Characterization of the Mouse Islet-Specific Glucose-6-Phosphatase Catalytic Subunit–Related Protein Gene Promoter by In Situ Footprinting: Correlation With Fusion Gene Expression in the Islet-Derived βTC-3 and Hamster Insulinoma Tumor Cell Lines

[α-³²P]deoxyadenosine 5′-triphosphate (dATP) (>3,000 Ci/mmol⁻¹) and [γ-³²P]ATP (>5,000 Ci/mmol⁻¹) were obtained from Amersham. [³H]acetic acid sodium salt (>10 Ci/mmol⁻¹) was purchased from ICN. Proteinase K, deoxynucleoside triphosphates (dNTPs; dATP, deoxycytidine 5′-triphosphate, deoxyguanosine 5′-triphosphate, and deoxythymidine 5′-triphosphate), yeast tRNA, and Triton X-100, were all obtained from Boehringer Mannheim. Dimethyl sulfate (DMS) (99+%) and piperidine (99%) were purchased from Aldrich Chemical Company. G-50 Sephadex was obtained from Amersham Pharmacia Biotech. RNase A was purchased from Sigma. Porcine gelatin was obtained from Difco Laboratories. Vent DNA polymerase was purchased from New England Biolabs. Specific antisera to hepatocyte nuclear factor (HNF)-3α (sc-6553 X), HNF-3β (sc-6554 X), and HNF-3γ (sc-5360 X) were obtained from Santa Cruz Biotechnology. All oligonucleotides were synthesized by the Vanderbilt University Medical Center Diabetes Core Laboratory.

The construction of a series of mouse G6Pase catalytic subunit-CAT fusion genes, with the 5′ end points shown in Fig. 12, has been previously described (25,26). The construction of a mouse IGRP-CAT fusion gene containing promoter sequence from −911 to +3, relative to the experimentally determined transcription start site, in the pCAT(An) expression vector has also been previously described (21). A series of truncated IGRP-CAT fusion genes was generated, with the 5′ end points shown in Figs. 1–3,FIG. 2,FIG. 3, using the −911 pCAT(An) construct as a template, by either restriction enzyme digestion or PCR, using standard techniques (27). Digestion of the −911 IGRP-CAT construct with HindIII and NheI, AflII, Xmn I, Bst NI, or AflIII, followed by Klenow treatment of the noncompatible ends, resulted in the generation of plasmids with the calculated 5′ end points of −306, −273, −197, −172, and −66. Specific 5′ PCR primers were used to generate plasmids with the calculated 5′ end points of −254, −234, −219, −129, and −97. The 3′ PCR primer was designed to conserve the sequence of the junction between the IGRP promoter and CAT reporter gene to be the same in all constructs. Promoter fragments generated by PCR were completely sequenced to ensure the absence of polymerase errors, whereas promoter fragments generated by restriction enzyme digestion were only sequenced to confirm the 5′ end points. All plasmid constructs were purified by centrifugation through cesium chloride gradients (27).

Mouse adrenocortical Y1 cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% (vol/vol) calf serum. HIT cells were grown in DMEM containing 2.5% (vol/vol) fetal bovine serum and 15% (vol/vol) horse serum and were transfected by the addition of a calcium phosphate-DNA coprecipitate as previously described (21). Mouse islet β-cell–derived βTC-3 cells were grown to ∼70% confluence in T150 flasks in DMEM containing 2.5% (vol/vol) fetal bovine serum and 15% (vol/vol) horse serum and were replated the day before use into 35-mm–diameter six-well culture plates (∼1 flask to 60 wells). Attached cells were incubated for 10 min in Opti-MEM reduced serum medium (GibcoBRL) before transfection using the lipofectamine reagent (GibcoBRL) according to the manufacturer’s instructions. Briefly, 2 μg of IGRP-CAT plasmid DNA and 0.5 μg of an expression vector encoding renilla luciferase (Promega) were mixed with 100 μl Opti-MEM reduced serum medium. A solution containing 12.5 μl lipofectamine and 100 μl Opti-MEM reduced serum medium was then added dropwise to the DNA solution. After a 30-min incubation at room temperature, this DNA/lipofectamine mixture was diluted with 800 μl of Opti-MEM reduced serum medium, and the resulting solution was added to the βTC-3 cells. The cells were then returned to 37°C in a humidified 5% CO₂/95% air atmosphere in a Nuaire cell culture incubator. After incubation for 5 h, the DNA/lipofectamine/Opti-MEM reduced serum medium was replaced with DMEM containing 2.5% (vol/vol) fetal bovine serum and 15% (vol/vol) horse serum; the cells were then incubated for an additional 18–20 h.

