Multiple roles and regulatory mechanisms of the transcription factor HNF4 in the intestine

Abstract

Hepatocyte nuclear factor 4-alpha (HNF4α) drives a complex array of transcriptional programs across multiple organs. Beyond its previously documented function in the liver, HNF4α has crucial roles in the kidney, intestine, and pancreas. In the intestine, a multitude of functions have been attributed to HNF4 and its accessory transcription factors, including but not limited to, intestinal maturation, differentiation, regeneration, and stem cell renewal. Functional redundancy between HNF4α and its intestine-restricted paralog HNF4γ, and co-regulation with other transcription factors drive these functions. Dysregulated expression of HNF4 results in a wide range of disease manifestations, including the development of a chronic inflammatory state in the intestine. In this review, we focus on the multiple molecular mechanisms of HNF4 in the intestine and explore translational opportunities. We aim to introduce new perspectives in understanding intestinal genetics and the complexity of gastrointestinal disorders through the lens of HNF4 transcription factors.

Keywords: HNF4; colon cancer; inflammatory bowel disease; intestinal differentiation; intestinal regeneration; intestine; redundancy; transcription factor.

Figure 1

Tissue-specific gene expression profiles of HNF4A and HNF4G. (A) Expression profiles of HNF4A and HNF4G from the Genotype-Tissue Expression Project (GTEx) were evaluated in tissues of healthy individuals (n=number of replicates per tissue examined for gene expression) (20). (B) HNF4A is expressed across multiple tissues of the gastrointestinal tract whereas HNF4G is primarily an intestine-restricted paralog (TPM: Transcripts Per Million). The Genotype-Tissue Expression (GTEx) Project was supported by the Common Fund of the Office of the Director of the National Institutes of Health, and by NCI, NHGRI, NHLBI, NIDA, NIMH, and NINDS. The data used for the analyses described in this manuscript were obtained from the GTEx Portal on 05/25/23. Adapted from “Human Internal Organs”, by BioRender.com (2023).

Figure 2

Distribution of HNF4 isoforms along the intestinal epithelium. (A) Isoforms of Hnf4a are generated through alternative splicing and usage of alternative promoters (proximal P1 and distal P2) that are forty kilobases (kb) apart in human and mouse (26). Enterocytes in the villi express high levels of P1 isoforms, whereas crypt stem cells express lower amounts (27). Studies in mouse and human colonic crypts show P1-driven HNF4α isoforms are expressed more robustly in the luminal, differentiated colonic epithelium, while P2-driven isoforms are enriched deeper in the proliferative crypt epithelium, co-localizing with the proliferative marker, Ki67. HNF4γ is expressed in both villus and crypt. (B) There are 12 isoforms of the HNF4A gene that are produced through the utilization of alternative promoters – P1 and P2. HNF4α1-6 are generated from the P1 promoter, while HNF4α7-12 are generated from the P2 promoter. The table shows the mRNA expression patterns of each isoform in the intestine from adult human tissue (25). Figure panel modeled off of the work of (25).

Figure 3

Structural differences between HNF4α and HNF4γ. HNF4α is typically a 474 amino acid protein amino acid protein, whereas HNF4γ1 is a 408 amino acid protein. Both paralogs have a DNA binding domain (DBD), a ligand binding domain (LBD), an AF-2 transactivation domain and a proline rich repressor domain (F) at the C-terminal. There is a 94% similarity between the DBDs and an 80% similarity between the LBDs (31). However, HNF4γ is missing an N-terminal, AF-1 transactivator domain. Figure generated using Illustrator for Biological Sequences, Version 1.0 (33).

