HNF4A and HNF1A exhibit tissue specific target gene regulation in pancreatic beta cells and hepatocytes

Abstract

HNF4A and HNF1A encode transcription factors that are important for the development and function of the pancreas and liver. Mutations in both genes have been directly linked to Maturity Onset Diabetes of the Young (MODY) and type 2 diabetes (T2D) risk. To better define the pleiotropic gene regulatory roles of HNF4A and HNF1A, we generated a comprehensive genome-wide map of their binding targets in pancreatic and hepatic cells using ChIP-Seq. HNF4A was found to bind and regulate known (ACY3, HAAO, HNF1A, MAP3K11) and previously unidentified (ABCD3, CDKN2AIP, USH1C, VIL1) loci in a tissue-dependent manner. Functional follow-up highlighted a potential role for HAAO and USH1C as regulators of beta cell function. Unlike the loss-of-function HNF4A/MODY1 variant I271fs, the T2D-associated HNF4A variant (rs1800961) was found to activate AKAP1, GAD2 and HOPX gene expression, potentially due to changes in DNA-binding affinity. We also found HNF1A to bind to and regulate GPR39 expression in beta cells. Overall, our studies provide a rich resource for uncovering downstream molecular targets of HNF4A and HNF1A that may contribute to beta cell or hepatic cell (dys)function, and set up a framework for gene discovery and functional validation.

Fig. 1. Comprehensive ChIP-Seq identifies HNF4A and HNF1A targets in pancreatic and hepatic cells.

a Schematic of ChIP-Seq pipeline using multiple human cell models to identify HNF4A- and HNF1A-bound targets in pancreatic and hepatic cells. Scale bar indicates 100 µm. b Immunofluorescence microscopy images showing HNF4A and HNF1A nuclear localization in beta cells using the antibodies for ChIP of HNF4A (R&D Systems, H1415) and HNF1A (ab96777). Scale bar indicates 50 µm. c HNF4A ChIP-qPCR fold enrichment at the HNF1A promoter across different cell types (n = 4 for D20 EP; n = 3 for all other tissues). d HNF1A ChIP-qPCR fold enrichment at the HNF4A P2 promoter across different cell types (n = 2 for islets; n = 3 for all other tissues). e Representative HNF4A or HNF1A motifs identified in the ChIP-Seq data and the total number of ChIP-Seq peaks identified based on q value cut-off of 0.1. f Peak count frequency profile of HNF4A (left) or HNF1A (right) ChIP-Seq peaks that map within 1 kb of the transcription start site (TSS) across multiple cell types and the corresponding genomic feature distributions mapped to the peaks. PP Pancreatic progenitors, EP Endocrine progenitors, βLC Beta-like cells, Hep Hepatoblasts. Data are presented as mean ± SEM. Each data point represents one independent experiment. * indicates p < 0.05, ** indicates p < 0.01, relative to IgG control, using one-way ANOVA with Fisher’s LSD post-hoc test. Source data and exact P values are provided in the Source Data file.

Fig. 2. Identification of HNF4A-bound targets in beta cells and hepatic cells reveal enrichment for common and distinct pathways across tissue types and identified beta cell target genes.

a Top Gene Ontology (GO) Biological Processes (BP) commonly identified in both EndoC-βH1 and human islet HNF4A ChIP-Seq. b Top GO BP identified in D35 βLC HNF4A ChIP-Seq, in side-by-side comparisons with EndoC-βH1 and human islet samples. c Top GO BP in EndoC-βH1 in comparison with HepG2 HNF4A ChIP-Seq. d Venn diagram showing overlaps in HNF4A-bound target genes in EndoC-βH1 cells, human islets, and D35 βLCs based on ChIP-Seq peaks mapping within 10 kb of the transcription start site (TSS) (number of total peaks with no filtering shown in brackets). The table provides a list of the common beta-cell target gene loci identified, of which some were also replicated in D20 EPs (in green). Analysis and visualization of GO data is based on the ChIPseeker R package (see Methods).

