Cooperative interactions among intestinal GATA factors in activating the rat liver fatty acid binding protein gene.

GATA-4, GATA-5, and GATA-6 are endodermal zinc-finger transcription factors that activate numerous enterocytic genes. GATA-4 and GATA-6 but not GATA-5 are present in adult murine small intestinal enterocytes, and we now report the simultaneous presence of all three GATA factors in murine small intestinal enterocytes before weaning age. An immunohistochemical survey detected enterocytic GATA-4 and GATA-6 at birth and 1 wk of age and GATA-5 at 1 wk but not birth. Interactions among GATA factors were explored utilizing a transgene constructed from the proximal promoter of the rat liver fatty acid binding protein gene (Fabp1). GATA-4 and GATA-5 but not GATA-6 activate the Fabp1 transgene through a cognate binding site at -128. A dose-response assay revealed a maximum in transgene activation by both factors, where additional factor did not further increase transgene activity. However, at saturated levels of GATA-4, additional transgene activation was achieved by adding GATA-5 expression construct, and vice versa. Similar cooperativity occurred with GATA-5 and GATA-6. Identical interactions were observed with a target transgene consisting of a single GATA site upstream of a minimal promoter. Furthermore, GATA-4 and GATA-5 or GATA-5 and GATA-6 bound to each other in solution. These results are consistent with tethering of one GATA factor to the Fabp1 promoter through interaction with a second GATA factor to produce increased target gene activation. Cooperative target gene activation was specific to an intestinal cell line and may represent a mechanism by which genes are activated in the small intestinal epithelium during the period before weaning.

Mechanisms of mutual functional interactions between HNF-4alpha and HNF-1alpha revealed by mutations that cause maturity onset diabetes of the young.

Hepatic nuclear factor (HNF)-4alpha and HNF-1alpha are key endodermal transcriptional regulators that physically and functionally interact. HNF-4alpha and HNF-1alpha cooperatively activate genes with binding sites for both factors, whereas suppressive interactions occur at regulatory sequences with a binding site for only one factor. The liver fatty acid binding protein gene (Fabp1) has binding sites for both factors, and chromatin precipitation assays were utilized to demonstrate that HNF-4alpha increased HNF-1alpha Fabp1 promoter occupancy during cooperative transcriptional activation. The HNF4 P2 promoter contains a HNF-1 but not HNF-4 binding site, and HNF-4alpha suppressed HNF-1alpha HNF4 P2 activation and decreased promoter HNF-1alpha occupancy. The apolipoprotein C III (APOC3) promoter contains a HNF-4 but not HNF-1 binding site, and HNF-1alpha suppressed HNF-4alpha APOC3 activation and decreased HNF-4alpha promoter occupancy. Maturity onset diabetes of the young (MODY) as well as defects in hepatic lipid metabolism result from mutations in either HNF-4alpha or HNF-1alpha. We found that MODY missense mutant R127W HNF-4alpha retained wild-type individual Fabp1 activation and bound to HNF-1alpha better than wild-type HNF-4alpha, yet did not cooperate with HNF-1alpha or increase HNF-1alpha Fabp1 promoter occupancy. The R127W mutant was also defective in both suppressing HNF-1alpha activation of HNF4 P2 and decreasing HNF-1alpha promoter occupancy. The HNF-1alpha R131Q MODY mutant also retained wild-type Fabp1 activation and bound to HNF-4alpha as well as the wild type but was defective in both suppressing HNF-4alpha APOC3 activation and decreasing HNF-4alpha promoter occupancy. These results suggest HNF-1alpha-HNF-4alpha functional interactions are accomplished by regulating factor promoter occupancy and that defective factor-factor interactions may contribute to the MODY phenotype.

C/EBP and Cdx family factors regulate liver fatty acid binding protein transgene expression in the small intestinal epithelium.

