Lymphatic Endothelial Branched-Chain Amino Acid Catabolic Defects Undermine Cardiac Lymphatic Integrity and Drive HFpEF

Abstract

Background: Heart failure with preserved ejection fraction (HFpEF) has become the most prevalent type of heart failure, but effective treatments are lacking. Cardiac lymphatics play a crucial role in maintaining heart health by draining fluids and immune cells. However, their involvement in HFpEF remains largely unexplored.

Methods: We examined cardiac lymphatic alterations in mice with HFpEF with comorbid obesity and hypertension, and in heart tissues from patients with HFpEF. Using genetically engineered mouse models and various cellular and molecular techniques, we investigated the role of cardiac lymphatics in HFpEF and the underlying mechanisms.

Results: In mice with HFpEF, cardiac lymphatics displayed substantial structural and functional anomalies, including decreased lymphatic endothelial cell (LEC) density, vessel fragmentation, reduced branch connections, and impaired capacity to drain fluids and immune cells. LEC numbers and marker expression levels were also decreased in heart tissues from patients with HFpEF. Stimulating lymphangiogenesis with an adeno-associated virus expressing an engineered variant of vascular endothelial growth factor C (VEGFC^C156S) that selectively activates vascular endothelial growth factor receptor 3 (VEGFR3) in LECs restored cardiac lymphatic integrity and substantially alleviated HFpEF. Through discovery-driven approaches, defective branched-chain amino acid (BCAA) catabolism was identified as a predominant metabolic signature in HFpEF cardiac LECs. Overexpression of branched-chain ketoacid dehydrogenase kinase (encoded by the Bckdk gene), which inactivates branched-chain ketoacid dehydrogenase (the rate-limiting enzyme in BCAA catabolism), resulted in spontaneous lymphangiogenic defects in LECs. In mice, inducible Bckdk gene deletion in LECs to enhance their BCAA catabolism preserved cardiac lymphatic integrity and protected against HFpEF. BCAA catabolic defects caused ligand-independent phosphorylation of VEGFR3 in the cytoplasm by Src kinase, leading to lysosomal degradation of VEGFR3 instead of its trafficking to the cell membrane. Reduced VEGFR3 availability on the cell surface impeded downstream Akt (protein kinase B) activation, hindered glucose uptake and utilization, and inhibited lymphangiogenesis in LECs with BCAA catabolic defects.

Conclusions: Our study provides evidence that cardiac lymphatic disruption, driven by impaired BCAA catabolism in LECs, is a key factor contributing to HFpEF. These findings unravel the crucial role of BCAA catabolism in modulating lymphatic biology, and suggest that preserving cardiac lymphatic integrity may present a novel therapeutic strategy for HFpEF.

Keywords: amino acids, branched-chain; endothelial cells; heart failure; lymphangiogenesis; vascular endothelial growth factor receptor-3.

Figure 1.

Structural disruption and functional impairment of cardiac lymphatics in HFpEF. A, Whole heart clearing and whole-mount immunofluorescence staining for LYVE1 (lymphatic endothelial hyaluronic acid receptor 1) were performed to visualize the lymphatic vasculature in hearts obtained from control mice or mice challenged with a high-fat (HF) diet plus treatment with N^ω-nitro-L-arginine methyl ester (L-NAME) in drinking water for 6 or 12 weeks. Images were captured using a light sheet fluorescence microscope and the cardiac lymphatic volume was quantified (top panel). The 3-dimensional architecture of the cardiac lymphatic vasculature was reconstructed and displayed. Lymphatic segments (middle panel) and connecting branch points (bottom) were also quantified. B, Assessment of the capacity of cardiac lymphatics to drain immune cells in both control and heart failure with preserved ejection fraction (HFpEF) mice (HF+L-NAME 12 weeks) was performed. Splenocytes from donor mice were isolated and labeled with the live cell tracker CM-DiI. Equal numbers of labeled splenocytes were injected into myocardial apex of both control and HFpEF recipient mice. The tracheobronchial lymph node adjacent to the heart was harvested at 0, 12, and 24 hours, and quantification of CM-DiI–positive cells that had drained into the lymph node was performed. C, Assessment of the capacity of cardiac lymphatics to drain tissue fluid was performed. Equal volumes of Evans blue dye (EBD) were injected into the myocardial apex of controls and mice with HFpEF. Representative images show drainage of the dye by cardiac lymphatics 5 minutes after injection (top left) and retention of the dye in myocardial apex regions after 6 hours (top right). The remaining level of EBD was quantified by measuring the absorbance at 610 nm in the whole heart homogenates (bottom). D, Cardiac water content was evaluated using T2-weighted magnetic resonance imaging in the controls and mice with HFpEF. E, Cardiac edema was assessed by the wet-to-dry heart weight ratio. F, Cardiac lymphatic density was assessed by immunofluorescent staining for LYVE1 (green) and vascular endothelial growth factor receptor 3 (VEGFR3; red) in controls and mice with HFpEF. Tube-like structures demonstrating dual immunofluorescence positivity for LYVE1 and VEGFR3 were identified as lymphatic vessels and quantified. G, Protein levels of lymphatic endothelial cell markers LYVE1, PDPN (podoplanin), and PROX1 (prospero-related homeobox 1) in left ventricular tissues were analyzed by Western blot. β-Tubulin expression was set as loading control. H, mRNA levels of Lyve1, Pdpn, Prox1, Vegfr3, and Reln relative to 18S in left ventricular tissues were analyzed by real-time polymerase chain reaction. I, LYVE1 immunohistochemical staining was performed on heart tissue samples from patients with HFpEF and controls without HFpEF, followed by quantification of tube-like LYVE1⁺ lymphatics. J, Western blot analysis of cardiac tissue samples from patients with HFpEF and controls without HFpEF, specifically examining lymphatic endothelial cell markers LYVE1, PDPN, and PROX1, was performed. β-Tubulin protein expression was used as loading control. K, Real-time polymerase chain reaction was used to evaluate mRNA expression levels of lymphatic endothelial cell markers LYVE1, PDPN, VEGFR3, and PROX1, and the lymphoangiocrine factor RELN, in heart tissues from patients with HFpEF and controls without HFpEF. Data are presented as mean±SD. Each point represents an individual animal (A through H) or an individual participant (I through K). For statistical analysis, 1-way ANOVA followed by the Tukey multiple comparison test was used in A and 2-tailed Student t test was used in B through K.