Transfected HIT cells were harvested by trypsin digestion and then sonicated in 300 μl of 250 mmol/l Tris (pH 7.8) containing 2 mmol/l phenylmethylsulfonyl fluoride. The lysate was assayed for β-galactosidase activity as previously described (21). Transfected βTC-3 cells were harvested by trypsin digestion and then solubilized in passive lysis buffer (Promega). After two cycles of freeze/thawing, renilla luciferase activity was assayed as described previously (28). The remaining HIT and βTC-3 lysate was heated for 10 min at 65°C and cellular debris was removed by centrifugation. CAT assays were then performed on the supernatant as previously described (21). To correct for variations in transfection efficiency, the results are expressed as the ratio of CAT:β-galactosidase in HIT cell transfections or CAT:luciferase activity in βTC-3 cell transfections. In addition, three independent preparations of each IGRP-CAT plasmid construct were analyzed to obtain the data shown in each figure.

βTC-3 and Y1 cells were cultured as described above. Total RNA was isolated from βTC-3 cells by cesium chloride centrifugation as previously described (21) and from Y1 cells using the TRI Reagent (Molecular Research Center) according to the manufacturer’s instructions. Two primers were used for the primer extension analysis: a 29-bp primer (5′-GGGTCTCCAACATTGGACATAAAATTTAG-3′), complementary to exon 1 of the mouse IGRP gene, and a 30-bp primer (5′-ATGTCGAAGAACACGGTGGGGTTGACCATG-3′), complementary to exon 1 of the mouse, rat, and human cyclophilin A genes (29–31). Expression of the cyclophilin A gene was measured as an internal control (Fig. 4). After gel purification (27), both primers were 5′ end-labeled with [γ-³²P]ATP to a specific activity of ∼2 Ci/μmol⁻¹ (27) and then annealed, individually or in combination, to 50 μg of total βTC-3 or Y1 RNA for 1 h at 60°C. Primer extension was then performed as previously described and extension products were visualized by electrophoresis on polyacrylamide/urea/TBE (1× TBE = 89 mmol/l Tris, 89 mmol/l boric acid, and 2 nmol/l EDTA) gels (26). The IGRP primer used here was different than that used in previous IGRP primer extension analyses (21), but again gave three major extension products (Fig. 4). The sizes of these products were consistent with previous estimates of the major IGRP gene transcription start sites.

To analyze protein binding to the IGRP promoter, promoter methylation by DMS in vitro and in situ was compared; differences were interpreted as indicative of trans-acting factor binding in situ. Methylation of genomic DNA and strand cleavage by piperidine and heating were based on the method of DNA sequencing by Maxam and Gilbert (32) and yielded DNA products predominantly cleaved at guanine (G) residues with only weak cleavage of adenine (A) residues. The relative abundance of piperidine-cleaved DNA fragments, and hence the level of DNA methylation, was then assayed using the LMPCR footprinting technique (Fig. 5). For in situ methylation, βTC-3 and Y1 cells were grown in 55-cm² cell culture dishes until the cells were ∼90% confluent. The DNA was methylated by addition of 25 μl DMS to 10 ml of cell culture medium (including serum) and incubation for 2.5 min at room temperature. The medium was then aspirated and 10 ml of room temperature phosphate-buffered saline (PBS) was added. This PBS was immediately aspirated, and the cells were then washed with 10 ml PBS for 1 min at room temperature. The PBS was then aspirated, and 2 ml lysis solution (20 mmol/l Tris-HCl, pH 8.0; 20 mmol/l NaCl; 20 mmol/l EDTA; 1% SDS; and 600 μg proteinase K per ml) were added. High quality genomic DNA, as required for LMPCR, was then isolated as previously described (33,34). Briefly, the viscous lysate that formed was scraped from each dish into 2 ml lysis dilution buffer (150 mmol/l NaCl and 5 mmol/l EDTA). Proteinase K digestion then proceeded at 37°C for 3 h. The samples were then extracted with phenol and then phenol:chloroform (1:1) before ethanol precipitation. The nucleic acids were resuspended in 400 μl of a solution containing 50 mmol/l Tris-HCL (pH 7.5), 10 mmol/l MgCl₂, 5 mmol/l dithiothreitol (DTT), and 0.125 μg RNase A per μl and incubated for 1 h at 37°C. Each sample was then phenol:chloroform extracted; the DNA was ethanol precipitated and then resuspended in 180 μl water before piperidine cleavage (see below).

