Impact of a High-fat Diet on Tissue Acyl-CoA and Histone Acetylation Levels

Abstract

Cellular metabolism dynamically regulates the epigenome via availability of the metabolite substrates of chromatin-modifying enzymes. The impact of diet on the metabolism-epigenome axis is poorly understood but could alter gene expression and influence metabolic health. ATP citrate-lyase produces acetyl-CoA in the nucleus and cytosol and regulates histone acetylation levels in many cell types. Consumption of a high-fat diet (HFD) results in suppression of ATP citrate-lyase levels in tissues such as adipose and liver, but the impact of diet on acetyl-CoA and histone acetylation in these tissues remains unknown. Here we examined the effects of HFD on levels of acyl-CoAs and histone acetylation in mouse white adipose tissue (WAT), liver, and pancreas. We report that mice consuming a HFD have reduced levels of acetyl-CoA and/or acetyl-CoA:CoA ratio in these tissues. In WAT and the pancreas, HFD also impacted the levels of histone acetylation; in particular, histone H3 lysine 23 acetylation was lower in HFD-fed mice. Genetic deletion of Acly in cultured adipocytes also suppressed acetyl-CoA and histone acetylation levels. In the liver, no significant effects on histone acetylation were observed with a HFD despite lower acetyl-CoA levels. Intriguingly, acetylation of several histone lysines correlated with the acetyl-CoA: (iso)butyryl-CoA ratio in liver. Butyryl-CoA and isobutyryl-CoA interacted with the acetyltransferase P300/CBP-associated factor (PCAF) in liver lysates and inhibited its activity in vitro This study thus provides evidence that diet can impact tissue acyl-CoA and histone acetylation levels and that acetyl-CoA abundance correlates with acetylation of specific histone lysines in WAT but not in the liver.

Keywords: Acetyl-CoA; adipose tissue; diet; histone acetylation; liver.

FIGURE 1.

Mouse response to dietary intervention. A–D, C57Bl/6 mice were fed either RC (white columns) or an HFD (gray columns) for 5 days (n = 5/group) or 4 weeks (n = 10/group). Weight gain (in grams, compared with the weight at day 0) and blood glucose after 5 days (A and B, respectively) or 4 weeks (C and D) in the indicated diet. Graphs represent mean ± S.D. Significance was assessed by two-tailed unpaired t test. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

FIGURE 2.

Suppression of ACLY, ACSS2, and FASN levels by HFD. A–F, Western blotting analysis for proteins involved in de novo lipogenesis in pgWAT (A, quantified in B), liver (C, quantified in D), and pancreas (E, quantified in F). B.T. indicates below background threshold set for detection in ImageJ. Graphs represent mean ± S.D. Dots show individual data points. Significance was assessed by two-tailed t test. *, p < 0.05; **, p < 0.01.

FIGURE 3.

Acetyl-CoA and/or the acetyl-CoA:CoA ratio is/are suppressed by HFD. A–F, quantification of acyl-CoA species in the liver (A and B; n = 9 for RC, n = 10 for HFD), perigonadal WAT (C and D; n = 10 for RC, n = 10 for HFD), and pancreas (E and F; n = 4 for RC, n = 5 for HFD) of mice fed with either RC (white columns) or HFD (gray columns) for 4 weeks. Acyl-CoA measurements were normalized to internal standards, as described under “Experimental Procedures” and expressed as relative levels, with RC mean set to 1. For acetyl-CoA, CoA, the ratio between relative abundance of acetyl-CoA and relative abundance of CoA for each animal is depicted. (iso)Butyryl-CoA shows the abundance of both isobutyryl-CoA and butyryl-CoA, as the two isomers could not be discriminated. Data are graphed as mean ± S.D., with data from individual animals represented as dots. For A, C, and E, statistical significance was assessed by two-tailed t test (*, p < 0.05; **, p < 0.01; or as indicated). For B, D, and F, significance was assessed by two-tailed t tests and corrected for multiple comparisons using the Holm-Sidak method (α = 0.05). The p values are indicated on the graphs.

FIGURE 4.

Specific histone acetylation marks are suppressed by HFD in WAT and pancreas. A–C, histones were extracted from the liver (A), perigonadal WAT (B), and pancreas (C) of mice fed either RC (white columns) or HFD (gray columns) for 4 weeks (n = 5/group). Values are expressed as percent of each lysine that is acetylated. Data are graphed as mean ± S.D., with data from individual animals represented as dots. Statistical significance was assessed by two-tailed t test (*, p < 0.05; **, p < 0.01; or as indicated).

FIGURE 5.

Correlation between acetyl-CoA and histone acetylation in pgWAT. Spearman correlation (r) was calculated for the indicated relationships (n = 9 animals). Significant correlations are with acetyl-CoA, H4K8ac (r = 0.933, p = 0.0007) and with acetyl-CoA:CoA, H4K8ac (r = 0.8333, p = 0.0083).

FIGURE 6.

Genetic deletion of Acly decreases global histone acetylation. Acly^fl/fl preadipocytes (5A) were differentiated and then either mock-treated (mock, white columns) or infected with Adeno-Cre virus (Cre, gray columns). After 4 days, cells were analyzed. A, Oil Red O staining for neutral lipids. The whole plate is shown on the left. B, mRNA expression of the indicated genes. C, Western blotting analysis for the indicated cytosolic proteins (top panel) and histone marks and nuclear fraction (center panel). Ponceau is shown in the bottom panel as a loading control for isolated histones. Shown are two biological replicates. D, quantification of acetyl-CoA, CoA, and the acetyl-CoA:CoA ratio (n = 6/group). Measurements were normalized to internal standards and expressed as relative levels, with mock mean set to 1. Data are graphed as mean ± S.D., with data from individual replicate represented as dots. For B and D, statistical significance was assessed by two-tailed t test (*, p < 0.05; **, p < 0.01; ***, p < 0.001).

FIGURE 7.

Correlation between the acetyl-CoA:(iso)butyryl-CoA ratio and histone acetylation in the liver. A, correlation between the levels of histone acetylation and acetyl-CoA measurements or acetyl-CoA:metabolite ratios in the livers of RC- and HFD-fed mice (n = 17). Spearman correlations (r) are graphed. Significant correlations with acetyl-CoA:(iso)butyryl-CoA ratio include H3K9ac (r = 0.6176, p = 0.0096), H3K14ac (r = 0.4926, p = 0.0465), H3K23ac (r = 0.5098, p = 0.0366), and H4K5ac (r = 0.4975, p = 0.0421). B, schematic for the workflow used to profile KAT activity as described previously (25). C–F, pulldown of PCAF by the KAT chemoproteomic probe in the presence of the indicated competitors, shown by Western blot (C and E) and quantified in D and F, respectively. Proteomes were obtained from pools of livers (n = 3/group) from mice fed either with RC (white columns) or HFD (gray columns). N.A., data not acquired G, rate of in vitro acetylation of H3K14 by PCAF in the presence of acetyl-CoA with the indicated concentrations of butyryl-CoA or (iso)butyryl-CoA. Graphs represent mean ± S.D. Significance was assessed by two-tailed t test, and each condition was compared with mock. ***, p < 0.001.