Fig. 3. Prioritization and functional validation of HNF4A-bound beta cell targets.

a IGV tracks showing HNF4A ChIP-Seq peaks in selected loci in D35 βLCs, EndoC-βH1 cells, and human islets. The scale used to visualize peaks in IGV is indicated on the right side of each track. The chromosomal location near the peak region is indicated. b Gene expression analysis of selected target genes in EndoC-βH1 cells with HNF4A siRNA-mediated knockdown (n = 5). * indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001 relative to si-NT control, based on unpaired two-tailed Students’ t test. c Gene expression analysis of selected target genes in EndoC-βH1 cells overexpressing GFP (empty vector), HNF4A WT, T139I variant, or MODY1 I271fs variant constructs (n = 4). * indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001 relative to GFP control; $ indicates p < 0.05 relative to WT, based on one-way ANOVA with Tukey’s post-hoc test. d Transactivation activities at selected target promoters in EndoC-βH1 cells using luciferase reporter assays (n = 3 for ACY3/HAAO/USH1C; n = 4 for VIL1). * indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001 relative to Empty si-HNF4A control; $ indicates p < 0.01 relative to WT, based on one-way ANOVA with Tukey’s post-hoc test. e Glucose-stimulated insulin secretion (GSIS) in EndoC-βH1 cells with siRNA-mediated KD of selected target genes (n = 4). * indicates p < 0.05 relative to si-NT control, based on two-way ANOVA with Dunnett’s multiple comparisons test. f Insulin secreted as a percentage of total insulin in EndoC-βH1 cells with siRNA-mediated KD of selected target genes, at 16.7 mM glucose stimulation (n = 4). P value indicated is based on paired two-tailed Students’ t test. Data are presented as mean ± SD. Each data point represents one independent experiment. Source data and exact P values are provided in the Source Data file.

Fig. 4. HNF4A downstream targets signal changes across hepatic cells and pinpoint common targets in both beta cells and hepatic cells.

a Topmost common and distinct Gene Ontology (GO) biological processes (BP) of HNF4A ChIP-Seq targets in D8 hepatoblasts and HepG2 cells. Analysis and visualization of GO data is based on the ChIPseeker R package (see Methods). b Venn diagram showing overlaps in HNF4A-bound beta cell and hepatic cell target genes based on ChIP-Seq peaks within 10 kb of the transcription start site (TSS). Table provides a list of gene loci identified in both cell types. c IGV tracks showing ChIP-Seq peaks that map to the nearest genes in selected loci in hepatic cells. The scale used to visualize peaks in IGV is indicated on the right side of each track. The chromosomal location near the peak region is indicated. d Luciferase reporter analysis of CDKN2AIP, HAAO and MAP3K11 promoter activities in HepG2 cells (n = 3 for HAAO/MAP3K11; n = 4 for CDKN2AIP). Data are presented as mean ± SD. Each data point represents one independent experiment. *** indicates p < 0.001, ** indicates p < 0.01, relative to Empty si-HNF4A control in the presence of the promoter. $ indicates p < 0.05 relative to WT, based on one-way ANOVA with Tukey’s post-hoc test. Source data and exact P values are provided in the Source Data file.

Fig. 5. Identification of HNF1A-bound targets in pancreatic endocrine cells highlight several target genes regulated by HNF1A.

a Topmost gene ontology (GO) biological processes (BP) found in HNF1A ChIP-Seq targets in human islets. Analysis and visualization of GO data is based on the ChIPseeker R package (see Methods). b Venn diagram showing overlaps in HNF1A-bound target genes in D20 EPs and human islets based on ChIP-Seq peaks within 10 kb of the transcription start site (TSS) (number of total peaks with no filtering shown in brackets). Table provides a consensus list of the beta cell target genes identified. c IGV tracks showing HNF1A ChIP-Seq peaks that map to the nearest genes in selected loci in D20 EPs and human islets. The scale used to visualize peaks in IGV is indicated on the right side of each track. The chromosomal location near the peak region is indicated. d Gene expression analysis of selected target genes in EndoC-βH1 cells with HNF1A siRNA-mediated knockdown (n = 3). * indicates p < 0.05, ** indicates p < 0.01 relative to si-NT (non-targeting control) using paired two-tailed Students’ t test. e Transactivation activity of GPR39 promoter in EndoC-βH1 cells (n = 3). * indicates p < 0.05 relative to Empty si-HNF1A control in the presence of the GPR39 promoter; $ indicates p < 0.01 relative to WT, based on one-way ANOVA with Tukey’s post-hoc test. Data are presented as mean ± SD. Each data point represents one independent experiment. Source data and exact P values are provided in the Source Data file.