Genomic DNA was isolated from control, non-DMS-treated cells grown in 55-cm² cell culture dishes using the same method. After ethanol precipitation, this nonmethylated DNA was either resuspended in 10 mmol/l Tris-HCl (pH 8.0) and 1 mmol/l EDTA to a final concentration of 1 μg/μl for use as a noncleaved control in the LMPCR assay (see below) or in 99.5 μl water for in vitro methylation. Genomic DNA was methylated in vitro by addition of 0.5 μl DMS for 30 s at room temperature. Ice cold stop solution (25 μl of 1.5 mol/l NaOAc, pH 7.0; 1 mol/l β-mercaptoethanol; and 100 μg tRNA per ml) was then added, followed by 375 μl ethanol to precipitate the DNA. This DNA was then resuspended in 180 μl water before piperidine cleavage (see below). The typical yield of genomic DNA from βTC-3 and Y1 cells was ∼200 and ∼100 μg per 55-cm² cell culture dish, respectively.

Piperidine was used to cleave DNA methylated in situ or in vitro. Methylated DNA that had been resuspended in 180 μl water, as described above, was incubated with 20 μl piperidine at 90°C for 30 min. The samples were then lyophilized to remove the piperidine. The DNA was resuspended in 100 μl water and lyophilized another two times. The DNA was then resuspended in 100 μl water, centrifuged briefly to pellet any debris, and ethanol precipitated. Finally, the DNA was resuspended in 10 mmol/l Tris-HCl (pH 8.0) and 1 mmol/l EDTA to a final concentration of 1 μg/μl.

The LMPCR technique (24,35) (Fig. 5) was used to analyze protein DNA interactions at the IGRP promoter in intact βTC-3 and Y1 cells by quantitating the level of G residue methylation in situ and in vitro. Four sets of oligonucleotides complementary to the mouse IGRP promoter were designed for the analysis of both strands of the IGRP promoter from approximately bases −50 to −350 (Table 1). Each set comprised three nested oligonucleotides: oligonucleotide 1 was used in the first strand synthesis reaction, oligonucleotide 2 was used for the PCR-based amplification step, and oligonucleotide 3 was used for the labeling reaction (24,35) (Fig. 5). These oligonucleotides were designed such that the calculated melting temperatures (http://research.nwfsc.noaa.gov/protocols/oligoTMcalc.html) increased progressively by 2–8°C. The relative calculated melting temperatures for oligonucleotides 1, 2, and 3 were similar among all four sets of oligonucleotides (Table 1). Two additional generic oligonucleotides were used for the linker ligation and amplification steps (24) (Fig. 5). These consisted of a 25-bp oligonucleotide (25mer), 5′ GCGGTGACCCGGGAGATCTGAATTC 3′, and an 11-bp oligonucleotide (11mer), 5′ GAATTCAGATC 3′ (24). All oligonucleotides, with the exception of the 11mer, which was not used in a DNA polymerization step, were purified by electrophoresis on 15% polyacrylamide/urea/TBE gels (27).

The first strand synthesis reaction (35) (Fig. 5) contained 2 μg genomic DNA template in a final reaction volume of 15 μl, which comprised 40 mmol/l NaCl, 10 mmol/l Tris-HCl (pH 8.9), 5 mmol/l MgSO₄, 0.01% gelatin, 0.24 mmol/l dNTPs, 0.01 μmol/l oligonucleotide 1, and 0.033 units of Vent DNA polymerase per μl. A MiniCycler (MJ Research) was used for all steps of the LMPCR. The first strand synthesis conditions were 95°C for 5 min, 47°C for 30 min, and 76°C for 10 min. Next, for the linker ligation, 22.5 μl of a linker cocktail (49 mmol/l Tris-HCl, pH 7.5; 84 μg/ml bovine serum albumin; 13.3 mmol/l MgCl₂; 33.3 mmol/l dithiothreitol; 1.68 mmol/l ATP; 2.2 μmol/l annealed 25mer and 11mer linkers; and 0.1 units T4 ligase per μl) were added. The linker ligation proceeded for 12–20 h at 16°C. The nucleic acids were then precipitated by adding 5 μg tRNA (0.5 μl of 10 μg/μl), 3.8 μl 3 mol/l NaOAc (pH 5.2), and 100 μl ethanol.

For the amplification reaction (35) (Fig. 5), the precipitated nucleic acids were dissolved in 35 μl water, and then 15 μl amplification cocktail was added to obtain final concentrations of 40 mmol/l NaCl, 20 mmol/l Tris-HCl (pH 8.9), 5 mmol/l MgSO₄, 0.01% gelatin, 0.1% Triton X-100, 0.2 mmol/l dNTPs, 0.1 μmol/l oligonucleotide 2, 0.1 μmol/l 25mer oligonucleotide, and 0.03 units Vent DNA polymerase per μl. The conditions for the amplification reaction were one cycle of 95°C for 3 min, 62°C for 2 min, and 76°C for 3 min, followed by 15 cycles of 95°C for 1 min, 62°C for 2 min, and 76°C for 3 min plus 5 s added per cycle.

