The pancreatic and duodenal homeobox factor-1 (PDX-1), also known as IDX-1/STF-1/IPF1, a homeodomain-containing transcription factor, plays a central role in regulating pancreatic development and insulin gene transcription. Furthermore, even in adults, PDX-1 is associated with islet neogenesis and differentiation of insulin-producing cells from progenitor cells. Here, we report for the first time that PDX-1 protein can permeate cells due to an Antennapedia-like protein transduction domain sequence in its structure and that transduced PDX-1 functions similarly to endogenous PDX-1; it binds to the insulin promoter and activates its expression. PDX-1 protein can also permeate into isolated pancreatic islets, which leads to stimulation of insulin gene expression. Moreover, PDX-1 protein transduced into cultures of pancreatic ducts, thought to be islet progenitor cells, induces insulin gene expression. These data suggest that PDX-1 protein transduction could be a safe and valuable strategy for enhancing insulin gene transcription and for facilitating differentiation of ductal progenitor cells into insulin-producing cells without requiring gene transfer technology.
Since the discovery of protein transduction domains (PTDs) that allow proteins to be translocated across the plasma membrane and into nuclei without endocytosis, there has been increasing interest in their potential for the delivery of bioactive peptides and proteins into eukaryotic cells as a valuable strategy for the transduction of therapeutic proteins into patients. The small PTDs from TAT protein of HIV-1 (1) and from VP22 protein of Herpes simplex virus (2) have been fused to proteins with the remarkable result of delivery of proteins into many mammalian tissues (1,2). The presence of a PTD in the third α-helix of the homeodomain of Antennapedia, a Drosophila transcription factor, has been shown to be necessary and sufficient for an “unconventional” translocation of this peptide into the cytoplasm and nuclei of living cells (3).
We noticed that the transcription factor PDX-1 protein, which itself contains an Antennapedia-like homeodomain in its structure, has the same amino acid sequence, except for one amino acid toward the NH2-terminus, as the Antennapedia PTD (Fig. 1A). The presence of a PTD in PDX-1 is intriguing because PDX-1 plays such a crucial role in pancreatic development (4–6), β-cell differentiation (7–9), and maintenance of normal β-cell function by its regulation of multiple important β-cell genes (10,11), including insulin (12,13). Adenoviral-mediated introduction of PDX-1 induced insulin expression in the liver and improved the glucose tolerance of streptozotocin-induced diabetic mice (8). The ability of the PDX-1 protein to both regulate its own gene expression through an A-box element in its promoter (14) and to induce the expression of other β-cell genes is thought to be the basis of this induced insulin expression.
Here, we report for the first time that PDX-1 protein can permeate cells due to the Antennapedia-like protein transduction domain sequence in its structure and that transduced PDX-1 functions similarly to endogenous PDX-1.
RESEARCH DESIGN AND METHODS
Construction of vectors and purification of recombinant PDX-1 proteins.
Full-length PDX-1 cDNA was amplified by PCR using appropriate linker-primers and then subcloned into the NdeI and XhoI sites of pET21a (+) (Novagen, Madison, WI) using a ligation kit (TaKaRa, Tokyo, Japan). For deletion of the PTD, sequences before and after the PTD of PDX-1 cDNA were amplified by PCR using appropriate linker-primers and then subcloned into the NdeI-BamHI and BamHI-XhoI sites of pET21a (+), respectively. For EGFP-11 arginine (11R), full-length enhanced green fluorescent protein cDNA was amplified by PCR using appropriate linker-primers, and 11R sequence was synthesized by Genosys-Sigma (The Woodlands, TX) and then subcloned into the NdeI-BamHI and NotI-XhoI sites of pET21a (+). BL21 (DE3) cells containing the expression plasmids were grown at 37°C to an OD600 of 0.8. Isopropyl-β-d-thiogalactopyranoside was added to a final concentration of 0.1 mmol/l, and the cells were then incubated for 12 h at 24°C. Cells were sonicated, and the supernatants were recovered and applied to a column of Ni-NTA agarose (Invitrogen, San Diego, CA). Purified PDX-1 proteins were conjugated using a fluorescein isothiocyanate (FITC)-labeling kit (American Qualex Antibodies, San Clemente, CA). FITC-labeled PDX-1 PTD peptide was synthesized by Genosys-Sigma.
Western blotting.
