Regulation and impact of cardiac lymphangiogenesis in pressure-overload-induced heart failure

Abstract

Aims: Lymphatics are essential for cardiac health, and insufficient lymphatic expansion (lymphangiogenesis) contributes to development of heart failure (HF) after myocardial infarction. However, the regulation and impact of lymphangiogenesis in non-ischaemic cardiomyopathy following pressure-overload remains to be determined. Here, we investigated cardiac lymphangiogenesis following transversal aortic constriction (TAC) in C57Bl/6 and Balb/c mice, and in end-stage HF patients.

Methods and results: Cardiac function was evaluated by echocardiography, and cardiac hypertrophy, lymphatics, inflammation, oedema, and fibrosis by immunohistochemistry, flow cytometry, microgravimetry, and gene expression analysis. Treatment with neutralizing anti-VEGFR3 antibodies was applied to inhibit cardiac lymphangiogenesis in mice. We found that VEGFR3-signalling was essential to prevent cardiac lymphatic rarefaction after TAC in C57Bl/6 mice. While anti-VEGFR3-induced lymphatic rarefaction did not significantly aggravate myocardial oedema post-TAC, cardiac immune cell levels were increased, notably myeloid cells at 3 weeks and T lymphocytes at 8 weeks. Moreover, whereas inhibition of lymphangiogenesis did not aggravate interstitial fibrosis, it increased perivascular fibrosis and accelerated development of left ventricular (LV) dilation and dysfunction. In clinical HF samples, cardiac lymphatic density tended to increase, although lymphatic sizes decreased, notably in patients with dilated cardiomyopathy. Similarly, comparing C57Bl/6 and Balb/c mice, lymphatic remodelling post-TAC was linked to LV dilation rather than to hypertrophy. The striking lymphangiogenesis in Balb/c was associated with reduced cardiac levels of macrophages, B cells, and perivascular fibrosis at 8 weeks post-TAC, as compared with C57Bl/6 mice that displayed weak lymphangiogenesis. Surprisingly, however, it did not suffice to resolve myocardial oedema, nor prevent HF development.

Conclusions: We demonstrate for the first time that endogenous lymphangiogenesis limits TAC-induced cardiac inflammation and perivascular fibrosis, delaying HF development in C57Bl/6 but not in Balb/c mice. While the functional impact of lymphatic remodelling remains to be determined in HF patients, our findings suggest that under settings of pressure-overload poor cardiac lymphangiogenesis may accelerate HF development.

Keywords: CCL21; Hypertrophy; Inflammation; Vegfc; Vegfd; Wall stress; mF4-31C1.

Graphical Abstract

Pressure-overload in mice leads to strain-dependent cardiac responses, with the development of either compensated or decompensated hypertrophy, accompanied by left ventricular (LV) dilation causing increased ventricular wall stress (WS). The compensated hypertrophic phenotype post-transversal aortic constriction (post-TAC), seen in female C57Bl/6 mice, was associated with transient and weak cardiac lymphangiogenesis, accumulation of immune cells and perivascular fibrosis. In contrast, both male C57Bl/6 and female Balb/c mice developed a dilated phenotype post-TAC, characterized by increased cardiac gene expression of ANP and IL1β, and stimulation of cardiac lymphangiogenesis. Inhibition of lymphangiogenesis during pressure-overload uncovered lymphatic rarefaction in the heart, with increased cardiac pro-inflammatory immune cell levels and aggravation of perivascular fibrosis, leading to accelerated development of cardiac remodelling and decompensation.

Figure 1

Evaluation of cardiac hypertrophy and remodelling at 8 weeks post-TAC. Examples (A, scale bar 1 mm) and morphometric assessment of LV weights normalized to tibia lengths (B) at 8 weeks in male C57Bl/6 mice: sham (open circles, n = 13) or TAC (closed circles, n = 15); female C57Bl/6 mice: sham (open triangles, n = 14) or TAC (closed triangles, n = 10); and in Balb/c female mice: sham (open square, n = 14) or TAC (closed square, n = 8). Analysis of cardiomyocyte cross-sectional area (C, n = 3–8 animals per group) and calculation of LV hypertrophy dilatation index (D) at 8 weeks. Cardiac expression analyses at 8 weeks of Nppa and Nppb (E, n = 5–10 per group). Examples (F) and evaluation at 8 weeks post-TAC of blood vessel to cardiomyocyte ratios (G, n = 3–8 animals per group). WGA, white, CD31, purple, Dapi, blue (scale bar 20 µm). Groups were compared pair-wise by two-way ANOVA followed by Sidak’s multiple comparison test (for morphometry and cardiac expression analyses) or by non-parametric Kruskal Wallis followed by Dunn’s post hoc test (for immunohistochemistry) *P < 0.05, **P < 0.01, ***P < 0.001 vs. sham.

