Wilson Disease: Intersecting DNA Methylation and Histone Acetylation Regulation of Gene Expression in a Mouse Model of Hepatic Copper Accumulation - PubMed (original) (raw)
Wilson Disease: Intersecting DNA Methylation and Histone Acetylation Regulation of Gene Expression in a Mouse Model of Hepatic Copper Accumulation
Gaurav V Sarode et al. Cell Mol Gastroenterol Hepatol. 2021.
Abstract
Background & aims: The pathogenesis of Wilson disease (WD) involves hepatic and brain copper accumulation resulting from pathogenic variants affecting the ATP7B gene and downstream epigenetic and metabolic mechanisms. Prior methylome investigations in human WD liver and blood and in the Jackson Laboratory (Bar Harbor, ME) C3He-Atp7btx-j/J (tx-j) WD mouse model revealed an epigenetic signature of WD, including changes in histone deacetylase (HDAC) 5. We tested the hypothesis that histone acetylation is altered with respect to copper overload and aberrant DNA methylation in WD.
Methods: We investigated class IIa HDAC4 and HDAC5 and H3K9/H3K27 histone acetylation in tx-j mouse livers compared with C3HeB/FeJ (C3H) control in response to 3 treatments: 60% kcal fat diet, D-penicillamine (copper chelator), and choline (methyl group donor). Experiments with copper-loaded hepatoma G2 cells were conducted to validate in vivo studies.
Results: In 9-week tx-j mice, HDAC5 levels increased significantly after 8 days of a 60% kcal fat diet compared with chow. In 24-week tx-j mice, HDAC4/5 levels were reduced 5- to 10-fold compared with C3H, likely through mechanisms involving HDAC phosphorylation. HDAC4/5 levels were affected by disease progression and accompanied by increased acetylation. D-penicillamine and choline partially restored HDAC4/5 and H3K9ac/H3K27ac to C3H levels. Integrated RNA and chromatin immunoprecipitation sequencing analyses revealed genes regulating energy metabolism and cellular stress/development, which, in turn, were regulated by histone acetylation in tx-j mice compared with C3H mice, with Pparα and Pparγ among the most relevant targets.
Conclusions: These results suggest dietary modulation of class IIa HDAC4/5, and subsequent H3K9/H3K27 acetylation/deacetylation can regulate gene expression in key metabolic pathways in the pathogenesis of WD.
Keywords: Copper; Histone Deacetylase; Liver; Metabolism.
Copyright © 2021 The Authors. Published by Elsevier Inc. All rights reserved.
Figures
Graphical abstract
Figure 1
Animal study design. C3H, C3HeB/FeJ; DI, deionized; HFD, high-fat diet; PCA, D-penicillamine; PPD6, postpartum day 6; tx-j, C3He-Atp7btx-j/J; WT, wild-type.
Figure 2
Expression of HDAC4 and HDAC5 in mouse and HepG2 models of WD. Immunoblot densitometry analyses are normalized to β-actin; data are represented as means ± SEM and statistical significance was determined by Student's t -test (∗_P < .05, ∗∗_P < .01, and ∗∗∗∗P < .0001). (A) Total protein liver lysate densitometries for HDAC4 and HDAC5 protein expression in tx-j mice compared with C3H control at postpartum day 6 (PPD6; C3H, n = 4 males/4 females; tx-j, n = 5 males/4 females), 9 weeks (C3H, n = 4 males/3 females; tx-j, n = 4 males/4 females), 12 weeks (C3H, n = 4 males/4 females; tx-j, n = 4 males/3 females), and 24 weeks (C3H, n = 10 males/12 females; tx-j, n = 11 males/11 females). (B) Total protein liver lysate densitometry analyses for HDAC4 and HDAC5 protein expression in 16-week-old _Atp7b_-/- mice (n = 7 males/4 females) and wild-type (WT, n = 6 males/5 females). (C) Immunohistochemical analysis of 24-week-old C3H and tx-j mouse livers for HDAC5 (red) and 4′,6-diamidino-2-phenylindole (DAPI, blue). Images display cytosolic and nuclear HDAC5 localization. White arrows indicate nuclei. Scale bar: 50 μm. Bar graphs represent HDAC5 optical density in cytosol, nuclei, and the nucleus/cytosol ratio; n = 3 mice/group, 60 cells/mouse. (D) HDAC5 immunoblot of HepG2 cell lysates treated with CuSO4 (0–100 μmol/L; n = 3 per treatment) for 24 hours.
