Loss of Hepatic Angiotensinogen Attenuates Diastolic Dysfunction in Heart Failure with Preserved Ejection Fraction

Abstract

Heart failure with preserved ejection fraction (HFpEF) is a prevalent complex syndrome characterized by diastolic dysfunction with limited therapeutic options. While the renin-angiotensin system (RAS) is implicated in heart failure pathogenesis, the causal contribution of angiotensinogen (AGT), the unique precursor of the RAS, to HFpEF remains undefined. Using a two-hits mouse HFpEF model (high-fat diet + L-NAME), consistent upregulation of hepatic and plasma AGT is identified in wild-type mice of both sexes. Critically, hepatocyte-specific AGT deletion directly ameliorated diastolic dysfunction in male and female HFpEF mice, whereas systemic angiotensin II blockade (losartan) failed to improve cardiac diastolic function. Mechanistically, hepatic AGT drove HFpEF through LRP2-mediated internalization in cardiac endothelial cells, suppressing the GATA2/Pim3 signaling axis, which inhibited microvascular angiogenesis and ultimately exacerbated diastolic dysfunction. To validate therapeutic potential, it is demonstrated that 18β-glycyrrhetinic acid - identified as a potent hepatic AGT inhibitor - significantly improved cardiac diastolic function in HFpEF mice. These findings establish hepatic AGT as a causal contributor to HFpEF pathogenesis and reveal its therapeutic targeting as a promising strategy.

Keywords: 18β‐glycyrrhetinic acid; angiotensin II‐independent mechanism; angiotensinogen; heart failure with preserved ejection fraction; microvascular angiogenesis.

Figure 1

HFpEF increased hepatic AGT abundances and plasma AGT concentrations in female mice. A) Experimental workflow and analysis for HFpEF model establishment in female mice (n = 7 for CHOW group, and n = 8 for HFD+L‐NAME group). B) Representative echocardiography images obtained from female mice at baseline (n = 7 for the CHOW group, and n = 8 for the HFD+L‐NAME group). C) Representative echocardiography images obtained from female mice fed with a high‐fat diet and L‐NAME for 5 weeks (n = 7 for CHOW group, and n = 8 for HFD+L‐NAME group). D) Representative echocardiography images obtained from female mice fed with a high‐fat diet and L‐NAME for 15 weeks (n = 7 for CHOW group, and n = 8 for HFD+L‐NAME group). E) Left ventricular ejection fraction (LVEF%) and left ventricular fraction shortening (LVFS%) were quantified via echocardiography (n = 7 for the CHOW group, and n = 8 for the HFD+L‐NAME group). F) E/A ratio was quantified via echocardiography (n = 7 for the CHOW group, and n = 8 for the HFD+L‐NAME group). G) E/e’ ratio was quantified via echocardiography (n = 7 for the CHOW group, and n = 8 for the HFD+L‐NAME group). H) IVRT was quantified via echocardiography (n = 7 for the CHOW group, and n = 8 for the HFD+L‐NAME group). I) Tei index was quantified via echocardiography (n = 7 for the CHOW group, and n = 8 for the HFD+L‐NAME group). J) Ratio of heart weight to tibia length (HW/TL) was measured in female mice 15 weeks after feeding with HFD+L‐NAME (n = 7 for CHOW group, and n = 8 for HFD+L‐NAME group). K) Western blotting detection and quantification of AGT protein abundances in livers obtained from female mice 15 weeks after feeding with HFD and L‐NAME (n = 6 for each group). L) Plasma AGT concentrations in female mice fed with HFD and L‐NAME for 15 weeks were measured by ELISA (n = 6 for each group). Two‐way ANOVA was used for statistical analysis in E–I, and Student's t test was used for statistical analysis in J–L.