After three washings with PBS, cells were scrapped from the dish and sonicated. Ten micrograms of cell extracts were fractionated by 10% SDS-PAGE and transferred to polyvinylideno fluoride membranes (Immun-Blot PVDF Membrane; Bio-RAD, Hercules, CA) using transfer buffer containing 20% methanol, 25 mmol/l Tris base, and 192 mmol/l glycine (300 mA, 2 h). After blocking at room temperature for 1 h in 50 mmol/l Tris-HCl, 150 mmol/l NaCl, and 0.1% Tween-20 (Tris-buffered saline with Tween [TBST]) with 5% nonfat dry milk, the membranes were incubated overnight at 4°C in TBST using 5% nonfat dry milk containing rabbit anti-PDX-1 (6), anti–6 His antibody coupled to horseradish peroxidase (HRP) (1:5,000; Invitrogen), or mouse anti-actin antibody (1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA), and then for 1 h at room temperature in TBST with 5% nonfat dry milk containing anti-rabbit (PDX-1) or anti-mouse IgG (actin) antibody coupled to HRP (1:1,000; Bio-RAD).
Immunostaining.
Cells were fixed with 4% paraformaldehyde in PBS buffer. After blocking with donkey serum for 30 min at room temperature, cells were incubated overnight at 4°C with rabbit anti–PDX-1 antiserum (6) (1:1,000 in PBS containing 1% BSA) and then for 1 h at room temperature with FITC-conjugated donkey anti-rabbit IgG (1:100; Jackson Immunochemicals, West Grove, PA).
Gel-mobility shift assay.
Nuclear extract (2 μg) was incubated with 2 μg poly (dI-dC), 10 mmol/l HEPES (pH 7.8), 0.1 mmol/l EDTA, 75 mmol/l KCl, 2.5 mmol/l MgCl2, and 1 mmol/l dithiothreitol, 3% Ficoll, at room temperature. The binding reaction was initiated by adding [32P]-labeled double-stranded oligonucleotide probes. A double-stranded oligonucleotide reproducing the rat insulin gene II PDX-1 binding region and surrounding sequences (ACGTCC TCTTAAGAC TCTAATTACCCTACG T) (Sigma Genosys) was used as a binding probe. In some of the binding assays, anti-PDX-1 (6), anti–6 His, and preimmune antisera were added to the reaction mixture 1 h before addition of the DNA probes.
Gene transfection and luciferase assay.
Rat insulin II promoter-reporter (luciferase) plasmid (1.0 μg) containing 238-bp 5′-flanking sequences of the rat insulin II promoter region or A3-box mutated insulin II promoter-reporter plasmid (1.0 μg) (13,15) was transfected into cells with LipofectAMINE (Invitrogen) using the conditions recommended by the manufacturer. Forty-eight hours after transfection, cells were harvested and assayed (Promega, Madison, WI).
Isolation and culture of rat pancreatic islets and duct cells.
Islets and pancreatic ducts were isolated from pancreases of Sprague-Dawley rats (Taconic Farms, Germantown, NY). All animal procedures were approved by the Animal Care Committee of the Joslin Diabetes Center. For islet isolation, the common bile duct was cannulated and injected with 6 ml cold M199 medium containing 1.5 mg/ml collagenase (Roche Boehringer Mannheim, Indianapolis, IN). The islets were separated on Histopaque 1077 (Sigma, St. Louis, MO) density gradient, hand-picked under a dissecting microscope to ensure a pure islet preparation, and used immediately afterward. Common pancreatic ducts were isolated using a modified islet isolation method and were cultured in Dulbecco’s modified Eagle’s medium/F12. After cells attached and spread, nonductal cells were removed mechanically with a rubber scrapper and the remaining ductal cells passaged; this was repeated four to five times. The ductal cells used in these experiments were from passages 7–9. In these experiments islets and ductal cells were cultured in RPMI medium with 10% FCS.
Semiquantitative radioactive multiplex PCR.