Figure 2.

Stimulating lymphangiogenesis alleviates cardiac lymphatic anomalies and improves HFpEF. A, Mice were injected intramyocardially with adeno-associated virus serotype 9 (AAV9) carrying a variant of vascular endothelial growth factor C (VEGFC) specifically targeting vascular endothelial growth factor receptor 3 (VEGFR3; AAV9-VEGFC^C156S) or an empty vector control (AAV9-Vector). Next, they were randomly divided into groups treated with or without high-fat (HF) diet plus N^ω-nitro-L-arginine methyl ester (L-NAME) for 12 weeks. Whole-mount immunohistochemical staining for LYVE1 (lymphatic endothelial hyaluronic acid receptor 1) was performed to assess cardiac lymphatic structure, and lymphatic density was quantified. B, Early to late diastolic transmitral flow velocity ratio (E/A) and early transmitral flow velocity to mitral annular early diastolic velocity ratio (E/e′) values were measured using echocardiography to assess cardiac diastolic function. C, Left ventricular end-diastolic pressure (LVEDP) was measured using hemodynamic examination. D, Treadmill running distance was measured to assess exercise tolerance. E, Representative images of lung hematoxylin & eosin (HE) staining are shown. F, Pulmonary congestion was assessed by the ratio of wet-to-dry lung weight. G, Assessment of cardiac hypertrophy was performed using wheat germ agglutinin (WGA) staining. Representative images (left) and cardiomyocyte cross-sectional area (CSA) quantification (right) are shown. H, Assessment of interstitial fibrosis was performed using Masson trichrome staining. Representative images (left) and fibrotic area quantification (right) are shown. I, Assessment of cardiac inflammatory infiltration was performed using immunohistochemical staining for CD45. Representative images (left) and CD45-positive cell quantification (right) are shown. Data are presented as mean±SD. Each point represents an individual animal. Two-way ANOVA followed by the Sidak multiple comparison test was used for statistical analysis. HFpEF indicates heart failure with preserved ejection fraction.

Figure 3.