Figure 2

Deficiency of hepatic AGT alleviated diastolic dysfunction in female HFpEF mice. A) Experimental workflow and analysis for the effects of hepatic AGT deletion on HFpEF in female mice (n = 6 for hepAGT+/+ CHOW group, n = 6 for hepAGT‐/‐ CHOW group, n = 15 for hepAGT+/+ HFD+L‐NAME group, n = 13 for hepAGT‐/‐ HFD+L‐NAME group). B) Representative echocardiography images obtained from female mice at 15w after feeding with chow or HFD+L‐NAME (n = 6 for hepAGT+/+ CHOW group, n = 6 for hepAGT‐/‐ CHOW group, n = 15 for hepAGT+/+ HFD+L‐NAME group, n = 13 for hepAGT‐/‐ HFD+L‐NAME group). C) Left ventricular ejection fraction (LVEF%) was quantified via echocardiography (n = 6 for hepAGT+/+ CHOW group, n = 6 for hepAGT‐/‐ CHOW group, n = 15 for hepAGT+/+ HFD+L‐NAME group, n = 13 for hepAGT‐/‐ HFD+L‐NAME group). D) Left ventricular fraction shortening (LVFS%) was quantified via echocardiography (n = 6 for hepAGT+/+ CHOW group, n = 6 for hepAGT‐/‐ CHOW group, n = 15 for hepAGT+/+ HFD+L‐NAME group, n = 13 for hepAGT‐/‐ HFD+L‐NAME group). E) E/A ratio was quantified via echocardiography (n = 6 for hepAGT+/+ CHOW group, n = 6 for hepAGT‐/‐ CHOW group, n = 15 for hepAGT+/+ HFD+L‐NAME group, n = 13 for hepAGT‐/‐ HFD+L‐NAME group). F) E/e’ ratio was quantified via echocardiography (n = 6 for hepAGT+/+ CHOW group, n = 6 for hepAGT‐/‐ CHOW group, n = 15 for hepAGT+/+ HFD+L‐NAME group, n = 13 for hepAGT‐/‐ HFD+L‐NAME group). G) IVRT was quantified via echocardiography (n = 6 for hepAGT+/+ CHOW group, n = 6 for hepAGT‐/‐ CHOW group, n = 15 for hepAGT+/+ HFD+L‐NAME group, n = 13 for hepAGT‐/‐ HFD+L‐NAME group). H) Tei index was quantified via echocardiography (n = 6 for hepAGT+/+ CHOW group, n = 6 for hepAGT‐/‐ CHOW group, n = 15 for hepAGT+/+ HFD+L‐NAME group, n = 13 for hepAGT‐/‐ HFD+L‐NAME group). I) mRNA abundance of cardiac atrial natriuretic peptide (ANP) was assessed in female HFpEF hepAGT+/+ and female HFpEF hepAGT‐/‐ mice (n = 6 for hepAGT+/+ CHOW group, n = 6 for hepAGT‐/‐ CHOW group, n = 15 for hepAGT+/+ HFD+L‐NAME group, n = 13 for hepAGT‐/‐ HFD+L‐NAME group). J) mRNA abundance of cardiac brain natriuretic peptide (BNP) was assessed in female HFpEF hepAGT+/+ and female HFpEF hepAGT‐/‐ mice (n = 6 for hepAGT+/+ CHOW group, n = 6 for hepAGT‐/‐ CHOW group, n = 15 for hepAGT+/+ HFD+L‐NAME group, n = 13 for hepAGT‐/‐ HFD+L‐NAME group). Two‐way ANOVA was used for statistical analysis.