PCRs were performed in a Perkin-Elmer 9700 Thermocycler with 3 μl cDNA (20 ng RNA equivalents), 160 μmol/l cold dNTPs, 2.5 μCi [α-32P]dCTP (3,000 Ci/mmol), 10 pmol appropriate oligonucleotide primers, 1.5 mmol/l MgCl2, and 5 units AmpliTaq Gold DNA polymerase (Perkin-Elmer, Norwalk, CT). The oligonucleotide primers and cycle number used for semiquantitative-radioactive multiplex PCR were insulin (forward) TCTTCTACACACCCATGTCCC, (reverse) GGTGCAGCACTGATCCAC, 15 cycles for islets, 28 cycles for duct cells; PDX-1 (forward) CGGACATCTCCCCATACG, (reverse) AAAGGGAGATGAACGCGG, 28 cycles; islet amyloid polypeptide (IAPP) (forward) CAACCCTCAGGTGGACAAAC, (reverse) CTCTGCCACATTCCTCTTCC, 35 cycles; glucokinase (GK): (forward) TGACAGAGCCAGGATGGAG, (reverse) TCTTCACGCTCCACTGCC, 35 cycles; Nkx6.1: (forward) TCTTCTGGCCTGGGGTGATG, (reverse) GGCTGCGTGCTTCTTTCTCCA, 35 cycles; Cyclophilin: (forward) AACCCCACCGTGTTCTTC, (reverse) TGCCTTCTTTCACCTTCCC, 28 cycles, and the primers for rRNA were purchased from AMBION (Austin, TX) (15 cycles). The thermal cycle profile used a 10-min denaturing step at 94°C followed by amplification cycles (1 min denaturation at 94°C, 1 min annealing at 55°C, and 1 min extension at 72°C) and an extension step of 10 min at 72°C. The steps taken to validate these measurements were previously reported (16).
RESULTS
Transduction of PDX-1 protein into cells.
To test whether purified PDX-1 protein with its native PTD sequence can be transduced into cells, cervix-derived HeLa, liver-derived HepG2, and β-cell-derived MIN6 cells were treated with FITC-conjugated PDX-1 protein. Six hours after this treatment, PDX-1 was observed as a fluorescence signal in almost all of the HeLa (Fig. 1C), HepG2, and MIN6 cells (data not shown). To test whether the 16 amino acids of PDX-1 truly form a PTD, we compared the penetration of FITC-conjugated full-length PDX-1, PTD-deleted PDX-1 (mutant PDX-1), and the PTD peptide (PDX-1 PTD) (Fig. 1B). The peptide of the 16 amino acids similar to Antennapedia PTD transduced HeLa cells and was seen in the nuclei and cytoplasm in living cells, but PTD-deleted PDX-1 protein did not penetrate the cells (Fig. 1C). These data show that the 16 amino acids of PDX-1 form a functional PTD.
Effects of transduced PDX-1 on clonal cells.
To test the effectiveness and stability of the transduced PDX-1 protein, HepG2, HeLa, and MIN6 cells were treated with PDX-1 protein at several concentrations and for variable times. In MIN6 (Fig. 2A), HepG2 (Fig. 2B), and HeLa (Fig. 2C) cells, PDX-1 protein was transduced in a dose-dependent manner up to 1 μmol/l and was stable for at least 48 h. In treated MIN6 cells, the band for the transduced PDX-1 could be separated from the endogenous PDX-1 by running the gel for a longer time. Additionally, the transduced PDX-1 could be detected by using an antibody to the histidine tag included in this synthesized protein.
To test whether transduced PDX-1 protein binds to the insulin enhancer element, we used gel shift assays. Using nuclear extracts from PDX-1–treated MIN6 cells there was increased A-box binding complex (Fig. 3B). Additionally, the A-box binding complex was observed in nuclear extracts from treated HeLa (Fig. 3A) and HepG2 cells (data not shown). Treatment with either mutant PDX-1 or EGFP-11R, which can be transduced into cells (Fig. 1C) (17), did not induce A-box complexes in the non-islet-derived cells nor increase them in MIN6 cells.
To test whether transduced PDX-1 protein can activate insulin gene transcription, we examined its effect on insulin promoter activity using a luciferase assay. Insulin promoter activity was increased by treatment with PDX-1 protein in HepG2, MIN6, and HeLa cells (Fig. 4) but not with mutant PDX-1 or EGFP-11R protein. Mutated insulin promoter activity was not significantly changed by any treatments.
Together these results suggest that exogenous PDX-1 protein can be transduced into cells and their nuclei, bind to the A-box, and activate the insulin promoter.
Effect of PDX-1 on islets.
