Combining three independent pathological stressors induces a heart failure with preserved ejection fraction phenotype

Abstract

Heart failure (HF) with preserved ejection fraction (HFpEF) is defined as HF with an ejection fraction (EF) ≥ 50% and elevated cardiac diastolic filling pressures. The underlying causes of HFpEF are multifactorial and not well-defined. A transgenic mouse with low levels of cardiomyocyte (CM)-specific inducible Cavβ2a expression (β2a-Tg mice) showed increased cytosolic CM Ca²⁺, and modest levels of CM hypertrophy, and fibrosis. This study aimed to determine if β2a-Tg mice develop an HFpEF phenotype when challenged with two additional stressors, high-fat diet (HFD) and N^ω-nitro-l-arginine methyl ester (l-NAME, LN). Four-month-old wild-type (WT) and β2a-Tg mice were given either normal chow (WT-N, β2a-N) or HFD and/or l-NAME (WT-HFD, WT-LN, WT-HFD-LN, β2a-HFD, β2a-LN, and β2a-HFD-LN). Some animals were treated with the histone deacetylase (HDAC) (hypertrophy regulators) inhibitor suberoylanilide hydroxamic acid (SAHA) (β2a-HFD-LN-SAHA). Echocardiography was performed monthly. After 4 mo of treatment, terminal studies were performed including invasive hemodynamics and organs weight measurements. Cardiac tissue was collected. Four months of HFD plus l-NAME treatment did not induce a profound HFpEF phenotype in FVB WT mice. β2a-HFD-LN (3-Hit) mice developed features of HFpEF, including increased atrial natriuretic peptide (ANP) levels, preserved EF, diastolic dysfunction, robust CM hypertrophy, increased M₂-macrophage population, and myocardial fibrosis. SAHA reduced the HFpEF phenotype in the 3-Hit mouse model, by attenuating these effects. The 3-Hit mouse model induced a reliable HFpEF phenotype with CM hypertrophy, cardiac fibrosis, and increased M₂-macrophage population. This model could be used for identifying and preclinical testing of novel therapeutic strategies.NEW & NOTEWORTHY Our study shows that three independent pathological stressors (increased Ca²⁺ influx, high-fat diet, and l-NAME) together produce a profound HFpEF phenotype. The primary mechanisms include HDAC-dependent-CM hypertrophy, necrosis, increased M₂-macrophage population, fibroblast activation, and myocardial fibrosis. A role for HDAC activation in the HFpEF phenotype was shown in studies with SAHA treatment, which prevented the severe HFpEF phenotype. This "3-Hit" mouse model could be helpful in identifying novel therapeutic strategies to treat HFpEF.

Keywords: M2-macrophage; cardiac hypertrophy; heart failure with preserved ejection fraction; histone deacetylases; suberoylanilide hydroxamic acid.

Graphical abstract

Figure 1.

Effects of high-fat diet (HFD), N^ω-nitro-l-arginine methyl ester (l-NAME, LN), or HFD + l-NAME on the heart failure with preserved ejection fraction (HFpEF) phenotype in wild-type (WT) mice. A: survival rate from 4-mo follow-up. Body weight (BW; B), heart weight (HW; C), and ratio of HW to tibia length (HW/TL; D) at the time of euthanasia. Conventional and sophisticated echocardiography data showing left ventricular (LV) wall thickness (E and F), LV ejection fraction (LVEF; G), LV longitudinal strain (H), left atrium (LA) diameter (I), and ratio between early mitral inflow velocity (E) and mitral annular early diastolic velocity (e′) (E/e′; J). Hemodynamics data showing LV end-diastolic pressure (LVEDP; K) and maximum rate of pressure decay (dP/dt_min; L). M: representative images of hearts stained with Picrosirius red and wheat germ agglutinin (WGA). N: quantification of the percentage of Picrosirius red-positive area. O: quantification of cardiomyocyte cross-sectional area (CSA). DAPI, 4′,6-diamidino-2-phenylindole; LVAWd, end-diastolic left ventricular anterior wall thicknesses; LVEF, left ventricular ejection fraction; LVPWd, end-diastolic left ventricular posterior wall thickness; N, normal chow diet. Data shown are means ± SE. Tukey post hoc multiple comparison adjusted P values: E–J: P < 0.05, *WT-N vs. WT-HFD-LN, #WT-N vs. WT-HFD, and &WT-N vs. WT-LN; and other panels: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Total number of animals (n) and number of females and males included in each group are reported in the Supplemental Table.

