund H 1996 The morphogenesis of the pancreatic mesenchyme is uncoupled from that of the pancreatic epithelium in IPF1/PDX1-deficient mice. Development 122:1409–1416
Jonsson J, Carlsson L, Edlund T, Edlund H 1994 Insulin-promoter-factor 1 is required for pancreas development in mice. Nature 371:606–609
Ahlgren U, Jonsson J, Jonsson L, Simu K, Edlund H 1998 ?-Cell-specific inactivation of the mouse Ipf1/Pdx1 gene results in loss of the ?-cell phenotype and maturity onset diabetes. Genes Dev 12:1763–1768
Ohlsson H, Karlsson K, Edlund T 1993 IPF1, a homeodomain-containing transactivator of the insulin gene. EMBO J 12:4251–4259(Lisa I. Jepeal, Yoshio Fu)