Fig. 6. HNF4A and HNF1A commonly bind to several beta cell targets and govern overlapping pathways in both beta cells and hepatic cells.

a Topmost common and distinct gene ontology (GO) biological processes (BP) from HNF4A- and HNF1A-bound target genes in human islets. b Venn diagram showing overlap in HNF4A- and HNF1A-bound target genes in human islets based on ChIP-Seq peaks within 10 kb of the transcription start site (TSS) (number of total peaks with no filtering shown in brackets). Table provides a consensus list of the commonly-bound target genes. c Venn diagram showing overlaps in HNF4A- and HNF1A-bound target genes in HepG2 cells based on ChIP-Seq peaks within 10 kb of the transcription start site (TSS) (number of total peaks with no filtering shown in brackets). HNF4A-bound targets in HepG2 cells are from the consensus list of genes. d Topmost common and distinct gene ontology (GO) biological processes (BP) from HNF4A- and HNF1A-bound target genes in HepG2 cells. Analysis and visualization of GO data is based on the ChIPseeker R package (see Methods).

Fig. 7. Investigation of the effects of HNF4A T2D risk variant rs1800961 on gene regulation using FLAG ChIP-Seq.

a Western blot analysis of cell lysates from EndoC-βH1 cells stably overexpressing FLAG-tagged HNF4A WT and T2D variant constructs. b Immunofluorescence microscopy images showing nuclear localization of FLAG-tagged HNF4A WT and T2D variants overexpressed in EndoC-βH1 cells, with constitutive GFP signal from the vector in green, FLAG-tagged protein in red and DAPI in blue. Enlarged images from the insets are shown in the last panel. Scale bar indicates 100 µm. c HNF4A FLAG ChIP-qPCR fold enrichment at the HNF1A promoter in the stable EndoC-βH1 cell lines (n = 3). * indicates p < 0.05 relative to GFP control based on one-way ANOVA with Dunnett’s post-hoc test. d Representative HNF4A motif identified in the ChIP-Seq samples and the total number of ChIP-Seq peaks identified in each sample based on q value cut-off of 0.1. e Peak count frequency profile of HNF4A FLAG ChIP-Seq peaks that map within 1 kb of the transcription start site (TSS) across the EndoC-βH1 stable lines and the corresponding genomic feature distributions. Data are presented as mean ± SD. Each data point represents one independent experiment. Source data and exact P values are provided in the Source Data file.

Fig. 8. Determining the regulation of target genes by HNF4A T2D risk variant rs1800961.

a IGV tracks showing FLAG ChIP-Seq peaks that map to the AKAP1, GAD2, and HOPX loci, in which differential binding between HNF4A8 WT and T117I was observed. Tracks within each locus have been adjusted to the same scale as specified on the right. The chromosomal location near the peak region is indicated. b FLAG ChIP-qPCR fold enrichment at the selected target regions in the stable EndoC-βH1 cell lines expressing HNF4A8 WT or T117I (n = 3). * indicates p < 0.05, ** indicates p < 0.01 based on two-way ANOVA. c Gene expression analyses of differentially-bound target genes AKAP1, GAD2, and HOPX in EndoC-βH1 cells expressing HNF4A8 WT or T117I (n = 3). * indicates p < 0.05, ** indicates p < 0.01 based on two-way ANOVA with Bonferroni’s multiple comparisons test. d Structural modeling shows the crystal structure of homodimeric HNF4A (upstream subunit in green, downstream subunit in cyan) bound to its DNA response element (orange), coactivator peptides (salmon and white) and myristoate (PDB code 4IQR) (left), and the final trajectory structure from a representative molecular dynamics (MD) simulation of phosphorylated HNF4A complexed with DNA (right). Phosphorylated T117 (pT117) in the hinge region of the downstream subunit and nearby DNA nucleotides are shown in sticks. Data are presented as mean ± SD. Each data point represents one independent experiment. Source data and exact P values are provided in the Source Data file.