- Email: [email protected]
Involvement of the Homeodomain-Containing Transcription Factor PDX-1 in Islet Amyloid Polypeptide Gene Transcription
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
229, 746–751 (1996)
1875
Involvement of the Homeodomain-Containing Transcription Factor PDX-1 in Islet Amyloid Polypeptide Gene Transcription Hirotaka Watada,1 Yoshitaka Kajimoto,2 Hideaki Kaneto, Taka-aki Matsuoka, Yoshio Fujitani, Jun-ichi Miyazaki,* and Yoshimitsu Yamasaki First Department of Medicine, and *Department of Nutrition and Physiological Chemistry, Osaka University School of Medicine, Suita 565, Japan Received November 7, 1996 The AT-rich cis-motif A elements of the insulin gene promoter contribute to directing the gene’s expression to pancreatic b-cells, bound by a homeodomain-containing transcription factor, PDX-1/IPF1/STF-1/ IDX-1. The islet amyloid polypeptide (IAPP; amylin) gene, which is also expressed in limited tissues such as pancreatic b- and d-cells, contained similar AT-rich sequences in its regulatory sequences. To understand the molecular basis of IAPP gene regulation, we evaluated the possible physiological significance of the motif in human IAPP gene regulation. All of the three typical A element-like sequences that matched the CT-box consensus (AT-1, -207/-202; AT-2, -154/-142; and AT-3, -88/-83) were shown to bind specifically to a nuclear factor in the b-cell-derived MIN6 cells, which was subsequently identified immunologically as the insulin gene transcription factor PDX-1. When the promoter activity was examined in MIN6 cells, the disruption of AT-1 or AT-3 but not of AT-2 caused a marked reduction in the IAPP gene promoter. Thus, despite the observation that all the three A element-like regions could bind to PDX-1, the AT-2 site may not be involved in mediating the PDX-1 action in vivo. These observations suggest the involvement of PDX-1 in human IAPP gene regulation, which seems to be mediated through at least two A elementlike cis-motifs in the gene promoter. q 1996 Academic Press
INTRODUCTION
Islet amyloid polypeptide (IAPP), also known as amylin, is a 37-amino acid peptide of the calcitonin gene family isolated from amyloid deposits of non-insulin-dependent diabetic pancreas and insulinoma (1). In pancreatic b-cells, IAPP is co-stored with insulin in the secretory granules and coreleased with insulin into the circulation in response to a variety of secretagogues. While its physiological role is yet to be clarified, IAPP has been suggested to be toxic to b-cells and to be involved in the development of non-insulin-dependent diabetes (1-3). Like insulin, the major site of IAPP synthesis is the islet b-cell (1). However, unlike insulin, detectable quantities of IAPP are also produced in other cells including d cells in pancreatic islets (4), stomach, duodenum, and jejunum (5). The similar but slightly wider tissue distribution pattern of IAPP suggests that some, if not all, of the gene regulating machinery of the insulin gene may be shared with the IAPP gene. The 5*-flanking region of the human IAPP gene contains multiple sequences similar to the insulin gene A elements, the cis-motif important for the b-cell specific expression of the gene (6-10). A group of proteins that belong to the homeodomain family of transcription factors (11-15) were shown to bind to the motif. Among them, PDX-1/IPF1/STF-1/IDX-1 (13-15), is 1 Recipient of a fellowship and grant from the Japan Society for the Promotion of Science for Japanese Junior Scientists. 2 Corresponding author: First Department of Medicine, Osaka University School of Medicine, 2-2 Yamadaoka, Suita City, Osaka Pref. 565, Japan. Fax: /81-6-879-3639. E-mail: [email protected]. Abbreviations: IAPP, islet amyloid polypeptide; bp, base pair(s); PCR, polymerase chain reaction.
746 0006-291X/96 $18.00 Copyright q 1996 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 5834
/
6914$$$141
12-06-96 06:14:14
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
the factor for which physiological significance has been demonstrated (16). PDX-1 appears before insulin during the ontogeny of the mouse pancreas, where it eventually becomes restricted to the b cells in the adult (13, 17). As a step toward elucidating transcriptional control of the IAPP gene, we investigated whether the A-element-like motif and the insulin gene transcription factor PDX-1 are involved in the regulation of IAPP gene expression. MATERIALS AND METHODS Preparation of nuclear extracts and gel mobility-shift analyses. Nuclear extracts were prepared, following the procedure described by Sadowski and Gilman (18). The binding reactions were performed as described previously (19) using 3 mg of each nuclear extract. The samples were fractionated on a 5% polyacrylamide gel pre-cooled to 4 7C in 0.5 1 TBE (45 mM Tris base, 45 mM boric acid, 1 mM EDTA). Reporter gene constructs. The reporter plasmid pA3hIAPPLuc/-410 was generated by polymerase chain reaction (PCR) using human genomic DNA as a template, with subsequent ligation to the HindIII pre-digested pA3Luc plasmid (20). The PCR procedures were performed. The plasmids pA3hIAPPLuc/-410/MAT-1, pA3hIAPPLuc/-410/MAT-2, and pA3hIAPPLuc/-410/MAT-3 were prepared by PCR-mediated site-directed mutagenesis. These reporter plasmids were designed to contain 32 bp of the exon 1 untranslated region in addition to the 5*-flanking sequences. Gene transfer studies. MIN6 cells (21) and HepG2 cells were cultured as described previously (21, 22). For gene transfection, cells were replated in six-well tissue culture plates 24 h before transfection. One microgram of each reporter plasmid and 1 mg of the pEFb-Gal plasmid were cotransfected into the cells by the lipofection method using lipofectAMINE reagent (Life Technologies, Tokyo, Japan). For evaluation of PDX-1 effects, 0.7 mg of PDX-1 (IPF1) expression plasmid, pcDNA3IPF1 or pcDNA3IPF1-A containing the PDX-1 coding region in the reverse direction (19) was cotransfected into HepG2 cells with 0.7 mg of each reporter plasmid and 0.7 mg of the pEFb-Gal plasmid. Transfected cells were incubated for 24 h, then fresh medium was added. After another 24-h incubation, luciferase and b-galactosidase assays were performed as we described previously (22), allowing normalization of the luciferase results with respect to transfection efficiency.