Figure 3

Inhibition of systemic AngII by losartan exhibited no effects on cardiac diastolic function in female HFpEF mice. A) Plasma AngII concentrations in female HFpEF hepAGT+/+ and female HFpEF hepAGT‐/‐ mice were measured by ELISA (n = 10 for hepAGT+/+ CHOW group, n = 10 for hepAGT‐/‐ CHOW group, n = 9 for hepAGT+/+ HFD+L‐NAME group, n = 10 for hepAGT‐/‐ HFD+L‐NAME group). B) Experimental workflow and analysis for the effects of losartan treatment on HFpEF in female mice (n = 9 for PBS CHOW group, n = 10 for Losartan CHOW group, n = 9 for PBS HFD+L‐NAME group, n = 7 for Losartan HFD+L‐NAME group). C) Representative echocardiography images obtained from female HFpEF mice at 15w after feeding with PBS or losartan (n = 9 for PBS CHOW group, n = 10 for Losartan CHOW group, n = 9 for PBS HFD+L‐NAME group, n = 7 for Losartan HFD+L‐NAME group). D) Left ventricular ejection fraction (LVEF%) was quantified via echocardiography (n = 9 for PBS CHOW group, n = 10 for Losartan CHOW group, n = 9 for PBS HFD+L‐NAME group, n = 7 for Losartan HFD+L‐NAME group). E) Left ventricular fraction shortening (LVFS%) was quantified via echocardiography (n = 9 for PBS CHOW group, n = 10 for Losartan CHOW group, n = 9 for PBS HFD+L‐NAME group, n = 7 for Losartan HFD+L‐NAME group). F) E/A ratio was quantified via echocardiography (n = 9 for PBS CHOW group, n = 10 for Losartan CHOW group, n = 9 for PBS HFD+L‐NAME group, n = 7 for Losartan HFD+L‐NAME group). G) E/e’ ratio was quantified via echocardiography (n = 9 for PBS CHOW group, n = 10 for Losartan CHOW group, n = 9 for PBS HFD+L‐NAME group, n = 7 for Losartan HFD+L‐NAME group). H) IVRT was quantified via echocardiography (n = 9 for PBS CHOW group, n = 10 for Losartan CHOW group, n = 9 for PBS HFD+L‐NAME group, n = 7 for Losartan HFD+L‐NAME group). I) Tei index was quantified via echocardiography (n = 9 for PBS CHOW group, n = 10 for Losartan CHOW group, n = 9 for PBS HFD+L‐NAME group, n = 7 for Losartan HFD+L‐NAME group). J) mRNA abundances of cardiac atrial natriuretic peptide (ANP) and cardiac brain natriuretic peptide (BNP) were assessed in female HFpEF mice feeding with PBS or losartan (n = 9 for PBS HFD+L‐NAME group, n = 7 for Losartan HFD+L‐NAME group). Two‐ way ANOVA was used for statistical analysis in A and D–I, and Student's t test was used for statistical analysis in J.

Figure 4

Deficiency of hepatic AGT promoted microvascular angiogenesis in HFpEF mice. A) Representative IB4 immune‐staining images of cardiac tissues obtained from female hepAGT+/+ and female hepAGT‐/‐ mice at 15w after feeding with Chow or HFD+L‐NAME, then IB4 density of cardiac tissues was calculated in female HFpEF hepAGT+/+ and female HFpEF hepAGT‐/‐ mice (n = 8 for each group, Bar = 100 µm). B) Representative IB4 immune‐staining images of cardiac tissues obtained from male hepAGT+/+ and male hepAGT‐/‐ mice at 15w after feeding with Chow or HFD+L‐NAME, then IB4 density of cardiac tissues was calculated in male HFpEF hepAGT+/+ and male HFpEF hepAGT‐/‐ mice (n = 8 for each group, Bar = 100 µm). C) Representative IB4 immune‐staining images of cardiac tissues obtained from female HFpEF mice at 15w after feeding with PBS or losartan, then IB4 density of cardiac tissues was calculated in female HFpEF mice fed with PBS or losartan (n = 6 for each group, Bar = 100 µm). D) Representative IB4 immune‐staining images of cardiac tissues obtained from male HFpEF mice at 15w after feeding with PBS or losartan, then IB4 density of cardiac tissues was calculated in male HFpEF mice fed with PBS or losartan (n = 6 for each group, Bar = 100 µm). E) Western blotting detection of AGT protein abundances in hepatocyte supernatants obtained from HFpEF hepAGT+/+ and HFpEF hepAGT‐/‐ mice (n = 3 for each group). F) Mouse cardiac endothelial cell lines (MCECs) were incubated with hepatocyte supernatants obtained from HFpEF hepAGT+/+ and HFpEF hepAGT‐/‐ mice. Subsequently, the representative tube formation images were acquired (n = 6 for each group). Tube lengths were finally calculated in hepAGT+/+ and hepAGT‐/‐ groups, respectively (n = 6 for each group, Bar = 1mm). Two‐ way ANOVA was used for statistical analysis in A–D, and Student's t test was used for statistical analysis in F.