Figure 2.

Effects of cardiac-specific L-type Ca²⁺ channel (LTCC) β2a-subunit expression together with high-fat diet (HFD) and N^ω-nitro-l-arginine methyl ester (l-NAME, LN) treatment (3-Hit) on the heart failure with preserved ejection fraction (HFpEF) phenotype. A: survival rate from 4-mo follow-up. Body weight (BW; B), heart weight (HW; C), and ratio of HW to tibia length (HW/TL; D) at the time of euthanasia. Conventional and sophisticated echocardiography data showing left ventricular (LV) wall thickness (E and F), LV ejection fraction (LVEF; G), LV longitudinal strain (H), left atrium (LA) diameter (I), and ratio between early mitral inflow velocity (E), and mitral annular early diastolic velocity (e′) (E/e′; J). Hemodynamics data showing LV end-diastolic pressure (LVEDP; K) and maximum rate of pressure decay (dP/dt_min; L). M: representative images of hearts stained with Picrosirius red and wheat germ agglutinin (WGA). N: quantification of the percentage of Picrosirius red-positive area. O: quantification of cardiomyocyte cross-sectional area (CSA). β2a, transgenic mouse with low levels of cardiomyocyte (CM)-specific inducible Cavβ2a expression; DAPI, 4′,6-diamidino-2-phenylindole; LVAWd: end-diastolic left ventricular anterior wall thicknesses; LVPWd, end-diastolic left ventricular posterior wall thickness; N, normal chow diet. Data shown are means ± SE. Tukey post hoc multiple comparison adjusted P values: E–J: P < 0.05, *β2a-N vs. β2a-HFD-LN, #β2a-N vs. β2a-HFD, &β2a-N vs. β2a-LN, ^β2a-HFD vs. β2a-HFD-LN, +β2a-LN vs. β2a-HFD-LN; and other panels: *P < 0.05, **P < 0.01, ****P < 0.0001. Total number of animals (n) and number of females and males included in each group are reported in the Supplemental Table.

Figure 3.

The 3-Hit mouse model produces a profound heart failure with preserved ejection fraction (HFpEF) phenotype, which can be reduced by suberoylanilide hydroxamic acid (SAHA) treatment. β2a, transgenic mouse with low levels of cardiomyocyte (CM)-specific inducible Cavβ2a expression; HFD, high-fat diet; l-NAME (or LN), N^ω-nitro-l-arginine methyl ester; N, normal diet; WT, wild type. Data shown for WT-N, WT-HFD-LN, β2a-N, and β2a-HFD-LN groups in are the same as those shown in Figs. 1 and 2, and Supplemental Fig. S2. Statistical comparisons being made here are unique. The β2a-HFD-LN-SAHA group was added into the statistic comparation. A: survival rate from 4-mo follow-up. B: body weight (BW) at the time of euthanasia. Conventional and sophisticated echocardiography data showing left ventricular (LV) wall thickness (C and D), LV ejection fraction (LVEF; E), LV longitudinal strain (F), and left atrium (LA) diameter (G). H: LA weight-to-BW ratio at the time of euthanasia. I: conventional echocardiography data showing ratio between early mitral inflow velocity (E) and mitral annular early diastolic velocity (e′) (E/e′). Hemodynamics data showing end-diastolic pressure (EDP; J) and LV diastolic time constant (τ; K). L: expression level of atrial natriuretic peptide (ANP) in heart tissues by real-time polymerase chain reaction. Relative expression was calculated with respect to the WT-N group. HW, heart weight; LVAWd, end-diastolic left ventricular anterior wall thicknesses; LVPWd, end-diastolic left ventricular posterior wall thickness; TL, tibia length. Data shown are means ± SE. Tukey post hoc multiple comparison adjusted P values: C–G and I: P < 0.05, *WT-N vs. β2a-HFD-LN, #WT-N vs. β2a-N,@WT-N vs. WT-HFD-LN, &β2a-N vs. β2a-HFD-LN, and $WT-HFD-LN vs. β2a-HFD-LN; and other panels: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Total number of animals (n) and number of females and males included in each group are reported in the Supplemental Table.

Figure 4.