RESULTS
PDX-1 has high affinity for each A-element-like sequence. The regulatory region of the human IAPP gene (9, 10) contains three regions that match the consensus sequence for the A elements (CT-box consensus: C/TTAATG). They were located at -88/-83, -154/-142, and, in reverse orientation, -207/-202. For the sake of convenience, we call those three regions AT1, AT-2, and AT-3, respectively. To investigate the possible role of each as a cis-acting element in the IAPP gene regulation, we first performed gel mobility-shift analyses using nuclear extracts isolated from the pancreatic b-cell derived MIN6 cells. As shown in Fig. 1, the oligonucleotide probes containing AT-1, AT-2, or AT-3 site and their flanking sequences became bound to two nuclear factors in MIN6 cells. Among the two bands, however, the upper ones seemed non-specific because they were influenced even by the addition of mutatedtype competitors (Fig. 1). To identify the transcription factor involved in the lower band formation, we investigated whether the anti-PDX-1 (IPF1) antibody could recognize the DNAprotein complexes. In contrast to the pre-immune serum which displayed no significant effects on the complex, the addition of anti-PDX-1 (IPF1) antiserum totally eliminated the lower bands (arrows; Fig. 1). These results suggested that all the three A-element-like regions (AT1, AT-2, and AT-3) in the human IAPP gene are bound by PDX-1 in MIN6 cells. Site-directed mutagenesis at A-element-like motifs of the IAPP gene promoter. Further to examine whether the A-element-like sequences play a role in generating promoter activity of the IAPP gene, we performed reporter gene analyses using a wild-type and mutant IAPP gene promoters linked to a luciferase reporter (Fig. 2A). The reporter plasmids were transfected into MIN6 cells in which the features of gene regulation and biosynthesis of IAPP resembled those of islet cells from normal animal (21) and IAPP non-producing HepG2 cells. Figure 2B presents the activities of the wild-type and mutant IAPP gene promoters in those two cell lines as relative luciferase activities. The results (Fig. 2B) showed that, while the IAPP gene promoters generally displayed poor activities in HepG2 cells, they generated much higher activities in MIN6 cells than did the control promoter-less plasmid pA3Luc. In terms of the effects of 747
AID
BBRC 5834
/
6914$$$141
12-06-96 06:14:14
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Analyses of protein bindings to A element-like regions. Gel-mobility shift analyses were performed to evaluate the specific protein bindings to the A element-like motifs in the human IAPP gene. The 5*-end-[32P]labeled probe was incubated with nuclear extracts isolated from MIN6 cells. Where specified, 50-fold excess of unlabeled competitors, or preimmune serum or anti-PDX-1 (IPF1) antiserum, was also added to the binding reactions. Arrows and arrowheads indicate the disappearance and supershift of the DNA–protein complexes by the antiserum, respectively. Nucleotide sequences of sense strands of the double-stranded oligonucleotide probes and competitors were as follows: AT-1 probe, ATGACACACCATTAACTGCACAAG; mutated-type AT-1 probe, ATGACACACCAGGAACTGCACAAG; AT-2 probe, GATGAGTTAATGTAATAATGACCC; mutated-type AT-2 probe, GATGAGTTCCTGTAAGCATGACCC; AT-3 probe GGATGGAAATTAATGACAGAGGCT, mutated-type AT-3 probe, GGATGGAAATTCCTGACAGAGGCT. The exposure time was 12 h.