A transgene constructed from the rat liver fatty acid binding protein gene (Fabp1) promoter is active in all murine small intestinal crypt and villus epithelial cells. Coincident Cdx and C/EBP transcription factor binding sites were identified spanning Fabp1 nucleotides -90 to -78. CDX-1, CDX-2, C/EBPalpha, and C/EBPbeta activated the Fabp1 transgene in CaCo-2 cells, and mutagenizing the -78 site prevented activation by these factors. CDX but not C/EBP factors bound to the site in vitro, although C/EBP factors competed with CDX factors for transgene activation. The -78 site adjoins an HNF-1 site, and CDX and C/EBP family factors cooperated with HNF-1alpha but not HNF-1beta to activate the transgene. Furthermore, CDX-1, CDX-2, C/EBPalpha, and C/EBPbeta bound to HNF-1alpha and HNF-1beta. The transgene with a mutagenized -78 site was silenced in vivo specifically in small intestinal crypt epithelial cells but remained active in villus cells. These results demonstrate functional interactions between HNF-1, C/EBP, and CDX family factors and suggest that these interactions may contribute to differential transcriptional regulation in the small intestinal crypt and villus compartments.

GATA-4, GATA-5, and GATA-6 activate the rat liver fatty acid binding protein gene in concert with HNF-1alpha.

Transcriptional regulation by GATA-4, GATA-5, and GATA-6 in intestine and liver was explored using a transgene constructed from the proximal promoter of the rat liver fatty acid binding protein gene (Fabpl). An immunohistochemical survey detected GATA-4 and GATA-6 in enterocytes, GATA-6 in hepatocytes, and GATA-5 in neither cell type in adult animals. In cell transfection assays, GATA-4 or GATA-5 but not GATA-6 activated the Fabpl transgene solely through the most proximal of three GATA binding sites in the Fabpl promoter. However, all three factors activated transgenes constructed from each Fabpl site upstream of a minimal viral promoter. GATA factors interact with hepatic nuclear factor (HNF)-1alpha, and the proximal Fabpl GATA site adjoins an HNF-1 site. GATA-4, GATA-5, or GATA-6 bounded to HNF-1alpha in solution, and all cooperated with HNF-1alpha to activate the Fabpl transgene. Mutagenizing all Fabpl GATA sites abrogated transgene activation by GATA factors, but GATA-4 activated the mutagenized transgene in the presence of HNF-1alpha. These in vitro results suggested GATA/HNF-1alpha interactions function in Fabpl regulation, and in vivo relevance was determined with subsequent experiments. In mice, the Fabpl transgene was active in enterocytes and hepatocytes, a transgene with mutagenized HNF-1 site was silent, and a transgene with mutagenized GATA sites had identical expression as the native transgene. Mice mosaic for biallelic Gata4 inactivation lost intestinal but not hepatic Fabpl expression in Gata4-deficient cells but not wild-type cells. These results demonstrate GATA-4 is critical for intestinal gene expression in vivo and suggest a specific GATA-4/HNF-1alpha physical and functional interaction in Fabpl activation.

HNF-1alpha and endodermal transcription factors cooperatively activate Fabpl: MODY3 mutations abrogate cooperativity.

Hepatocyte nuclear factor (HNF)-1alpha plays a central role in intestinal and hepatic gene regulation and is required for hepatic expression of the liver fatty acid binding protein gene (Fabpl). An Fabpl transgene was directly activated through cognate sites by HNF-1alpha and HNF-1beta, as well as five other endodermal factors: CDX-1, C/EBPbeta, GATA-4, FoxA2, and HNF-4alpha. HNF-1alpha activated the Fabpl transgene by as much as 60-fold greater in the presence of the other five endodermal factors than in their absence, accounting for up to one-half the total transgene activation by the group of six factors. This degree of synergistic interaction suggests that multifactor cooperativity is a critical determinant of endodermal gene activation by HNF-1alpha. Mutations in HNF-1alpha that result in maturity onset diabetes of the young (MODY3) provide evidence for the in vivo significance of these synergistic interactions. An R131Q HNF-1alpha MODY3 mutant exhibits complete loss of synergistic activation in concert with the other endodermal transcription factors despite wild-type transactivation ability in their absence. Furthermore, whereas wild-type HNF-1alpha exhibited pairwise cooperative synergy with each of the other five factors, the R131Q mutant could synergize only with GATA-4 and C/EBPbeta. Selective loss of synergy with other endodermal transcription factors accompanied by retention of native transactivation ability in an HNF-1alpha MODY mutant suggests in vivo significance for cooperative synergy.