For the final step in the LMPCR assay (35) (Fig. 5), 1 pmol oligonucleotide 3 was end labeled using [γ-³²P]ATP and T4 polynucleotide kinase. The labeled oligonucleotide was purified on G-50 Sephadex spin columns (27), precipitated with ethanol, and then resuspended in 2.5 μl of 40 mmol/l NaCl, 20 mmol/l Tris-HCl (pH 8.9), 5 mmol/l MgSO₄, 0.01% gelatin, 0.1% Triton X-100, 2 mmol/l dNTPs, and 0.2 units Vent per μl. This solution was then added to the amplification reaction at 4°C. The conditions for the LMPCR labeling reaction consisted of two cycles of 95°C for 3 min, 66°C for 2 min, and 76°C for 10 min. The entire reaction (52.5 μl) was then mixed with 52.5 μl of 10 mmol/l Tris-HCl (pH 8.0), 1 mmol/l EDTA, and 295 μl stop solution (10 mmol/l Tris-HCl, pH 7.5), 4 mmol/l EDTA, 260 mmol/l NaOAc (pH 7.0), and 67 μg tRNA per ml and extracted with 350 μl phenol:chloroform (1:1). The aqueous phase (370 μl) was then removed and the nucleic acids were precipitated by addition of 1 ml ethanol. For gel electrophoresis, the precipitated nucleic acids were resuspended in 10 μl loading buffer (80% formamide, 1X TBE, 0.04% xylene cyanol, and 0.04% bromophenol blue) and electrophoresed on 6% polyacrylamide/urea/TBE gels (27).

Two important controls in the LMPCR assay are TE (10 mmol/l Tris-HCl [pH 7.5] and 1 mmol/l EDTA) without DNA and genomic DNA that has not been methylated; these controls test for the presence of background bands from the LMPCR reactions, which would not be specific for the DMS-methylated DNA. No background bands were detected with either of these controls in our experiments (data not shown).

Complementary oligonucleotides representing the wild type (WT) IGRP promoter sequence between −247 and −220, containing a putative HNF-3 binding site (Figs. 10 and 11A), were gel purified, annealed, and then labeled with [α-³²P]dATP using the Klenow fragment of Escherichia coli DNA polymerase I to a specific activity of ∼2.5 μCi/pmol (22). Labeled oligonucleotides (∼7 fmol) were incubated with βTC-3 cell nuclear extract (4 μg), prepared as previously described (22), for 10 min at room temperature in the presence of 20 mmol/l HEPES (pH 7.8), 50 mmol/l KC1, 1 mmol/l DTT, 1 μg poly(dG-dC) · poly(dG-dC), and 10% glycerol (vol/vol) in a final volume of 20 μl. After incubation, the reactants were loaded onto a 6% polyacrylamide gel and electrophoresed at room temperature for 90 min at 150 V in a buffer containing 25 mmol/l Tris-HC1 (pH 7.8), 190 mmol/l glycine, and 1 mmol/l EDTA (22). After electrophoresis, the gels were dried and exposed to Kodak XAR5 film, and binding was analyzed by autoradiography. For competition experiments, the indicated unlabeled competitor DNA (25-fold molar excess) was mixed with the radiolabeled oligomer before addition of nuclear extract. For supershift experiments, specific antisera (1 μl) raised against HNF-3α, -3β, or -3γ were preincubated with βTC-3 nuclear extract for 15 min at room temperature before the addition of the labeled IGRP WT oligonucleotide probe and incubation for an additional 10 min at room temperature.

We have previously demonstrated that the proximal IGRP promoter region located between −306 and +3, relative to the transcription start site, is sufficient to confer maximal basal IGRP-CAT fusion gene expression in HIT cells (21). Figure 1A shows that this region of the IGRP promoter confers a level of basal reporter gene expression that is ∼30% of that obtained with the well- characterized rat insulin II gene promoter (36,37) in HIT cells. In another islet β-cell–derived cell line, namely βTC-3 cells, this difference between the relative strengths of the IGRP and insulin promoters is slightly reduced (Fig. 1B).