the mutations, the results varied depending upon the mutation site; destruction of AT-1 and AT-3 caused major loss of the promoter activity in MIN6 cells (56% and 83% decrease, respectively) but not that of the AT-2 site (12% increase; Fig. 2B). No significant reduction of the promoter activities was observed with any mutations when examined with HepG2 cells (Fig. 2B). These results thus suggested that AT-1 and AT-3 but not AT-2 function as cisacting motifs for the IAPP gene promoter. Activation of IAPP gene promoter by PDX-1. To examine whether PDX-1 can function as a positive regulator of the IAPP gene transcription, PDX-1 expression was induced in non-bcell derived HepG2 cells. As shown in Fig. 3, this expression caused 4.0-fold increase of the 410-bp IAPP gene promoter activity. All three mutations introduced at any of the A-elementlike sequences caused some deleterious effects on the promoter activity. However, in agreement with the observations for the promoter activities in MIN6 cells (Fig. 2B), the destruction of AT-1 or AT-3 conferred more damage on the induction ratio by PDX-1 (2.5-fold, 1.5-fold, respectively) as compared with the AT-2 mutation (3.1-fold; Fig. 3). These results suggested that PDX-1 activates the IAPP gene promoter and that the activation is mediated mainly by AT-1 and AT-3 sites. DISCUSSION
In this study, we showed that PDX-1 can activate the human IAPP gene promoter and this is mediated by at least two of the A-element-like regions in the 5*-flanking sequence. As we recently characterized the human glucokinase gene b-cell-type promoter and identified PDX1 as an essential transcription factor for it (19), PDX-1 has now been shown to be an important 748
AID
BBRC 5834
/
6914$$$141
12-06-96 06:14:14
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. Effects of mutations of A element-like regions on promoter activities. (A) Structure of the reporter gene constructs used in the reporter gene analyses. The pA3IAPPLuc/-410 plasmid, shown as ‘‘Wild Type,’’ contained the 410-bp 5*-flanking sequence and 32-bp untranslated exon 1 sequences of the human IAPP gene. The three mutant reporter plasmids pA3hIAPPLuc/-410/MAT-1 (shown as ‘‘MAT-1’’), pA3hIAPPLuc/-410/MAT-2 (‘‘MAT-2’’), and pA3IAPPLuc/-420/MAT-3 (‘‘MAT-3’’) were prepared by introducing mutations to the AT-1, AT-2, and AT-3 regions, respectively. Substituted nucleotides are underlined. Sequences of the rat insulin I gene FLAT region and of the AT2 region are aligned to facilitate comparison. (B) Effects of the mutation introduced to AT-1 (shown as ‘‘MAT-1’’), AT-2 (‘‘MAT-2’’), and AT-3 (‘‘MAT-3’’) on the promoter activities in MIN6 cells (open bars) and HepG2 cells (filled bars). Each reporter plasmid was individually transfected into MIN6 cells or HepG2 cells, and luciferase activity was measured 48 h later. The results shown here are expressed as relative luciferase activities with that of the mutationless promoter (‘‘Wild Type’’) in MIN6 cells arbitrarily set at 1. They were normalized with respect to those of pA3RSVLuc in order to allow comparison of the results obtained with MIN6 cells and HepG2 cells. All data are presented as means { SE of at least four individual transfection experiments.
transcription factor for at least three b-cell-specific genes. This broad range of target genes for which PDX-1 functions as a transcription factor is consistent with the phenotype of the PDX-1 (IPF1)-knock out mouse demonstrating that PDX-1 is required for pancreas development (16). Recently, Bretherton-Watt et al. reported that PDX-1 bind the same motifs that we pointed out in this study, although, unlike us, they did not examine the function of the motifs (23). An interesting point that became evident in our present study was a discrepancy between the gel-mobility shift and reporter gene results obtained with respect to the AT-2 site. Among the three A-element-like motifs, only AT-1 and AT-3 appeared to play a physiological role in the IAPP gene transcription (Fig. 2B). In contrast, the AT-2 site, to which PDX-1 can also bind 749
AID
BBRC 5834
/
6914$$$141
12-06-96 06:14:14
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Effects of PDX-1 on IAPP promoter activities. Expression of PDX-1 in HepG2 cells was induced by transient transfection of an expression plasmid (pcDNA3IPF1 (24)), and the effects on the IAPP gene promoter activities were examined. The transfected plasmids (‘‘Wild-type’’, ‘‘MAT-1’’, ‘‘MAT-2’’, and ‘‘MAT-3’’) are described in the legend to Fig. 2. Each reporter plasmid was individually cotransfected into HepG2 cells with pcDNA3IPF1 (24) or the control plasmid pcDNA3IPF1-A. The results are expressed as n-fold induction over the promoter activity of each reporter plasmid in the control pcDNA3IPF1-A-transfected HepG2 cells. All data are presented as means { SE of four individual transfection experiments.
at least in vitro (Fig. 1), seemed not to be important, or to be less important, as a cis-acting motif in MIN6 cells (Fig. 2B & 3). Although this tends to be speculative, we may need to refer to the structural characteristics of this AT-2 region in order to discuss the reason for this discrepancy. When the sequences are compared with other b-cell specific promoters, we find that AT-2 is structurally similar to the A3/A4 (FLAT) region of the rat insulin I gene (Fig. 2A). According to reports by German et al. (24), the FLAT region is composed of two negatively interacting elements. Also, they pointed out that the phenotypes of the FLAT region mutation vary depending upon which nucleotides of the region are mutated. Although PDX1 was also shown to bind to the FLAT region as well as to the AT-2 element (6), these observations suggest that some factors other than PDX-1 may also bind to these regions and thus make it difficult to interpret their function. The in vitro gel-mobility shift assay may have failed to detect the band corresponding to this, for example, because of low affinity. Recently, Wang et al. reported that transcription of the rat IAPP gene is activated by the LIM domain homeobox protein isl-1 (25). They also performed gel-mobility shift assays using the A-element-like regions in the rat IAPP gene and thereby detected multiple bands that were specific to the CT-box (A element) consensus motif. They showed one of those contained isl1. In contrast to their observations, we as well as Bretherton-Watt et al. used the binding probe reproducing the human IAPP gene A element-like region and detected only PDX-1 as the binding factor. Although this did not totally exclude the possibility of isl-1 involvement, it suggested that PDX-1 is the dominant A element-like region-binding factor, at least in MIN6 cells (this study) and bTC3 cells (23). To clarify the physiological significance of PDX-1, we recently established a stable transfectant of aTC1 clone 6 cells that expresses PDX-1. Despite its failure to drive insulin, glucokinase or GLUT2 gene expression, the exogenously expressed PDX-1 could induce IAPP gene expression in aTC1 clone 6 cells (26). Also, the tissues expressing IAPP correspond very well to those expressing PDX-1; pancreatic b and d cells and duodenum (7, 8, 17-19). These observations, along with our present data, provide support for the physiological significance of PDX1 as a primary determinant of IAPP gene expression. ACKNOWLEDGMENTS We thank Dr. W. M. Wood of University of Colorado Health Science Center and Dr. D. R. Helinski of University of California San Diego for the pA3Luc and pA3RSVLuc plasmids. We also thank Ms. N. Fujita for the excellent technical assistance. 750
AID
BBRC 5834
/
6914$$$141
12-06-96 06:14:14
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
This work was supported in part by grants from Japan Diabetes Foundation (to Y.K.), and Kyowa Hakko Kogyo Co. Ltd. (to Y.K. and Y.Y.).
REFERENCES 1. Cooper, G. J. S. (1994) Endocr. Rev. 15, 163–201. 2. Lorenzo, A., Razzaboni, B., Weir, G. C., and Yankner, B. A. (1994) Nature 368, 756–760. 3. Verchere, C. B., D’ Alessio, D. A., Palmiter, R. D., Weir, G. C., Bonner-Weir, S., Baskin, D. G., and Kahn, A. E. (1996) Proc. Natl. Acad. Sci. U.S.A. 93, 3492–3496. 4. De Vroede, M., Foriers, A., Van de Winkel, M., Madsen, O., and Pipeleers, D. (1992) Biochem. Biophys. Res. Commun. 182, 886–893. 5. Nakazato, M., Asai, J., Kangawa, K., Matsukura, S., and Matsuo, H. (1989) Biochem. Biophys. Res. Commun. 164, 394–399. 6. German, M., Ashcroft, S., Docherty, K., Edlund, H., Edlund, T., Goodison, T., Imura, H., Kennedy, G., Madsen, O., Melloul, D., Moss, L., Olson, K., Permutt, M. A., Philippe, J., Robertson, J. P., Rutter, W. J., Serup, P., Stein, R., Steiner, D., Tsai, M.-J., and Walker, M. D. (1995) Diabetes 44, 1002–1004. 7. Ohlsson, H., Karlsson, O., and Edlund, T. (1988) Proc. Natl. Acad. Sci. U.S.A. 85, 4228–4231. 8. Ohlsson, H., Thor, S., and Edlund, T. (1991) Mol. Endocrinol. 5, 897–904. 9. Mosselman, S., Hoppener, J. W. M., de Wit, L., Soeller, W., Lips, C. J. M., and Jansz, H. S. (1990) FEBS Lett. 271, 33–36. 10. van Mansfield, A. D. M., Mosselman, S., Hoppener, J. W. M., Zandberg, J., van Teeffelen, H. A. A. M., Baas, P. D., Lips, C. J. M., and Jansz, H. S. (1990) Biochem. Biophys. Acta. 1087, 235–240. 11. Karlsson, O., Thor, S., Norberg, T., Ohlsson, H., and Edlund, T. (1990) Nature 344, 879–882. 12. German, M. S., Wang, J., Chadwick, R. B., and Rutter, W. J. (1992) Genes Dev. 6, 2165–2176. 13. Ohlsson, H., Karlsson, K., and Edlund, T. (1993) EMBO J. 12, 4251–4259. 14. Leonard, J., Peers, B., Johnson, T., Ferreri, K., Lee, S., and Montminy, M. R. (1993) Mol. Endocrinol. 7, 1275– 1283. 15. Miller, C. P., McGehee, R. E., and Habener, J. F. (1994) EMBO J. 13, 1145–1156. 16. Jonsson, J., Carlsson, L., Edlund, T., and Edlund, H. (1994) Nature 371, 606–609. 17. Guz, Y., Montminy, M. R., Stein, R., Leonard, J., Gamer, L. W., Wright, C. V., and Teitelman, G. (1995) Development 121, 11–18. 18. Sadowski, H. B., and Gilman, M. Z. (1993) Nature 362, 79–83. 19. Watada, H., Kajimoto, Y., Umayahara, Y., Matsuoka, T., Kaneto, Y., Fujitani, Y., Kamada, T., Kawamori, R., and Yamasaki, Y. (1996) Diabetes 45, 1478–1488. 20. Maxwell, I. H., Harrison, G. S., Wood, W. M., and Maxwel, F. (1989) Biotechniques 7, 276–280. 21. Kanatsuka, A., Makino, H., Yamaguchi, T., Ohsawa, H., Tokuyama, Y., Saitou, T., Yamaura, K., Miyazaki, J., and Yoshida, S. (1992) Diabetes 41, 1409–1414. 22. Umayahara, Y., Kawamori, R., Watada, H., Imano, E., Iwama, N., Morishima, T., Yamasaki, Y., Kajimoto, Y., and Kamada, T. (1994) J. Biol. Chem. 269, 16433–16442. 23. Bretherton-Watt, D., Gore, N., and Boam, D. S. W. (1996) Biochem. J. 313, 495–502. 24. German, M. S., Moss, L. G., Wang, J., and Rutter, W. J. (1992) Mol. Cell. Biol. 12, 1777–1788. 25. Wang, M., and Drucker, D. J. (1996) Mol. Endocrinol. 10, 243–251. 26. Watada, H., Kajimoto, Y., Miyagawa, J., Hanafusa, T., Hamaguchi, K., Matsuoka, T., Yamamoto, K., Matsuzawa, Y., Kawamori, R., and Yamasaki, Y. (1996) Diabetes, in press.
751
AID
BBRC 5834
/
6914$$$141
12-06-96 06:14:14
bbrca
229, 746–751 (1996)
1875
Involvement of the Homeodomain-Containing Transcription Factor PDX-1 in Islet Amyloid Polypeptide Gene Transcription Hirotaka Watada,1 Yoshitaka Kajimoto,2 Hideaki Kaneto, Taka-aki Matsuoka, Yoshio Fujitani, Jun-ichi Miyazaki,* and Yoshimitsu Yamasaki First Department of Medicine, and *Department of Nutrition and Physiological Chemistry, Osaka University School of Medicine, Suita 565, Japan Received November 7, 1996 The AT-rich cis-motif A elements of the insulin gene promoter contribute to directing the gene’s expression to pancreatic b-cells, bound by a homeodomain-containing transcription factor, PDX-1/IPF1/STF-1/ IDX-1. The islet amyloid polypeptide (IAPP; amylin) gene, which is also expressed in limited tissues such as pancreatic b- and d-cells, contained similar AT-rich sequences in its regulatory sequences. To understand the molecular basis of IAPP gene regulation, we evaluated the possible physiological significance of the motif in human IAPP gene regulation. All of the three typical A element-like sequences that matched the CT-box consensus (AT-1, -207/-202; AT-2, -154/-142; and AT-3, -88/-83) were shown to bind specifically to a nuclear factor in the b-cell-derived MIN6 cells, which was subsequently identified immunologically as the insulin gene transcription factor PDX-1. When the promoter activity was examined in MIN6 cells, the disruption of AT-1 or AT-3 but not of AT-2 caused a marked reduction in the IAPP gene promoter. Thus, despite the observation that all the three A element-like regions could bind to PDX-1, the AT-2 site may not be involved in mediating the PDX-1 action in vivo. These observations suggest the involvement of PDX-1 in human IAPP gene regulation, which seems to be mediated through at least two A elementlike cis-motifs in the gene promoter. q 1996 Academic Press
INTRODUCTION
Islet amyloid polypeptide (IAPP), also known as amylin, is a 37-amino acid peptide of the calcitonin gene family isolated from amyloid deposits of non-insulin-dependent diabetic pancreas and insulinoma (1). In pancreatic b-cells, IAPP is co-stored with insulin in the secretory granules and coreleased with insulin into the circulation in response to a variety of secretagogues. While its physiological role is yet to be clarified, IAPP has been suggested to be toxic to b-cells and to be involved in the development of non-insulin-dependent diabetes (1-3). Like insulin, the major site of IAPP synthesis is the islet b-cell (1). However, unlike insulin, detectable quantities of IAPP are also produced in other cells including d cells in pancreatic islets (4), stomach, duodenum, and jejunum (5). The similar but slightly wider tissue distribution pattern of IAPP suggests that some, if not all, of the gene regulating machinery of the insulin gene may be shared with the IAPP gene. The 5*-flanking region of the human IAPP gene contains multiple sequences similar to the insulin gene A elements, the cis-motif important for the b-cell specific expression of the gene (6-10). A group of proteins that belong to the homeodomain family of transcription factors (11-15) were shown to bind to the motif. Among them, PDX-1/IPF1/STF-1/IDX-1 (13-15), is 1 Recipient of a fellowship and grant from the Japan Society for the Promotion of Science for Japanese Junior Scientists. 2 Corresponding author: First Department of Medicine, Osaka University School of Medicine, 2-2 Yamadaoka, Suita City, Osaka Pref. 565, Japan. Fax: /81-6-879-3639. E-mail: [email protected]. Abbreviations: IAPP, islet amyloid polypeptide; bp, base pair(s); PCR, polymerase chain reaction.
746 0006-291X/96 $18.00 Copyright q 1996 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 5834
/
6914$$$141
12-06-96 06:14:14
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
the factor for which physiological significance has been demonstrated (16). PDX-1 appears before insulin during the ontogeny of the mouse pancreas, where it eventually becomes restricted to the b cells in the adult (13, 17). As a step toward elucidating transcriptional control of the IAPP gene, we investigated whether the A-element-like motif and the insulin gene transcription factor PDX-1 are involved in the regulation of IAPP gene expression. MATERIALS AND METHODS Preparation of nuclear extracts and gel mobility-shift analyses. Nuclear extracts were prepared, following the procedure described by Sadowski and Gilman (18). The binding reactions were performed as described previously (19) using 3 mg of each nuclear extract. The samples were fractionated on a 5% polyacrylamide gel pre-cooled to 4 7C in 0.5 1 TBE (45 mM Tris base, 45 mM boric acid, 1 mM EDTA). Reporter gene constructs. The reporter plasmid pA3hIAPPLuc/-410 was generated by polymerase chain reaction (PCR) using human genomic DNA as a template, with subsequent ligation to the HindIII pre-digested pA3Luc plasmid (20). The PCR procedures were performed. The plasmids pA3hIAPPLuc/-410/MAT-1, pA3hIAPPLuc/-410/MAT-2, and pA3hIAPPLuc/-410/MAT-3 were prepared by PCR-mediated site-directed mutagenesis. These reporter plasmids were designed to contain 32 bp of the exon 1 untranslated region in addition to the 5*-flanking sequences. Gene transfer studies. MIN6 cells (21) and HepG2 cells were cultured as described previously (21, 22). For gene transfection, cells were replated in six-well tissue culture plates 24 h before transfection. One microgram of each reporter plasmid and 1 mg of the pEFb-Gal plasmid were cotransfected into the cells by the lipofection method using lipofectAMINE reagent (Life Technologies, Tokyo, Japan). For evaluation of PDX-1 effects, 0.7 mg of PDX-1 (IPF1) expression plasmid, pcDNA3IPF1 or pcDNA3IPF1-A containing the PDX-1 coding region in the reverse direction (19) was cotransfected into HepG2 cells with 0.7 mg of each reporter plasmid and 0.7 mg of the pEFb-Gal plasmid. Transfected cells were incubated for 24 h, then fresh medium was added. After another 24-h incubation, luciferase and b-galactosidase assays were performed as we described previously (22), allowing normalization of the luciferase results with respect to transfection efficiency.
RESULTS
PDX-1 has high affinity for each A-element-like sequence. The regulatory region of the human IAPP gene (9, 10) contains three regions that match the consensus sequence for the A elements (CT-box consensus: C/TTAATG). They were located at -88/-83, -154/-142, and, in reverse orientation, -207/-202. For the sake of convenience, we call those three regions AT1, AT-2, and AT-3, respectively. To investigate the possible role of each as a cis-acting element in the IAPP gene regulation, we first performed gel mobility-shift analyses using nuclear extracts isolated from the pancreatic b-cell derived MIN6 cells. As shown in Fig. 1, the oligonucleotide probes containing AT-1, AT-2, or AT-3 site and their flanking sequences became bound to two nuclear factors in MIN6 cells. Among the two bands, however, the upper ones seemed non-specific because they were influenced even by the addition of mutatedtype competitors (Fig. 1). To identify the transcription factor involved in the lower band formation, we investigated whether the anti-PDX-1 (IPF1) antibody could recognize the DNAprotein complexes. In contrast to the pre-immune serum which displayed no significant effects on the complex, the addition of anti-PDX-1 (IPF1) antiserum totally eliminated the lower bands (arrows; Fig. 1). These results suggested that all the three A-element-like regions (AT1, AT-2, and AT-3) in the human IAPP gene are bound by PDX-1 in MIN6 cells. Site-directed mutagenesis at A-element-like motifs of the IAPP gene promoter. Further to examine whether the A-element-like sequences play a role in generating promoter activity of the IAPP gene, we performed reporter gene analyses using a wild-type and mutant IAPP gene promoters linked to a luciferase reporter (Fig. 2A). The reporter plasmids were transfected into MIN6 cells in which the features of gene regulation and biosynthesis of IAPP resembled those of islet cells from normal animal (21) and IAPP non-producing HepG2 cells. Figure 2B presents the activities of the wild-type and mutant IAPP gene promoters in those two cell lines as relative luciferase activities. The results (Fig. 2B) showed that, while the IAPP gene promoters generally displayed poor activities in HepG2 cells, they generated much higher activities in MIN6 cells than did the control promoter-less plasmid pA3Luc. In terms of the effects of 747
AID
BBRC 5834
/
6914$$$141
12-06-96 06:14:14
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Analyses of protein bindings to A element-like regions. Gel-mobility shift analyses were performed to evaluate the specific protein bindings to the A element-like motifs in the human IAPP gene. The 5*-end-[32P]labeled probe was incubated with nuclear extracts isolated from MIN6 cells. Where specified, 50-fold excess of unlabeled competitors, or preimmune serum or anti-PDX-1 (IPF1) antiserum, was also added to the binding reactions. Arrows and arrowheads indicate the disappearance and supershift of the DNA–protein complexes by the antiserum, respectively. Nucleotide sequences of sense strands of the double-stranded oligonucleotide probes and competitors were as follows: AT-1 probe, ATGACACACCATTAACTGCACAAG; mutated-type AT-1 probe, ATGACACACCAGGAACTGCACAAG; AT-2 probe, GATGAGTTAATGTAATAATGACCC; mutated-type AT-2 probe, GATGAGTTCCTGTAAGCATGACCC; AT-3 probe GGATGGAAATTAATGACAGAGGCT, mutated-type AT-3 probe, GGATGGAAATTCCTGACAGAGGCT. The exposure time was 12 h.