Figure 4

HNF4α is regulated at the protein level by post-translational modifications. The most frequently recurring protein modification is phosphorylation of serine and threonine residues by several enzymes including protein kinase C, AMP-activated kinase, ERK1/2 kinase, protein kinase A, and p38 kinase (–45). Phosphorylation at the DBD residue S78, by protein kinase C impairs the DNA binding capacity of HNF4α and its nuclear localization (44). cAMP-induced protein kinase A phosphorylates S134 in the DBD and inhibits HNF4α’s recruitment to target genes (41). AMP activated kinase phosphorylates HNFα at S313 in the LBD and destabilizes the protein (42). ERK1/2 kinase can phosphorylate multiple residues such as S95, S262/S265, S451, and T457/T459, which reduce the transactivation capacity of HNF4α (45). Src kinase phosphorylates Y14 followed by Y277/279 of P1-HNF4α causing mislocalization of the protein to the cytoplasm (46). In contrast, p38 kinase mediated phosphorylation at S158 increases the transactivation potential of HNF4α (43). Further modifications include acetylation at DBD residues K97, K99, K117 and K118 mediated by the CREB-binding protein (CBP), in association with co-activators, p300 and DDX3 and methylation of R91 in the DBD by protein arginine methyltransferase 1 (PRMT1). Acetylation is likely required for the nuclear retention of the protein (47). HNF4α also undergoes SUMOylation at the C-terminal consensus site (Ψ-K-x-D/E) which destabilizes HNF4α in a ubiquitin-dependent manner (48). Ubiquitylation of HNF4α occurs at residues K234 and K307, and targets HNF4 for proteasomal degradation (49). Figure generated using Illustrator for Biological Sequences, Version 1.0 (33).

Figure 5

HNF4 plays diverse roles in the intestine. (a) HNF4α is a driver of repair and regeneration mechanisms in the intestine (58). It controls crypt survival and proliferating cell survival, which would allow epithelial repopulation (A), a phenomenon which is lost in HNF4α mutants (B). (b) HNF4α and HNF4γ are highly expressed in Lgr5+ intestinal stem cells (ISCs) and are required for ISC maintenance and renewal (A) (59). Dysregulation of fatty acid oxidation in stem cells causes exhaustion, loss of ISC and premature lineage commitment (Panel B) (60, 61). This effect is pronounced in HNF4α/γ double mutants, implying redundancy masks the phenotype in single mutants. (c) The HNF4α-CDX2 pair controls genes responsible for formation of the apical brush border in the intestine (62). HNF4 functions as a conserved and universal regulator of brush border genes by likely operating as a mechanosignaling sensor detecting changes in actin fibers and brush border transcripts upon mechanical stress (63). Electron microscopy cross sections through microvilli reveal a larger brush border diameter and shorter height in HNF4α/γ double mutant mice compared to wild type (A, B). (d) HNF4 paralogs, in association with the BMP/SMAD pathway mediate the formation of enterocytes in the villi. HNF4 and SMAD4 reciprocally activate each other’s expression in villi in vivo and in organoid models ex vivo, and disruption of this feed-forward loop results in loss of enterocyte fate in favor of a goblet cell fate (B) (17).

Figure 6

Spectrum of HNF4 dysfunctions in the intestine and beyond. (A) The mutational distribution of HNF4 generates 4 main disease phenotypes. An SNP in in the 3’-UTR of HNF4A has been associated with ulcerative colitis (rs6017342) (–85). Three SNPs each are significantly associated with increased susceptibility to childhood-onset Crohn’s disease (102) (rs2144908, rs1884613 and rs1884614) and colon cancer (46) (rs6031602, rs1063239, and rs6093980). A single SNP in the human locus of HNF4G (rs4735692) has been associated with obesity (103). (B) At the protein level, HNF4α dysfunctions are seen as metabolic impairments such as MODY1 (104), and Non-insulin dependent diabetes mellitus (NIDDM) (105). In MODY1, a low frequency missense mutation in HNF4α, Q268X, results in the deletion of 187 C-terminal amino acids (106), whereas a V393I substitution causes increase in susceptibility to NIDDM (107). Also, Fanconi renotubular syndrome is a unique phenotype comprising both MODY1 and atypical Fanconi syndrome, which occurs due to a heterozygous missense mutation, R76W in HNF4α (108).