A series of truncated IGRP-CAT fusion genes was constructed to begin to identify the cis-acting elements in the IGRP promoter that impart basal gene expression. The level of basal reporter gene expression directed by these fusion genes was analyzed by transient transfection in HIT cells (Fig. 2) and in βTC-3 cells (Fig. 3). Figure 2A shows that reporter gene expression did not change when the promoter region between −306 and −273 was deleted. However, deletion of the promoter regions −273 to −254 and −254 to −234 caused a marked reduction in reporter gene expression (Fig. 2A). Further deletion of the sequence between −234 and −197 had no effect (Fig. 2A). Although the level of reporter gene expression conferred by the −197 IGRP-CAT fusion gene was only ∼10% of that obtained with the full-length proximal promoter, CAT expression was still easily detectable. Thus, further deletion of the promoter sequence between −197 and −66 revealed the presence of additional cis-acting elements −172 to −129, −129 to −97, and −97 to −66, all of which contribute to basal reporter gene expression (Fig. 2B). The 5′ deletion analyses shown in Figs. 2A and B were repeated using a separate batch of HIT cells that we observed had a less rounded morphology. Although most of the results were the same between both batches of HIT cells, two differences were observed (Fig. 2C). Thus, in this alternate batch of HIT cells, a reduction in reporter gene expression was detected upon deletion of the promoter sequence between −306 and −273 (Fig. 2C). Such a decrease was observed in our initial characterization of the upstream IGRP promoter (21). In addition, in this alternate batch of HIT cells, an increase in reporter gene expression was detected upon deletion of the promoter sequence between −197 and −172 (Fig. 2C). These differences among batches of HIT cells presumably reflect variations in trans-acting factor expression and serve to emphasize the limitations in the use of cell lines as models of in vivo gene regulation.

Because of the variability in the results of the IGRP-CAT 5′ deletion analysis in HIT cells, these experiments were repeated in another islet β-cell–derived cell line, namely βTC-3 cells. Figure 3A shows that successive deletion of the IGRP promoter sequence between −306 and −234 resulted in a progressive reduction in reporter gene expression in βTC-3 cells with a pattern similar to that seen in HIT cells (Fig. 2). As in HIT cells (Figs. 2A and C), the effect of deleting the IGRP promoter sequence between −306 and −273 on reporter gene expression in βTC-3 cells was somewhat variable (Fig. 3A). Successive deletion of the IGRP promoter sequence between −197 and −97 also resulted in a progressive reduction in reporter gene expression in βTC-3 cells (Fig. 3B) with a pattern that was again similar to that seen in HIT cells (Fig. 2). However, in contrast to the data obtained in HIT cells, a clear reduction in reporter gene expression was obtained upon deletion of the IGRP promoter sequence between −197 and −172 in βTC-3 cells (Fig. 3B). This is one of the regions of the IGRP promoter that, when deleted, resulted in a variable change in fusion gene expression in HIT cells (compare Figs. 2A and C).

A key requirement for in situ footprinting is that the gene of interest must be expressed in the cell line under investigation. The hamster IGRP gene has not been cloned; therefore, whether the endogenous IGRP gene is expressed in HIT cells remains to be determined. However, primer extension analysis showed that the endogenous IGRP gene is expressed in the mouse βTC-3 cell line. Several IGRP gene transcription start sites were detected in βTC-3 cells, the most prominent of which were clustered over a ∼5-bp region (Fig. 4). Primer extension analysis also showed that the endogenous IGRP gene is not expressed in the mouse adrenocortical-derived Y1 cell line (Fig. 4). Y1 cells were therefore used as a control in in situ footprinting experiments (see below) to permit a comparison of protein binding to the IGRP promoter in cells that express the endogenous gene (βTC-3) and those that do not (Y1). Cyclophilin A gene expression was also assayed in both cell lines as a positive control for the integrity of the RNA. Figure 4 shows that the cyclophilin A gene is expressed at similar levels in both cell types. With both βTC-3 cell and Y1 cell RNA, the cyclophilin A primer gave a cluster of extension products between 70 and 72 bp (Fig. 4); the published transcription start site predicts a product of 70 bp with this primer (29).

To analyze protein binding to the IGRP promoter, promoter methylation by DMS in vitro and in situ was compared; differences were interpreted as indicative of trans-acting factor binding in situ. DMS methylates DNA on A and G residues and freely permeates cell membranes (38). Methylated genomic DNA was cleaved using piperidine under conditions that yielded DNA products predominantly cleaved at G residues with only weak cleavage of A residues (32). The relative abundance of these piperidine-cleaved DNA fragments, and hence the level of DNA methylation, was then assayed using the LMPCR footprinting technique (24,35) (Fig. 5). Thus, the individual bands in Figs. 6–9,FIG. 7,FIG. 8,FIG. 9 are indicative of the accessibility to DMS of mainly G residues in the IGRP promoter.