Figure 3
Activation of AMPKα signaling in 24-week-old tx-j mice and HepG2 cells. Data are represented as means ± SEM. Statistical significance was determined by Student's t -test (∗_P < .05, ∗∗_P < .01, and ∗∗∗∗P < .0001). (A) Transcript levels of Ampka1 normalized to Gapdh (C3H, n = 10 males/12 females; tx-j, n = 11 males/11 females). (B) Immunoblot densitometries of total AMPKα, pAMPKα, and pHDAC5 in total protein liver lysates obtained from 24-week-old C3H (n = 3 males/3 females) and tx-j (n = 3 males/2 females). Densitometry values were normalized to β-actin. (C) Relative density comparisons of total AMPKα with pAMPKα and total HDAC5 with pHDAC5, normalized to β-actin, in 24-week-old C3H and tx-j mice. (D) HDAC5 immunoblot densitometry analysis, normalized to β-actin, of HepG2 cell lysates treated with 50 μmol/L CuSO4 for 24 hours, followed by AICAR treatment (0–0.5 mmol/L; n = 3 per dose), an AMPK activator, for 24 hours. C, control; V, treated with vehicle (DMSO) only.
Figure 4
HDAC4 and HDAC5 down regulation is associated with increased histone acetylation and impaired methylation. Total protein liver lysate analyses of 24-week-old C3H (n = 10 males/12 females) and tx-j mice (n = 11 males/11 females), and tx-j mice treated with PCA (n = 11 males/10 females), choline (n = 8 males/13 females), and PCA+choline (n = 10 males/11 females). Immunoblot densitometries of HDAC4, HDAC5, H3ac, H3K9ac, H3K27ac, H3K9me3, and H3K27me3 were normalized to β-actin. Data are represented as means ± SEM and statistical analyses were performed with Kruskal–Wallis 1-way analysis of variance followed by an uncorrected Dunn's test (∗P < .05, ∗∗P < .01, ∗∗∗P < .001, and ∗∗∗∗P < .0001). (D) Transcript levels of Hat1 normalized to Gapdh (C3H, n = 10 males/12 females; tx-j, n = 11 males/11 females).
Figure 5
H3K27ac-associated ChIP-seq and RNA-seq data analyses in the liver of tx-j mice. ChIP-seq and RNA-seq were performed in livers of 24-week-old C3H, tx-j, tx-j+PCA, tx-j+choline, and tx-j+PCA+choline groups (n = 3 males/3 females per group). (A and B) Venn diagrams showing numbers of H3K27ac-associated differentially enriched genes detected by ChIP-seq and differentially expressed genes by RNA-seq. I, C3H vs tx-j; II, tx-j vs tx-j+PCA; III, tx-j vs tx-j+choline; IV, tx-j vs tx-j+PCA+choline. (C) Numbers of common significantly co–upregulated and co–downregulated genes between ChIP-seq and RNA-seq analyses. (D) Pathway enrichment analysis showing the top 10 significantly associated pathways with differentially expressed co–upregulated and co–downregulated genes between ChIP-seq and RNA-seq by g:Profiler.
Figure 6
Transcription factors within differentially expressed genes between C3H and tx-j mice. Using the ChIP-seq and RNA-seq data from livers of 24-week-old C3H and tx-j mice (3 males/3 females per group), differentially expressed genes in ChIP-seq and RNA-seq analyses were overlapped with total known mouse transcription factors. (A) Venn diagrams showing the number of common co–upregulated (174) and co–downregulated (80) transcription factor genes. (B) Bar graph showing the top 10 enriched TFs associated with differentially expressed co-regulated genes between ChIP-seq and RNA-seq by Lisa. (C) ChIP-seq peaks showing enrichment of H3K27ac for transcription factors, selected by their involvement in liver and metabolic diseases. (D) Heatmap displaying qPCR gene transcript levels of representative mouse transcription factors. Relative expression data are normalized to Gapdh (3 males/3 females per group). Data are represented as means ± SEM and statistical analyses were performed with Kruskal–Wallis 1-way analysis of variance followed by an uncorrected Dunn's test (∗P < .05, ∗∗P < .01, and ∗∗∗P < .001) for C3H vs tx-j and tx-j vs tx-j+PCA, choline, or combined treatments.