The 3-Hit mouse model has a more severe cardiomyocyte hypertrophy and necrosis, which can be reduced by suberoylanilide hydroxamic acid (SAHA) treatment. β2a, transgenic mouse with low levels of cardiomyocyte (CM)-specific inducible Cavβ2a expression; HFD, high-fat diet; N, normal chow diet, l-NAME (LN), N^ω-nitro-l-arginine methyl ester; WT, wild type. Representative images of wheat germ agglutinin (WGA)-stained hearts (C) and quantification of cardiomyocyte cross-sectional area (CSA) data (D) in WT-N, WT-HFD-LN, β2a-N, and β2a-HFD-LN groups are the same as those shown in Fig. 1, M and O, and Fig. 2, M and O. The statistical comparisons being made here are different than those made in Figs. 1 and 2. The β2a-HFD-LN-SAHA group was added into the statistic comparison. A and B: expression level of class I histone deacetylases (HDACs) 1, 2, 3 and 8, Class IIb HDAC6 (A) and class IIa HDACs 4 and 7 (B) in heart tissues by real-time polymerase chain reaction. Relative expression was calculated with respect to WT-N mice compared with different treatment groups. C: representative images of WGA-stained hearts and histological assessment of cardiac ventricular pathology by Von Kossa staining. D: quantification of cardiomyocyte CSA. E and F: representative images of terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL)-stained hearts (E) and quantification of TUNEL-positive myocyte nuclei from hearts (F): α-sarcomeric actin (α-SA), red; TUNEL, green; and 4′,6-diamidino-2-phenylindole (DAPI), blue. Data shown are means ± SE. Tukey post hoc multiple comparison adjusted P values are reported here. *P < 0.05, **P < 0.01, ****P < 0.0001.

Figure 5.

The 3-Hit model has robust cardiac M₂-macrophage infiltration and fibroblast activation, which can be reduced by suberoylanilide hydroxamic acid (SAHA) treatment. A: immunofluorescence staining of heart sections to show CD45⁺ (red) immune cell, CD45⁺ (red) CD68⁺ (green) macrophage, CD68⁺ (green) CD206⁺ (red) M₂-macrophage, and phosphorylated Smad2/3 (pSmad2/3, green) in fibroblast (red). 4′,6-Diamidino-2-phenylindole (DAPI) was stained as blue. Data are expressed as percentage of CD45⁺ cell/total cells (B), percentage of CD68⁺ cell/CD45⁺ cell (C), and CD206⁺CD68⁺/total cell (D). E: expression level of transforming growth factor-β (TGFβ) in heart tissues by real-time polymerase chain reaction. Relative expression was calculated with respect to wild-type/normal chow diet (WT-N) mice compared with different treatment groups. F: quantification of mean intensity of pSmad2/3 for each group. β2a, transgenic mouse with low levels of cardiomyocyte (CM)-specific inducible Cavβ2a expression; HFD, high-fat diet; l-NAME (LN), N^ω-nitro-l-arginine methyl ester; P4HB, protein disulfide isomerase/prolyl 4-hydroxylase. Data shown are means ± SE. Tukey post hoc multiple comparison adjusted P values are reported here. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 6.

The 3-Hit model has severe myocardial fibrosis, which can be reduced be suberoylanilide hydroxamic acid (SAHA) treatment. β2a, transgenic mouse with low levels of cardiomyocyte (CM)-specific inducible Cavβ2a expression; HFD, high-fat diet; N, normal chow diet, l-NAME (LN), N^ω-nitro-l-arginine methyl ester; WT, wild type. The representative images of Picrosirius red-stained hearts (A) and quantification of the percentage of Picrosirius red-positive area (C) data of WT-N, WT-HFD-LN, β2a-N, and β2a-HFD-LN groups are the same as in Fig. 1, M and N, and Fig. 2, M and N. The statistical comparisons being made are different than those made in Figs. 1 and 2. The β2a-HFD-LN-SAHA group was added into the statistic comparation. A: representative images of hearts from 4 groups with α-smooth muscle actin (α-SMA) (red) immunofluorescence staining and Picrosirius red staining. Quantification of the α-SMA (red) intensity (B) and the percentage of Picrosirius red-positive area (C). Expression level of fibronectin 1 (FN1; D) and collagen cross-linking enzyme lysyl oxidase (LOX; E) in heart tissues by real-time polymerase chain reaction. Relative expression was calculated with respect to WT-N mice compared with different treatment groups. DAPI: 4′,6-diamidino-2-phenylindole. Data shown are means ± SE. Tukey post hoc multiple comparison adjusted P values are reported here. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.