The binding of a trans-acting factor to a gene promoter can result in either an increase or decrease in the level of methylation (24,35,38). A decrease in the level of methylation is interpreted to occur as a result of steric hindrance, whereas an increase is interpreted to occur if the trans-acting factor causes a change in DNA conformation, leading to greater accessibility of a base to DMS (38,39). Differences in the frequency of methylation at a given G residue in situ result in the over- or underrepresentation of that piperidine-cleaved fragment in the subsequent PCR amplification (24,35) (Fig. 5). As an additional control, Figs. 6–9,FIG. 7,FIG. 8,FIG. 9 show a comparison of the level of IGRP promoter methylation by DMS in βTC-3 and Y1 cells both in vitro and in situ. Because the endogenous IGRP gene is not expressed in Y1 cells (Fig. 4), differences between the pattern of IGRP promoter methylation in vitro and in situ, which are specific to βTC-3 cells, may be indicative of the binding of trans-acting factors that stimulate IGRP gene expression.

A total of four sets of nested oligonucleotide primers were used to analyze trans-acting factor binding to the IGRP promoter in situ (Table 1). With the in situ footprinting method used, each set of primers allowed analysis of protein binding to ∼200 bp of promoter sequence. Thus, an analysis of trans-acting factor binding to the IGRP promoter region between approximately −350 and +1 required designing two sets of primers for both the sense and antisense strands of the promoter. These primer sets were designed such that the IGRP promoter regions analyzed overlapped. Figures 6 and 7 show the in situ footprinting analysis of trans-acting factor binding to the sense strand of the IGRP promoter; Figs. 8 and 9 show the in situ footprinting analysis of trans-acting factor binding to the antisense strand of the IGRP promoter. The positions of specific G residues in the IGRP promoter sequence are indicated, along with increases or decreases in the level of IGRP promoter methylation comparing in vitro and in situ methylated βTC-3 cell genomic DNA. Only changes that were consistent between duplicate experiments are indicated. There were some consistent changes in G residue methylation when in vitro and in situ methylated Y1 genomic DNA was compared, but these are not indicated because these changes were distinct from those seen with βTC-3 cell genomic DNA. An exception was a decreased methylation of the G residues at −61, −66, and −68 that occurred in both cell types (Fig. 8). Some of these changes in Y1 genomic DNA methylation may reflect the binding of factors that actively suppress IGRP gene transcription in Y1 cells because the DMS methylation technique does not detect the presence of nucleosomal packaging (40).

Figure 10 shows a summary of the in situ footprinting data derived from Figs. 6–9,FIG. 7,FIG. 8,FIG. 9. The multiple trans-acting factor binding sites that were identified in βTC-3 cells by in situ footprinting correlated with regions of the IGRP promoter identified as being important for basal IGRP-CAT fusion gene expression in βTC-3 cells (Fig. 3). Thus, the decreases in basal IGRP-CAT fusion gene expression upon deletion of the promoter regions −306 to −273, −273 to −254, and −254 to −234 (Fig. 3A) are all associated with trans-acting factor binding in situ. Similarly, the decreases in basal IGRP-CAT fusion gene expression upon deletion of the promoter regions −197 to −172, −172 to −129, and −129 to −97 (Fig. 3B) are also associated with trans-acting factor binding in situ.

Figure 10 also shows the position of putative trans-acting factor binding sites in the IGRP promoter, as predicted by the MatInspector sequence analysis software (41), which overlap with or are close to G residues whose level of methylation is altered by trans-acting factor binding in situ. Several of these trans-acting factors have not been previously implicated in the regulation of gene expression in islets. In contrast, HNF-3, which is predicted to bind the IGRP promoter region between −243 and −230 (Fig. 10), has been shown to contribute to basal glucagon, GLUT2, and pdx-1 gene expression (see discussion). The gel retardation assay was used to investigate whether HNF-3 can bind to this region of the IGRP promoter in vitro. When a labeled double-stranded oligonucleotide, designated IGRP WT, representing the WT IGRP promoter sequence from −247 to −220 (Fig. 11A), was incubated with nuclear extract prepared from βTC-3 cells, a broad protein-DNA complex was detected (Fig. 11B). To determine whether this protein-DNA complex contained HNF-3, the effect of preincubating βTC-3 nuclear extract with antisera specific for the α, β, and γ isoforms of HNF-3 on the migration of this complex was investigated (Fig. 11B). Both the polyclonal HNF-3α and -3β antisera reduced the formation of the broad protein-DNA complex and resulted in the generation of distinct but weak supershifted protein-DNA complexes (Fig. 11B; see arrows). By contrast, the HNF-3γ antiserum had no effect on DNA binding (Fig. 11B).

A competition experiment, in which a 25-fold molar excess of unlabeled DNA was included with the labeled probe, was used to correlate protein binding in vitro with the changes in methylation detected in situ. The IGRP WT oligonucleotide competed effectively for the binding of the factors in broad protein-DNA complex (Fig. 11C; see arrow). By contrast, an oligonucleotide, designated IGRP MUT (Fig. 11A), which contained point mutations of the two base pairs within the HNF-3 motif that were implicated in protein binding in situ (Fig. 10), showed a reduced ability to compete with the labeled probe for protein binding (Fig. 11C). Thus, the formation of this protein-DNA complex in vitro correlates with trans-acting factor binding in situ. It should be noted that the critical, nonvariant bases in the HNF-3 binding motif are A and thymine (T) residues (Fig. 11A), whereas the in situ footprinting method mainly detects protein-DNA contacts at G residues. Thus, because the double-stranded IGRP mutant (MUT) oligonucleotide contains mutations at G residues that are not essential for HNF-3 binding, just high-affinity binding, this presumably explains why this oligonucleotide still competes partially for binding of the protein-DNA complex. In summary, these in vitro protein binding experiments support the hypothesis that HNF-3 is binding the IGRP promoter in situ.

Although multiple cis-acting elements are required for maximal IGRP fusion gene expression in HIT cells (Fig. 2) and βTC-3 cells (Fig. 3), it remains to be determined whether any of these elements are particularly important for gene expression in islet cells, in contrast to those that can enhance basal gene expression in all cell types. We have taken an indirect approach to address this issue by analyzing the cis-acting elements that are required for basal G6Pase catalytic subunit gene expression in HIT cells. Thus, an alignment of the IGRP promoter sequence with that of the mouse G6Pase catalytic subunit promoter revealed a stretch of ∼50% identity, not including spaces, over the −252 to +1 region (21). We hypothesized that some of the conserved promoter regions may represent cis-acting elements that are important for not only both basal IGRP and G6Pase catalytic subunit gene expression in islet cells but also basal G6Pase catalytic subunit gene expression in hepatoma cells. To explore this hypothesis, the level of basal reporter gene expression directed by a series of truncated G6Pase catalytic subunit-CAT fusion genes was analyzed by transient transfection in HIT cells. Figure 12 shows the surprising result that basal G6Pase catalytic subunit-CAT fusion gene expression in HIT cells is conferred entirely by the proximal −85 to +66 G6Pase promoter region. This contrasts with a similar analysis in the HepG2 hepatoma cell line that demonstrates the involvement of multiple cis-acting elements in the −751 to −66 G6Pase catalytic subunit promoter region that contribute to basal fusion gene expression (25,26). The simplest interpretation of this result is that none of the cis-acting elements identified in the IGRP promoter between −306 and −66, which contribute to basal gene expression in HIT cells, are present in the equivalent region of the G6Pase catalytic subunit promoter. This would imply that the 50% identity between the IGRP and G6Pase catalytic subunit promoters is not of functional significance with respect to the control of basal gene expression. Although this experiment failed to distinguish cis-acting elements that might be selectively important for islet gene expression, the result is consistent with the observation that the IGRP promoter is completely inactive in HepG2 cells (21), again despite the high degree of apparent identity with the G6Pase catalytic subunit promoter.

An alignment of the mouse IGRP promoter sequence with that of the mouse G6Pase catalytic subunit reveals a stretch of ∼50% identity, not including spaces, over the −252 to +1 promoter region (21). However, the G6Pase catalytic subunit and IGRP gene promoters show highly selective activity in the liver-derived HepG2 and islet-derived HIT cell lines (21). This study began to explore the molecular basis for the islet-specific expression of the IGRP gene through a combination of transient transfection and in situ footprinting analyses. The results demonstrated that multiple regions of the IGRP promoter are required for maximal IGRP-CAT fusion gene expression, and importantly, these functional data correlate with the location of trans-acting factor binding sites in the IGRP promoter, as assessed by in situ footprinting.

Figure 10 shows the position of putative trans-acting factor binding sites in the IGRP promoter, as predicted by the MatInspector sequence analysis software (41), that overlap with or are close to G residues whose level of methylation is altered by trans-acting factor binding in situ. Based on comparison with the consensus binding sequences described in the TRANSFAC 4.0 matrices used by the MatInspector software (41), the IGRP promoter may bind the trans-acting factors Gfi-1 (42), Brn-2 (43), Myb (44), TCF11 (45), deltaEF1 (46), and NF-1 (47). Interestingly, none of these factors have previously been shown to regulate the expression of other genes in islets; however, the presence of these factors in islet cells remains to be demonstrated. In contrast, the IGRP promoter also contains putative binding sites for HNF-3 (48), NF-AT (49), and Sp1 (50), all of which have been implicated in the regulation of islet cell gene transcription. Thus, both HNF-3α (51) and HNF-3β (52) are important for the regulation of glucagon gene expression. In addition, HNF-3β has been implicated in the regulation of pdx-1 (53) and islet GLUT2 (54) gene expression and has also recently been shown to interact with NF-AT to form a calcium response element in the glucagon gene promoter (55). Finally, Sp1 is important in gastrin gene expression (56); gastrin is transiently expressed in fetal pancreatic islet cells, where it may play a role in islet cell neogenesis (57). Detailed studies on the insulin (36,37), glucagon (58), and islet amyloid polypeptide (amylin) (59) promoters suggest that although islet-enriched trans-acting factors do contribute to the expression of these genes, maximal islet-specific expression is conferred by the particular arrangement of trans-acting factors binding these promoters rather than by a single, islet-specific trans-acting factor. Based on the data available so far, this paradigm would also appear to hold true for the IGRP promoter.