Figure 2

Lymphangiogenesis is necessary to prevent cardiac lymphatic rarefaction post-TAC. Cardiac expression analyses at 8 weeks of Vegfc and Vegfd (A), and evaluation at 3 and 8 weeks of lymphatic vessel density (B) and open lymphatic density (C) in the LV subepicardium in male C57Bl/6 sham (open circles, n = 6), TAC controls (closed circles, n = 6–8), and anti-VEGFR3-treated TAC (closed circles, n = 6–7). Cardiac expression analyses at 8 weeks of lymphatic markers Pdpn, Ccl21, and Flt4 (D, n = 6–10 animals per group). Light sheet imaging of cardiac lymphatics (E) at 3 weeks post-TAC visualized by Lyve1 staining (scale bar 300 µm). Examples of 3D modelling of cardiac lymphatics visualized by confocal analyses (Lyve1) (F, scale bar 100 µm). Quantification of cardiac lymphatic volume density (G), frequency of short lymphatic branches (H), and volume:length ratio (I) (n = 3 mice per group). Groups were compared by non-parametric Kruskal Wallis followed by Dunn’s post hoc test (for immunohistochemistry) and by two-way ANOVA followed by Dunnett’s multiple comparisons test (for expression analyses). *P < 0.05 vs. sham, ##P < 0.01, ###P < 0.001 vs. TAC control.

Figure 3

Aggravation of cardiac hypertrophy and inflammation, but not oedema, by anti-VEGFR3 treatment. Morphometric assessment of LV weight normalized to body weight (A) at 3 or 8 weeks in male C57Bl/6 sham (open circles, n = 7–13), TAC controls (closed circles, n = 8–20); and anti-VEGFR3-treated TAC (closed circles, n = 12–20). Analysis of cardiac Nppa and Nppb (B) expression at 8 weeks (n = 6–10 per group). Assessment of cardiac (C) and pulmonary (D) water content at 3 weeks post-TAC in male C57Bl/6 (n = 6–9 per group). Flow cytometric evaluation (E) at 3 weeks post-TAC (n = 4–8 per group) of cardiac-infiltrating CD45⁺ immune cells (F), CD11b⁺ CD206⁺ monocytes (G) and CD11b⁺ SSC^high granulocytes (H). Data are reported as cells per mg cardiac tissue. Quantification by immunohistochemistry at 3 weeks post-TAC (n = 6–7 animals per group) of total cardiac-infiltrating CD68⁺ macrophages (I), and classical iNOS⁺ macrophages (J). Quantification by immunohistochemistry at 8 weeks post-TAC (n = 7–8 animals per group) of total cardiac-infiltrating CD3⁺ T cells (k), CD4⁺ T cells (L), and B220⁺ B cells (M). Examples of cardiac CD3⁺ total T cells (green) and CD4⁺ T cell subpopulation (grey) indicated by arrows (N). Scalebar 50 µm. Groups were compared by non-parametric Kruskal Wallis followed by Dunn’s post hoc test (for immunohistochemistry, microgravimetry, and flow cytometry) and by two-way ANOVA followed by Sidak’s multiple comparison test (for morphometric data) or Dunnett’s multiple comparisons test (for expression analyses). *P < 0.05, **P < 0.01, ***P < 0.001 vs. sham, ##P < 0.01, ###P < 0.001 vs. control TAC.

Figure 4

Evaluation of strain-dependent cardiac lymphatic remodelling in mice post-TAC and in HF patients. Cardiac expression analyses (n = 5–10 per group) at 8 weeks post-TAC of Vegfc and Vegfd (A) in female C57Bl/6 sham (open triangles) or TAC (closed triangles); and in female Balb/c sham (open square, n = 9) or TAC (closed square). Evaluation of total lymphatic density (B) and open lymphatic density (C) in the LV subepicardium at 3 or 8 weeks (n = 7–10 per group). Expression analyses of lymphatic markers Pdpn, Ccl21, and Flt4 (D, n = 5–10 per group). Examples by light sheet and confocal imaging of cardiac lymphatics (E) at 8 weeks post-TAC. Lyve1 (scale bar upper row 300 µm, lower row zoomed confocal views 100 µm). Quantification of perivascular lymphatic density (F), lymphatic lumen sizes (G), and examples of cardiac lymphatic density (H) in healthy cardiac donors vs. end-stage HF patients with ischaemic or DCM. