Relative contribution of PDX-1, MafA and E47/beta2 to the regulation of the human insulin promoter

Abstract

The insulin promoter binds a number of tissue-specific and ubiquitous transcription factors. Of these, the homoeodomain protein PDX-1 (pancreatic duodenal homeobox factor-1), the basic leucine zipper protein MafA and the basic helix-loop-helix heterodimer E47/BETA2 (beta-cell E box transactivator 2; referred to here as beta2) bind to important regulatory sites. Previous studies have shown that PDX-1 can interact synergistically with E47 and beta2 to activate the rat insulin 1 promoter. The aim of the present study was to determine the relative contribution of PDX-1, MafA and E47/beta2 in regulating the human insulin promoter, and whether these factors could interact synergistically in the context of the human promoter. Mutagenesis of the PDX-1, MafA and E47/beta2 binding sites reduced promoter activity by 60, 74 and 94% respectively, in INS-1 beta-cells. In the islet glucagonoma cell line alphaTC1.6, overexpression of PDX-1 and MafA separately increased promoter activity approx. 2.5-3-fold, and in combination approx. 6-fold, indicating that their overall effect was additive. Overexpression of E47 and beta2 had no effect. In HeLa cells, PDX-1 stimulated the basal promoter by approx. 40-fold, whereas MafA, E47 and beta2 each increased activity by less than 2-fold. There was no indication of any synergistic effects on the human insulin promoter. On the other hand, the rat insulin 1 promoter and a mutated version of the human insulin promoter, in which the relevant regulatory elements were separated by the same distances as in the rat insulin 1 promoter, did exhibit synergy. PDX-1 was shown further to activate the endogenous insulin 1 gene in alphaTC1.6 cells, whereas MafA activated the insulin 2 gene. In combination, PDX-1 and MafA activated both insulin genes. Chromatin immunoprecipitation assays confirmed that PDX-1 increased the association of acetylated histones H3 and H4 with the insulin 1 gene and MafA increased the association of acetylated histone H3 with the insulin 2 gene.

Figure 1. Structure of the human insulin promoter constructs

Schematic diagram showing the structure of the human insulin promoter constructs. The positions of the A1, A3, C1 and E1 sites are indicated along with the CREs (CRE3 and CRE4) that lie downstream of the transcription start site (arrow). The scale represents base pairs relative to the start site.

Figure 2. Insulin promoter activity in INS-1 cells

INS-1 cells were transfected with 250 ng of the indicated insulin promoter construct and 2.5 ng of phRL-TK with pcDNA3.1 added to a final total of 1 μg. After 48 h, the cells were harvested and assayed for firefly and Renilla luciferase activity. The data are expressed as percentage luciferase activity (firefly/Renilla) relative to cells transfected with unmutated phINS171LUC construct. Values are means±S.D. of between 3–9 separate experiments.

Figure 3. Effect of PDX-1, PDX-1/VP-16, MafA, E47 and β2, individually and in combination, on phINS171LUC promoter activity in αTC1.6 cells

αTC1.6 cells were transfected with 250 ng of phINS171LUC, 2.5 ng of phRL-TK and 250 ng of expression plasmids for the indicated transcription factors with pcDNA3.1 added to a final total of 1 μg. After 48 h, the cells were harvested and assayed for firefly and Renilla luciferase activity. (A) The effect of transcription factors on their own was assayed. (B) Transcription factors were assayed in pairs. (C) Multiple combinations of transcription factors were assayed. (D) PDX-1/VP-16 was used in place of PDX-1 in a series of experiments in combination with the indicated transcription factors. The data are expressed as luciferase activity (firefly/Renilla) relative to that measured in the absence of the indicated transcription factors. The data shown (means±S.D.) are representative of up to seven separate experiments.

Figure 4. Western blot showing the level of expression of exogenous gene products in transfected cells

Nuclear extracts were prepared from INS-1 or αTC1.6 cells and Western blots were performed using antibodies against PDX-1, MafA, E47 and β2 as indicated.

Figure 5. Effect of PDX-1, MafA and β2, individually and in combination, on rat insulin 1 and human insulin promoter activity in HeLa cells

HeLa cells were transfected with 250 ng of reporter plasmid indicated below, 0.25 ng of phRL-TK and 250 ng of expression plasmids for the indicated transcription factors with pcDNA3.1 added to a final total of 1 μg. The plasmids used were (A) pFOX410LUC containing the rat insulin 1 promoter, (B) phINS171LUC containing the human insulin promoter, (C) phINS171C+2EALUC, (D) phINS171CE+2ALUC, and (E) phINS171C+2E+2ALUC. After 48 h, the cells were harvested and assayed for firefly and Renilla luciferase activity. The data are expressed as luciferase activity (firefly/Renilla) relative to that measured in the absence of the indicated transcription factors for each reporter plasmid used. The data shown (means±S.D.) are representative of three separate experiments performed in triplicate.

Figure 6. Comparison of the rat insulin 1 and human insulin promoters

(A) Alignment of the DNA sequences of the relevant region of the rat insulin 1 promoter (−133 to −74) and the human insulin promoter (−130 to −75). The MafA, β2 and PDX-1 consensus sequences are underlined and bases inserted into the mutated forms of phINS171LUC are shown in bold. (B) Schematic diagram showing the effect of spacing between the regulatory elements on potential transcription factor alignments. The insulin promoter is depicted looking down the DNA helical axis and is represented as a circle. The transcription factors are denoted as differently shaded ovals and are displayed binding to the centres of their respective regulatory elements at the bottom of (A). Note that the human insulin promoter contains an additional PDX-1 binding sequence (GG2) starting 9 bp upstream of the region shown in (A). The potential differences in spatial orientation of MafA, β2 and PDX-1 relative to each other within the various promoter constructs are portrayed by showing the transcription factors coming in contact with the DNA at the centre of their respective consensus binding sequences. Each schematic diagram illustrates the effects of regulatory element spacing in the promoter to the immediate left in (A).

Figure 7. Effect of PDX-1 and MafA on endogenous gene expression in αTC1.6 cells

αTC1.6 cells were transfected with 250 ng of expression plasmids for PDX-1, MafA and PDX-1/VP16 as indicated, with pcDNA3.1 added to a final total of 2 μg. After 48 h, the cells were harvested and assayed by RT-PCR for mRNA encoding insulin 1 (INS1), insulin 2 (INS2), IAPP, somatostatin (SOM), glucagon (GLU) and GAPDH, as indicated. The data are representative of three separate experiments.

Figure 8. ChiP of the insulin 1 and 2 genes in αTC1.6 cells transfected with PDX-1 or MafA

αTC1.6 cells transfected with expression vectors for transcription factors (TF) PDX-1, MafA, PDX-1/VP16 or empty vector (control) were subjected to ChiP assays using anti-acetylated H3 (AcH3) or anti-acetylated H4 (AcH4) antibodies and PCR primers specific for insulin 1 (INS1), insulin 2 (INS2) and GAPDH. (A) PCR results representative of four separate experiments. (B) Quantification of the PCR products with the amounts of DNA in each band calculated relative to the pre-immune (P.I.) signal. The data represent the means±S.D. of the four separate experiments. Insulin 1, open bars; insulin 2, closed bars.