Impaired BCAA catabolism is a prominent metabolic signature in HFpEF cardiac LECs, resulting in lymphangiogenic defects. A, PDPN (podoplanin)–positive lymphatic endothelial cells (LECs) were isolated from the hearts of control mice and mice with heart failure with preserved ejection fraction (HFpEF) induced by a 12-week high-fat (HF) diet plus N^ω-nitro-L-arginine methyl ester (L-NAME) challenge using magnetic-activated cell sorting and RNA sequencing (RNA-seq) analysis was performed. Primary LECs in culture were treated with either vehicle or the vascular endothelial growth factor receptor 3 (VEGFR3) inhibitor MAZ51 to induce lymphangiogenic defects, followed by RNA-seq analysis. Among the downregulated pathways, the branched-chain amino acid (BCAA) catabolic pathway (valine, leucine, and isoleucine degradation) was found to be one of the most significant. B, A schematic diagram of the BCAA catabolic pathway. The rate-limiting enzyme in BCAA catabolism, branched-chain ketoacid dehydrogenase (BCKDH), can be inactivated through phosphorylation modification by branched-chain ketoacid dehydrogenase kinase (BCKDK), and can be activated by dephosphorylation modification by PP2Cm. C, Cardiac LECs were isolated from controls and mice with HFpEF through magnetic-activated cell sorting and mRNA levels of BCAA catabolic genes were measured. 18S expression was used as loading control. D, Protein levels of pBCKDHA^S293 and BCKDHA in cardiac LECs isolated from both controls and mice with HFpEF. HSP90 (heat shock protein 90) expression was used as loading control. E, Total intracellular BCAA levels in cardiac LECs isolated from controls and mice with HFpEF. F, Human dermal LECs were treated with vehicle or MAZ51 and mRNA levels of BCAA catabolic genes were measured. G, Human dermal LECs were transfected with adenovirus overexpressing BCKDK (Ad-BCKDK) to induce BCAA catabolic defects. Cell proliferation was measured using a CCK-8 assay. H, Cell migration was assessed using a wound-healing assay. I, Tube formation in Ad-Control and Ad-BCKDK groups. J, Sprouting in Ad-Control and Ad-BCKDK groups. Data are presented as mean±SD. In C through E, each point represents an individual animal. In F through J, each point represents an independent biologic repeat. A 2-tailed Student t test was used for statistical analysis.

Figure 4.

Genetically enhancing BCAA catabolism in LECs preserves cardiac lymphatic integrity and protects against HFpEF. A, Age-matched male Bckdk^fl/fl mice and inducible lymphatic endothelial cell (LEC)–specific Bckdk gene deletion mice (Bckdk^cKO) were randomly divided into groups treated with or without a 12-week high-fat (HF) diet plus N^ω-nitro-L-arginine methyl ester (L-NAME) challenge. Whole-mount immunohistochemical staining for LYVE1 (lymphatic endothelial hyaluronic acid receptor 1) was performed to assess cardiac lymphatic structure and density. B, Echocardiographic measurement of early to late diastolic transmitral flow velocity ratio (E/A) and early transmitral flow velocity to mitral annular early diastolic velocity ratio (E/e′) values to assess cardiac diastolic function. C, Hemodynamic measurement of left ventricular end-diastolic pressure (LVEDP) values. D, Treadmill measurement of running distance to evaluate exercise tolerance. E, Representative images of lung hematoxylin & eosin (HE) staining. F, Ratios of wet-to-dry lung weight to evaluate pulmonary congestion. G, Representative images of wheat germ agglutinin (WGA) staining (left) and quantification of the cardiomyocyte cross-sectional area (CSA; right). H, Representative images of Masson trichrome staining (left panel) and quantification of the fibrotic area (right). I, Representative images of CD45 immunohistochemical staining (left) and quantification of CD45-positive cells (right). Data are presented as mean±SD. Each point represents an individual animal. Two-way ANOVA followed by the Sidak multiple comparison test was used for statistical analysis. BCAA indicates branched-chain amino acid; and HFpEF, heart failure with preserved ejection fraction.

Figure 5.

BCAA catabolic defects in LECs hinder lymphangiogenesis by disrupting glucose uptake and utilization. A, Lymphatic endothelial cells (LECs) were transfected with an adenovirus overexpressing BCKDK (Ad-BCKDK) to induce defects in branched-chain amino acid (BCAA) catabolism. Differential pathways between LECs transfected with Ad-Control and those transfected with Ad-BCKDK were identified using RNA sequencing (RNA-seq) and visualized in a volcano plot. B, A heatmap was generated to display the expression of genes involved in glycolysis pathways. C, Whole protein samples and cell membrane protein samples were extracted from human dermal LECs for Western blot analysis. GLUT1 (glucose transporter-1) expression levels were determined, with HSP90 (heat shock protein 90) protein expression serving as the loading control for the whole protein sample and ATP1A1 (sodium/potassium-transporting ATPase subunit α1) protein expression serving as the loading control for the cell membrane protein sample. D, GLUT1 protein abundance at the cell surface was determined by immunofluorescent staining. E, Glycolytic flux was measured by the extracellular acidification rate (ECAR). Basal glycolysis, glycolytic capacity, and glycolytic reserve were calculated. F, Human dermal LECs transfected with Ad-Control and Ad-BCKDK were treated with adenovirus overexpressing GLUT1 (Ad-GLUT1) or left untreated. LEC proliferation in each group was assessed using a CCK-8 assay. G, LEC migration in each group was assessed using a wound-healing assay. H, Tube formation of LECs in each group. I, Sprouting of LECs in each group. Data are presented as mean±SD. Each point represents an independent biologic repeat. C through D were analyzed by 2-tailed Student t test. E through I were analyzed by 2-way ANOVA followed by the Sidak multiple comparison test.

Figure 6.

Restoring Akt activity corrects disrupted glucose metabolism and improves lymphangiogenesis in LECs with BCAA catabolic defects. A, Human dermal lymphatic endothelial cells (LECs) transfected with Ad-Control and Ad-BCKDK were treated with vascular endothelial growth factor C (VEGFC). Downstream Akt (protein kinase B) phosphorylation was assessed by Western blot at 0, 10, and 15 minutes after VEGFC treatment. B, Human dermal LECs were transfected with Ad-Control and Ad-BCKDK and treated with adenovirus overexpressing a constitutively active AKT isoform 1 (Ad-myr-AKT1) or left untreated. Glucose uptake levels in each group of cells were evaluated using a fluorescence-labeled 2-deoxy-glucose analog (2-NBDG). C, Glycolytic flux was assessed by the extracellular acidification rate (ECAR). Basal glycolysis, glycolytic capacity, and glycolytic reserve were calculated. D, LEC proliferation in each group was assessed using a CCK-8 assay. E, LEC migration in each group was assessed using a wound-healing assay. F, Tube formation of LECs in each group. G, Sprouting of LECs in each group. Data are presented as mean±SD. Each point represents an independent biologic repeat. A was analyzed by 2-tailed Student t test. B through G were analyzed by 2-way ANOVA followed by the Sidak multiple comparison test. BCAA indicates branched-chain amino acid.

Figure 7.

Src-mediated ligand-independent VEGFR3 phosphorylation impedes its cell membrane trafficking, disrupts glucose metabolism, and inhibits lymphangiogenesis in LECs with BCAA catabolic defects. A, VEGFR3 mRNA levels were assessed in human dermal lymphatic endothelial cells (LECs) transfected with either Ad-Control or Ad-BCKDK. B, Protein expression of vascular endothelial growth factor receptor 3 (VEGFR3) in LECs transfected with Ad-Control or Ad-BCKDK was examined. HSP90 (heat shock protein 90) protein expression was used as a loading control. C, Cell surface availability of VEGFR3 was evaluated through immunofluorescence staining, with ATP1A1 (sodium/potassium-transporting ATPase subunit α1) serving as a marker for the cell membrane. D, After serum deprivation to remove endogenous ligands, the ligand-independent phosphorylation levels of VEGFR3 in LECs transfected with Ad-Control and Ad-BCKDK were examined using immunoprecipitation. Both short and long exposure immunoblots are presented. E, Lysosomal trafficking of VEGFR3 was evaluated through immunofluorescence staining, with LAMP1 (lysosomal-associated membrane protein 1) serving as a marker for lysosomes. F, Phosphorylation levels at Y419 and the total protein expression of Src kinase were assessed in LECs transfected with Ad-Control and Ad-BCKDK. HSP90 protein expression was used as loading control. G, Human dermal LECs were transfected with Ad-Control and Ad-BCKDK and treated with vehicle or Src kinase inhibitor PP2. The ligand-independent phosphorylation levels of VEGFR3 were assessed using immunoprecipitation. Representative immunoblots under short and long exposure are shown. H, Lysosomal protein samples were extracted from human dermal LECs for Western blot analysis. VEGFR3 expression levels were determined, with LAMP1 protein expression serving as the loading control. I, Human dermal LECs were transfected with Ad-Control or Ad-BCKDK, followed by treatment with either vehicle or the Src kinase inhibitor PP2. Expression levels of VEGFR3, Akt (protein kinase B) phosphorylation, and total Akt protein were detected by Western blot analysis. J, Graphic illustration. Impaired branched-chain amino acid (BCAA) catabolism is a prominent metabolic signature in cardiac LECs under heart failure with preserved ejection fraction (HFpEF) risk factors. This metabolic alteration triggers ligand-independent phosphorylation of VEGFR3 by Src kinase, resulting in lysosomal degradation of VEGFR3 instead of its trafficking to cell membrane. Reduced VEGFR3 availability on cell surface hinders downstream Akt activation and glucose metabolism, thereby suppressing lymphangiogenesis. Cardiac lymphatic disruption, driven by defective BCAA catabolism in LECs, promotes the development of HFpEF. Data are presented as mean±SD. Each point represents an independent biologic repeat. A through F were analyzed by 2-tailed Student t test. G through I were analyzed by 2-way ANOVA followed by the Sidak multiple comparison test.