The analysis of basal IGRP-CAT fusion gene expression by progressive 5′ deletion of the promoter indicated that the sequence between −97 and +3 confers very low basal fusion gene expression in the absence of more distal sequences (Figs. 2 and 3). However, the in situ footprinting analysis revealed several trans-acting factor binding sites in this location (Fig. 10). Interestingly, two of these binding sites are E-box motifs (Fig. 10), elements that are known to be important for activity of the insulin promoter (36,37). The E-box motif in the insulin promoter binds a dimeric complex composed of an islet-enriched factor called BETA2/NeuroD and a ubiquitous factor, either E2A or HEB (36,37). Site-directed mutagenesis of the proximal E-box motifs, in the context of an otherwise intact promoter, will reveal the relative contribution of these elements to basal IGRP gene expression. A similar approach in the analysis of other gene promoters has led to the identification of cis-acting elements that are important for basal gene expression, but that are almost inactive in the absence of distal elements (60).

The proximal IGRP promoter region located between −306 and +3 is sufficient to confer maximal basal IGRP-CAT fusion gene expression in HIT cells (21). However, establishing that this region of the IGRP promoter is also sufficient to confer islet-specific expression in vivo will require the generation of transgenic mice that express a reporter gene whose expression is directed by this promoter fragment. Magnuson and colleagues have previously shown that the proximal region of the upstream glucokinase promoter, located between −280 and +14, is sufficient to confer maximal basal expression of that gene in HIT cells (61) and appropriate transgene expression in transgenic mice (62). Interestingly, Magnuson and colleagues have also shown that multiple cis-acting elements within the proximal region of the upstream glucokinase promoter are required for maximal fusion gene expression in HIT cells (61), as is the case with the IGRP promoter (Figs. 2 and 3).

Type 2 diabetes is characterized by defects in insulin secretion, insulin-dependent peripheral glucose utilization, and hepatic glucose production (63). As a consequence of insulin resistance, the ability of insulin to stimulate peripheral glucose utilization and repress hepatic glucose production is reduced (63). Whether the initial event in the development of type 2 diabetes is the onset of insulin resistance or an islet cell defect is still controversial (64,65). However, it is apparent that the transition from an insulin-resistant state to frank diabetes occurs when the rate of insulin secretion by the β-cells of the islet is unable to balance the degree of insulin resistance (64,65). The causes of the defect in insulin secretion in type 2 diabetes are unclear (66,67). However, in some forms of maturity-onset diabetes of the young, aberrant insulin secretion can arise because of mutations in the genes encoding the transcription factors HNF-1, HNF-4, and pdx-1 (68–70). There are no apparent binding sites for HNF-1 or HNF-4 in the IGRP promoter (Fig. 10). However, the region of the IGRP promoter located between −306 and −273 contains a TAAT motif, which forms the core of the DNA sequence recognized by pdx-1 and other homeodomain proteins, although the MatInspector software identified this sequence as a potential binding site for the homeodomain protein Brn-2. There are several trans-acting factor binding sites in the IGRP promoter that were identified by in situ footprinting, as shown in Fig. 10, for which no associated trans-acting factors were identified by the MatInspector software. The most exciting interpretation of this observation is that these may represent binding sites for novel islet-enriched trans-acting factors. The search for such proteins has gained increased importance, considering the realization that these factors are not only potential diabetogenes but are also important for pancreatic development (36,71).

Data analysis was performed in part through the use of the VUMC Cell Imaging Resource (CA68485 and DK20593). The research was supported by a grant from the Juvenile Diabetes Foundation (to R.M.O’B.) and the Vanderbilt Diabetes Core Laboratory (P60 DK20593). B.T.S. and L.A.H. are supported by the Vanderbilt Molecular Endocrinology Training Program (5 T 32 DK07563-12); C.C.M. is supported by the Vanderbilt Viruses, Nucleic Acids and Cancer Training Program (5T32 CA09385-17).

We thank Roland Stein and Eva Henderson for providing the HIT cell line and Shimon Efrat for providing the βTC-3 cell line. We also thank Steve Okino for assistance with the LMPCR assay.