the mutations, the results varied depending upon the mutation site; destruction of AT-1 and AT-3 caused major loss of the promoter activity in MIN6 cells (56% and 83% decrease, respectively) but not that of the AT-2 site (12% increase; Fig. 2B). No significant reduction of the promoter activities was observed with any mutations when examined with HepG2 cells (Fig. 2B). These results thus suggested that AT-1 and AT-3 but not AT-2 function as cisacting motifs for the IAPP gene promoter. Activation of IAPP gene promoter by PDX-1. To examine whether PDX-1 can function as a positive regulator of the IAPP gene transcription, PDX-1 expression was induced in non-bcell derived HepG2 cells. As shown in Fig. 3, this expression caused 4.0-fold increase of the 410-bp IAPP gene promoter activity. All three mutations introduced at any of the A-elementlike sequences caused some deleterious effects on the promoter activity. However, in agreement with the observations for the promoter activities in MIN6 cells (Fig. 2B), the destruction of AT-1 or AT-3 conferred more damage on the induction ratio by PDX-1 (2.5-fold, 1.5-fold, respectively) as compared with the AT-2 mutation (3.1-fold; Fig. 3). These results suggested that PDX-1 activates the IAPP gene promoter and that the activation is mediated mainly by AT-1 and AT-3 sites. DISCUSSION
In this study, we showed that PDX-1 can activate the human IAPP gene promoter and this is mediated by at least two of the A-element-like regions in the 5*-flanking sequence. As we recently characterized the human glucokinase gene b-cell-type promoter and identified PDX1 as an essential transcription factor for it (19), PDX-1 has now been shown to be an important 748
AID
BBRC 5834
/
6914$$$141
12-06-96 06:14:14
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. Effects of mutations of A element-like regions on promoter activities. (A) Structure of the reporter gene constructs used in the reporter gene analyses. The pA3IAPPLuc/-410 plasmid, shown as ‘‘Wild Type,’’ contained the 410-bp 5*-flanking sequence and 32-bp untranslated exon 1 sequences of the human IAPP gene. The three mutant reporter plasmids pA3hIAPPLuc/-410/MAT-1 (shown as ‘‘MAT-1’’), pA3hIAPPLuc/-410/MAT-2 (‘‘MAT-2’’), and pA3IAPPLuc/-420/MAT-3 (‘‘MAT-3’’) were prepared by introducing mutations to the AT-1, AT-2, and AT-3 regions, respectively. Substituted nucleotides are underlined. Sequences of the rat insulin I gene FLAT region and of the AT2 region are aligned to facilitate comparison. (B) Effects of the mutation introduced to AT-1 (shown as ‘‘MAT-1’’), AT-2 (‘‘MAT-2’’), and AT-3 (‘‘MAT-3’’) on the promoter activities in MIN6 cells (open bars) and HepG2 cells (filled bars). Each reporter plasmid was individually transfected into MIN6 cells or HepG2 cells, and luciferase activity was measured 48 h later. The results shown here are expressed as relative luciferase activities with that of the mutationless promoter (‘‘Wild Type’’) in MIN6 cells arbitrarily set at 1. They were normalized with respect to those of pA3RSVLuc in order to allow comparison of the results obtained with MIN6 cells and HepG2 cells. All data are presented as means { SE of at least four individual transfection experiments.
transcription factor for at least three b-cell-specific genes. This broad range of target genes for which PDX-1 functions as a transcription factor is consistent with the phenotype of the PDX-1 (IPF1)-knock out mouse demonstrating that PDX-1 is required for pancreas development (16). Recently, Bretherton-Watt et al. reported that PDX-1 bind the same motifs that we pointed out in this study, although, unlike us, they did not examine the function of the motifs (23). An interesting point that became evident in our present study was a discrepancy between the gel-mobility shift and reporter gene results obtained with respect to the AT-2 site. Among the three A-element-like motifs, only AT-1 and AT-3 appeared to play a physiological role in the IAPP gene transcription (Fig. 2B). In contrast, the AT-2 site, to which PDX-1 can also bind 749
AID
BBRC 5834
/
6914$$$141
12-06-96 06:14:14
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Effects of PDX-1 on IAPP promoter activities. Expression of PDX-1 in HepG2 cells was induced by transient transfection of an expression plasmid (pcDNA3IPF1 (24)), and the effects on the IAPP gene promoter activities were examined. The transfected plasmids (‘‘Wild-type’’, ‘‘MAT-1’’, ‘‘MAT-2’’, and ‘‘MAT-3’’) are described in the legend to Fig. 2. Each reporter plasmid was individually cotransfected into HepG2 cells with pcDNA3IPF1 (24) or the control plasmid pcDNA3IPF1-A. The results are expressed as n-fold induction over the promoter activity of each reporter plasmid in the control pcDNA3IPF1-A-transfected HepG2 cells. All data are presented as means { SE of four individual transfection experiments.
at least in vitro (Fig. 1), seemed not to be important, or to be less important, as a cis-acting motif in MIN6 cells (Fig. 2B & 3). Although this tends to be speculative, we may need to refer to the structural characteristics of this AT-2 region in order to discuss the reason for this discrepancy. When the sequences are compared with other b-cell specific promoters, we find that AT-2 is structurally similar to the A3/A4 (FLAT) region of the rat insulin I gene (Fig. 2A). According to reports by German et al. (24), the FLAT region is composed of two negatively interacting elements. Also, they pointed out that the phenotypes of the FLAT region mutation vary depending upon which nucleotides of the region are mutated. Although PDX1 was also shown to bind to the FLAT region as well as to the AT-2 element (6), these observations suggest that some factors other than PDX-1 may also bind to these regions and thus make it difficult to interpret their function. The in vitro gel-mobility shift assay may have failed to detect the band corresponding to this, for example, because of low affinity. Recently, Wang et al. reported that transcription of the rat IAPP gene is activated by the LIM domain homeobox protein isl-1 (25). They also performed gel-mobility shift assays using the A-element-like regions in the rat IAPP gene and thereby detected multiple bands that were specific to the CT-box (A element) consensus motif. They showed one of those contained isl1. In contrast to their observations, we as well as Bretherton-Watt et al. used the binding probe reproducing the human IAPP gene A element-like region and detected only PDX-1 as the binding factor. Although this did not totally exclude the possibility of isl-1 involvement, it suggested that PDX-1 is the dominant A element-like region-binding factor, at least in MIN6 cells (this study) and bTC3 cells (23). To clarify the physiological significance of PDX-1, we recently established a stable transfectant of aTC1 clone 6 cells that expresses PDX-1. Despite its failure to drive insulin, glucokinase or GLUT2 gene expression, the exogenously expressed PDX-1 could induce IAPP gene expression in aTC1 clone 6 cells (26). Also, the tissues expressing IAPP correspond very well to those expressing PDX-1; pancreatic b and d cells and duodenum (7, 8, 17-19). These observations, along with our present data, provide support for the physiological significance of PDX1 as a primary determinant of IAPP gene expression. ACKNOWLEDGMENTS We thank Dr. W. M. Wood of University of Colorado Health Science Center and Dr. D. R. Helinski of University of California San Diego for the pA3Luc and pA3RSVLuc plasmids. We also thank Ms. N. Fujita for the excellent technical assistance. 750
AID
BBRC 5834
/
6914$$$141
12-06-96 06:14:14
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
This work was supported in part by grants from Japan Diabetes Foundation (to Y.K.), and Kyowa Hakko Kogyo Co. Ltd. (to Y.K. and Y.Y.).