Figure 5

Deficiency of hepatic AGT promoted cardiac microvascular angiogenesis by activating the GATA2/Pim3 pathway. A) Clustered heat map of genes between hearts from HFpEF hepAGT‐/‐ mice and HFpEF hepAGT+/+ mice by RNA‐seq analysis (n = 4 for each group). B) Funnel plot images of genes between hearts from HFpEF hepAGT‐/‐ mice and HFpEF hepAGT+/+ mice by RNA‐seq analysis (n = 4 for each group). C) Western blotting detection and quantification of GATA2 and Pim3 protein abundances in hearts obtained from female HFpEF hepAGT+/+ and female HFpEF hepAGT‐/‐ mice (n = 6 for each group). D) Representative tube formation images and calculation of tube lengths were obtained from MCECs transfected with either scrambled or Pim3 shRNA incubated with hepatocyte supernatants derived from hepAGT+/+ and hepAGT‐/‐ mice (n = 6 for each group, Bar = 250µm). E) Representative tube formation images and calculation of tube lengths were obtained from MCECs transfected with either scrambled or GATA2 shRNA incubated with hepatocyte supernatants derived from hepAGT+/+ or hepAGT‐/‐ mice (n = 6 for each group, Bar = 250µm). F) Western blotting detection of GATA2 and Pim3 protein abundances in MCECs transfected with either scrambled or Pim3 shRNA incubated with hepatocyte supernatants derived from hepAGT+/+ or hepAGT‐/‐ mice (n = 3 for each group). G) Western blotting detection of GATA2 and Pim3 protein abundances in MCECs transfected with either scrambled or GATA2 shRNA incubated with hepatocyte supernatants derived from hepAGT+/+ or hepAGT‐/‐ mice (n = 3 for each group). Student's t test was used for statistical analysis in C. One‐ way ANOVA was used for statistical analysis in D–G.

Figure 6

LRP2 was required for hepatocyte‐derived AGT internalization in cardiac endothelial cells. A) Western blotting detection and quantification of AGT protein abundances in hearts obtained from female HFpEF hepAGT+/+ and female HFpEF hepAGT‐/‐ mice (n = 6 for each group). B) MCECs were incubated with His‐Tag labelled rAGT protein or PBS, and the His‐Tag protein abundances of MCECs were analyzed by Western blotting (n = 6 for each group). C) MCECs transfected with either scrambled or LRP2 siRNA were incubated with His‐Tag labelled rAGT protein or PBS, and the His‐Tag and AGT protein abundances of MCECs were analyzed by Western blotting (n = 3 for each group). D) Representative tube formation images and calculation of tube lengths obtained from MCECs transfected with either scrambled or LRP2 siRNA incubated with hepatocyte supernatants derived from hepAGT+/+ mice (n = 6 for each group, Bar = 250µm). E) Experimental workflow and analysis for the effects of Tie2‐AAV‐LRP2‐shRNA myocardial injection on HFpEF in female mice (n = 7 for Tie2‐AAV‐NC‐shRNA, and n = 13 for Tie2‐AAV‐LRP2‐shRNA). F) Representative echocardiography images obtained from female HFpEF mice at 15w after injection with Tie2‐AAV‐LRP2‐shRNA (n = 7 for Tie2‐AAV‐NC‐shRNA, and n = 13 for Tie2‐AAV‐LRP2‐shRNA). G) Left ventricular ejection fraction (LVEF%) and left ventricular fraction shortening (LVFS%) were quantified via echocardiography (n = 7 for Tie2‐AAV‐NC‐shRNA, and n = 13 for Tie2‐AAV‐LRP2‐shRNA). H) E/A ratio was quantified via echocardiography (n = 7 for Tie2‐AAV‐NC‐shRNA, and n = 13 for Tie2‐AAV‐LRP2‐shRNA). I) E/e’ ratio was quantified via echocardiography (n = 7 for Tie2‐AAV‐NC‐shRNA, and n = 13 for Tie2‐AAV‐LRP2‐shRNA). J) IVRT was quantified via echocardiography (n = 7 for Tie2‐AAV‐NC‐shRNA, and n = 13 for Tie2‐AAV‐LRP2‐shRNA). K) Tei index was quantified via echocardiography (n = 7 for Tie2‐AAV‐NC‐shRNA, and n = 13 for Tie2‐AAV‐LRP2‐shRNA). L) Ratio of heart weight to tibia length (HW/TL) was measured in female HFpEF mice injected with Tie2‐AAV‐LRP2‐shRNA (n = 7 for Tie2‐AAV‐NC‐shRNA, and n = 13 for Tie2‐AAV‐LRP2‐shRNA). M) mRNA abundance of cardiac brain natriuretic peptide (BNP) was assessed in female HFpEF mice injected with Tie2‐AAV‐LRP2‐shRNA (n = 7 for Tie2‐AAV‐NC‐shRNA, and n = 11 for Tie2‐AAV‐LRP2‐shRNA). Student's t test was used for statistical analysis in B‐C and G‐M. Mann‐Whitney U test was used for statistical analysis in A. Two‐way ANOVA was used for statistical analysis in D.