To examine whether exogenous PDX-1 protein can penetrate into primary islet cells, isolated islets from Sprague-Dawley rats were treated with PDX-1 protein. The FITC-conjugated PDX-1 was observed as a fluorescence signal in both nuclei and cytoplasm in islets (Fig. 5A), and PDX-1 was transduced in a dose-dependent manner up to 1 μmol/l and was stable for at least 48 h in islets (Fig. 5B) as shown in the cell lines. To test whether PDX-1 protein enhances insulin gene expression, islets were treated with PDX-1 protein for 3 days (protein added day 0 and day 3). The expression of insulin mRNA was enhanced by treatment with exogenous PDX-1 protein but not with the mutant PDX-1 nor EGFP-11R protein (Fig. 5C). These results show that exogenous PDX-1 protein can enhance insulin gene transcription in primary islets as well as in the MIN6 cell line.
Effect of PDX-1 on cultured pancreatic duct cells.
Since ductal cells can differentiate into islets in vitro (9) and in vivo (18), we examined whether exogenous PDX-1 protein could permeate into ductal cells. FITC-conjugated PDX-1 was clearly detected in nuclei and cytoplasm in treated ductal cells (Fig. 6A). Additionally, PDX-1 was clearly detected by immunostaining in nuclei in treated but not untreated ductal cells (Fig. 6A), and PDX-1 protein was transduced in a dose-dependent manner up to 1 μmol/l and was stable for at least 48 h (Fig. 6B).
To test whether purified PDX-1 protein could induce differentiation of ductal cells into insulin-producing cells, ductal cells were treated with PDX-1 protein for 1 week. Insulin mRNA was clearly detected in ducts treated with PDX-1 protein but not in untreated, mutant PDX-1-treated, or EGFP-11R-treated cells (Fig. 6C). We also examined PDX-1 mRNA levels in these cells since it has been reported that pdx-1 gene transcription is, at least in part, autoregulated by PDX-1 (14). As shown in Fig. 6C, under these RT-PCR conditions PDX-1 mRNA was not detected in untreated, mutant PDX-1-treated, or EGFP-11R-treated ductal cells but was clearly detected after PDX-1 treatment. IAPP, GK, and Nkx6.1 mRNA were also detected when a higher number of cycles (35 cycles) were used in the PCR in PDX-1-treated ductal cells but not in the untreated or mutant PDX-1- or EGFP-11R-treated cells (Fig. 6C). These results suggest that once some PDX-1 protein is transduced into ductal cells, endogenous pdx-1 gene transcription is amplified by this PDX-1 and may facilitate their differentiation to insulin-producing cells. Thus, purified PDX-1 protein could be a useful tool to trigger differentiation of progenitors/precursors to become insulin-producing cells.
DISCUSSION
PDX-1, also known as IDX-1/STF-1/IPF1, a homeodomain-containing transcription factor, plays a central role in regulating pancreatic development and insulin gene transcription (4–6,19,20). Furthermore, even in adults, PDX-1 is associated with islet neogenesis and differentiation of insulin-producing cells from progenitor cells (7–9). PDX-1 contains an Antennapedia-like homeodomain with only one amino acid substitution. We would expect that such sequence difference would not affect the function, since deletion of several NH2-terminal amino acids does not prevent the Antennapedia PTD from being transduced into cells (21). In fact, we have shown that the PTD of PDX-1 has the ability to drive PDX-1 protein into cells. The 16 amino acids of the PDX-1 PTD are extremely important in several other aspects of its function. This sequence includes both the nuclear localization signal of PDX-1 (the six amino acids RRMKWKK) that is necessary for its transport into the nucleus (22) and the KIWFQN motif that is particularly important in DNA binding (23).
Our study shows that PDX-1 protein has a native PTD and can without modification permeate into several cell types and enhance insulin expression in these cells. This finding of a transduced native protein that influences differentiation is unique. Our finding of a functional PTD in the transcription factor PDX-1 suggests that exogenous PDX-1 might be used as a novel approach for enhancing insulin gene transcription without requiring gene transfer technology.
Article Information
This research was supported by National Institutes of Health Grants DK61251 and DK44523; the Joslin DERC Animal, Media and Advanced Microscopy Cores (DK36836); the Diabetes Research and Wellness Foundation; and an important group of private donors. Hideaki Kaneto was a recipient of a fellowship and grant from the Japan Society for the Promotion of Science.
We thank Dr. Masayuki Matsushita and Dr. Masakiyo Sakaguchi (Okayama University) for valuable suggestions, Dr. Arun Sharma for helpful discussions, and Dr. Yoshitaka Kajimoto and Dr. Junichi Miyazaki (Osaka University) for kindly providing PDX-1 antibody and MIN6 cells, respectively.