Figure 7
Dietary modulation (high fat, choline, PCA) of HDAC5 and metabolic regulators PPARα and PPARγ. (A–D) Data are represented as means ± SEM. Statistical significance for all analyses was determined by Student's t_-test for comparison between 2 groups and Kruskal–Wallis 1-way analysis of variance followed by an uncorrected Dunn's test for multiple groups (∗_P < .05, ∗∗P < .001, ∗∗∗P < .001, and ∗∗∗∗P < .001). (A) Liver triglyceride and total cholesterol of mice on chow (C3H, n = 4 males/3 females; tx-j, n = 4 males/4 females) compared with C3H and tx-j (n = 4 males/4 females each) fed a HFD. Mice were challenged with a HFD for 8 days and tissues were collected at 9 weeks of age. (B) HDAC5 protein expression in total protein liver lysate from mice on chow or HFD. Immunoblot densitometry analysis was normalized to β-actin. (C) Liver transcript levels of Pparγ, Pparα, and Hmox1 normalized to Gapdh in mice on chow or HFD. (D) Immunoblot densitometries of PPARγ and PPARα, normalized to β-actin, in total protein liver lysate of 24-week-old C3H (n = 3 males/3 females), tx-j (n = 3 males/2 females), tx-j+PCA (n = 3 males/3 females), tx-j+choline (n = 3 males/3 females), and tx-j+PCA+choline (n = 3 males/3 females). (E) Heatmap representing qPCR transcript levels of PPARα- and PPARγ-related genes measured in livers of 24-week-old C3H vs tx-j and tx-j vs tx-j+PCA, choline, or combined treatments (n = 3 males/3 females per group). Relative expression data are normalized to Gapdh.
Figure 8
Proposed schematic of HDAC5-mediated H3K9ac and H3K27ac regulation of gene expression and affected biological pathways in WD. In WD, copper overload and oxidative stress might lead to phosphorylation of AMPK (active form). Increased phosphorylated AMPK could then phosphorylate HDAC5, which is subsequently exported to the cytosol. Lack of nuclear HDAC5 and increased histone acetyltransferase (HAT1) might cause an increase in acetylated histones (H3K9ac and H3K27ac) and decrease methylated histones (H3K9me3 and H3K27me3) with subsequent altered regulation of genes in WD. Our results show HDAC5 impacts the acetylation/methylation balance and serves as a critical regulator of genes central in metabolic regulation. ChIP-seq and RNA-seq revealed 3732 differentially expressed genes in the tx-j mouse model of WD, and the enrichment analysis of these genes included pathways related to lysine degradation, fatty acid metabolism, carbon metabolism, pyruvate metabolism, and signal transduction. PI3K-AKT, phosphatidylinositol-3-kinase and protein kinase B.
Comment in
- Wilson Disease: Epigenetic Factors Contribute to Genetic Mutations to Affect the Disease.
Dev S, Hamilton JP. Dev S, et al. Cell Mol Gastroenterol Hepatol. 2021;12(4):1507-1508. doi: 10.1016/j.jcmgh.2021.07.010. Epub 2021 Aug 2. Cell Mol Gastroenterol Hepatol. 2021. PMID: 34352219 Free PMC article. No abstract available.
References
- Einer C., Leitzinger C., Lichtmannegger J., Eberhagen C., Rieder T., Borchard S., Wimmer R., Denk G., Popper B., Neff F., Polishchuk E.V., Polishchuk R.S., Hauck S.M., von Toerne C., Müller J.C., Karst U., Baral B.S., DiSpirito A.A., Kremer A.E., Semrau J., Weiss K.H., Hohenester S., Zischka H. A high-calorie diet aggravates mitochondrial dysfunction and triggers severe liver damage in Wilson disease rats. Cell Mol Gastroenterol Hepatol. 2019;7:571–596. -PMC -PubMed
- Müller T., Langner C., Fuchsbichler A., Heinz-Erian P., Ellemunter H., Schlenck B., Bavdekar A.R., Pradhan A.M., Pandit A., Müller-Höcker J., Melter M., Kobayashi K., Nagasaka H., Kikuta H., Müller W., Tanner M.S., Sternlieb I., Zatloukal K., Denk H. Immunohistochemical analysis of Mallory bodies in Wilsonian and non-Wilsonian hepatic copper toxicosis. Hepatology. 2004;39:963–969. -PubMed
Publication types
MeSH terms
Substances
LinkOut - more resources
Full Text Sources
Medical
Molecular Biology Databases