Figure 7

18β‐glycyrrhetinic acid (18‐βGA) is a potent suppressor of hepAGT to treat HFpEF in female mice. A) Experimental workflow and analysis for the effects of 18β‐GA treatment on HFpEF in female mice (n = 14 for PBS group, and n = 13 for 18β‐GA group). B) Western blotting detection and quantification of AGT protein abundances in hearts obtained from female HFpEF mice treated with 18β‐GA (n = 12 for each group). C) Plasma AGT concentrations were measured in female HFpEF mice treated with 18β‐GA (n = 14 for the PBS group, and n = 13 for the 18β‐GA group). D) Representative echocardiography images obtained from female HFpEF mice at 15w after treating with 18β‐GA (n = 14 for PBS group, and n = 13 for 18β‐GA group). E) Left ventricular ejection fraction (LVEF%) was quantified via echocardiography (n = 14 for PBS group, and n = 13 for 18β‐GA group). F) Left ventricular fraction shortening (LVFS%) was quantified via echocardiography (n = 14 for PBS group, and n = 13 for 18β‐GA group). G) E/A ratio was quantified via echocardiography (n = 14 for PBS group, and n = 13 for 18β‐GA group). H) E/e’ ratio was quantified via echocardiography (n = 14 for PBS group, and n = 13 for 18β‐GA group). I) IVRT was quantified via echocardiography (n = 14 for PBS group, and n = 13 for 18β‐GA group). J) Tei index was quantified via echocardiography (n = 14 for PBS group, and n = 13 for 18β‐GA group). K) Representative IB4 immune‐staining images of cardiac tissues obtained from female HFpEF mice at 15w after treating with 18β‐GA (n = 8 for each group, Bar = 100 µm). IB4 density of cardiac tissues was then calculated in female HFpEF mice at 15w after treating with 18β‐GA (n = 8 for each group). Mann‐Whitney U test was used for statistical analysis in B. Student's t test was used for statistical analysis in C–K.

Figure 8

Potential mechanisms involved in hepatocyte‐derived angiotensinogen (hepAGT) contributing to HFpEF. Both interventions (hepatic AGT deletion vs systemic AngII blockade) showed similar outcomes in terms of blood pressure, myocardial hypertrophy, heart weight, cardiac fibrosis, and cardiac inflammation. However, they exhibited divergent effects on myocardial microvascular density and myocardial diastolic function in HFpEF models, irrespective of sex. These findings suggested that hepatic AGT may regulate HFpEF pathogenesis through an AngII‐independent mechanism centered on microvascular angiogenesis. Liver‐derived AGT was internalized by LRP2 in cardiac endothelial cells, subsequently contributing to myocardial diastolic dysfunction by suppressing microvascular angiogenesis via inhibiting the GATA2/Pim3 pathway. This cartoon illustration represents an original creation by our team.