REFERENCES 1. Cooper, G. J. S. (1994) Endocr. Rev. 15, 163–201. 2. Lorenzo, A., Razzaboni, B., Weir, G. C., and Yankner, B. A. (1994) Nature 368, 756–760. 3. Verchere, C. B., D’ Alessio, D. A., Palmiter, R. D., Weir, G. C., Bonner-Weir, S., Baskin, D. G., and Kahn, A. E. (1996) Proc. Natl. Acad. Sci. U.S.A. 93, 3492–3496. 4. De Vroede, M., Foriers, A., Van de Winkel, M., Madsen, O., and Pipeleers, D. (1992) Biochem. Biophys. Res. Commun. 182, 886–893. 5. Nakazato, M., Asai, J., Kangawa, K., Matsukura, S., and Matsuo, H. (1989) Biochem. Biophys. Res. Commun. 164, 394–399. 6. German, M., Ashcroft, S., Docherty, K., Edlund, H., Edlund, T., Goodison, T., Imura, H., Kennedy, G., Madsen, O., Melloul, D., Moss, L., Olson, K., Permutt, M. A., Philippe, J., Robertson, J. P., Rutter, W. J., Serup, P., Stein, R., Steiner, D., Tsai, M.-J., and Walker, M. D. (1995) Diabetes 44, 1002–1004. 7. Ohlsson, H., Karlsson, O., and Edlund, T. (1988) Proc. Natl. Acad. Sci. U.S.A. 85, 4228–4231. 8. Ohlsson, H., Thor, S., and Edlund, T. (1991) Mol. Endocrinol. 5, 897–904. 9. Mosselman, S., Hoppener, J. W. M., de Wit, L., Soeller, W., Lips, C. J. M., and Jansz, H. S. (1990) FEBS Lett. 271, 33–36. 10. van Mansfield, A. D. M., Mosselman, S., Hoppener, J. W. M., Zandberg, J., van Teeffelen, H. A. A. M., Baas, P. D., Lips, C. J. M., and Jansz, H. S. (1990) Biochem. Biophys. Acta. 1087, 235–240. 11. Karlsson, O., Thor, S., Norberg, T., Ohlsson, H., and Edlund, T. (1990) Nature 344, 879–882. 12. German, M. S., Wang, J., Chadwick, R. B., and Rutter, W. J. (1992) Genes Dev. 6, 2165–2176. 13. Ohlsson, H., Karlsson, K., and Edlund, T. (1993) EMBO J. 12, 4251–4259. 14. Leonard, J., Peers, B., Johnson, T., Ferreri, K., Lee, S., and Montminy, M. R. (1993) Mol. Endocrinol. 7, 1275– 1283. 15. Miller, C. P., McGehee, R. E., and Habener, J. F. (1994) EMBO J. 13, 1145–1156. 16. Jonsson, J., Carlsson, L., Edlund, T., and Edlund, H. (1994) Nature 371, 606–609. 17. Guz, Y., Montminy, M. R., Stein, R., Leonard, J., Gamer, L. W., Wright, C. V., and Teitelman, G. (1995) Development 121, 11–18. 18. Sadowski, H. B., and Gilman, M. Z. (1993) Nature 362, 79–83. 19. Watada, H., Kajimoto, Y., Umayahara, Y., Matsuoka, T., Kaneto, Y., Fujitani, Y., Kamada, T., Kawamori, R., and Yamasaki, Y. (1996) Diabetes 45, 1478–1488. 20. Maxwell, I. H., Harrison, G. S., Wood, W. M., and Maxwel, F. (1989) Biotechniques 7, 276–280. 21. Kanatsuka, A., Makino, H., Yamaguchi, T., Ohsawa, H., Tokuyama, Y., Saitou, T., Yamaura, K., Miyazaki, J., and Yoshida, S. (1992) Diabetes 41, 1409–1414. 22. Umayahara, Y., Kawamori, R., Watada, H., Imano, E., Iwama, N., Morishima, T., Yamasaki, Y., Kajimoto, Y., and Kamada, T. (1994) J. Biol. Chem. 269, 16433–16442. 23. Bretherton-Watt, D., Gore, N., and Boam, D. S. W. (1996) Biochem. J. 313, 495–502. 24. German, M. S., Moss, L. G., Wang, J., and Rutter, W. J. (1992) Mol. Cell. Biol. 12, 1777–1788. 25. Wang, M., and Drucker, D. J. (1996) Mol. Endocrinol. 10, 243–251. 26. Watada, H., Kajimoto, Y., Miyagawa, J., Hanafusa, T., Hamaguchi, K., Matsuoka, T., Yamamoto, K., Matsuzawa, Y., Kawamori, R., and Yamasaki, Y. (1996) Diabetes, in press.
751
AID
BBRC 5834
/
6914$$$141
12-06-96 06:14:14
bbrca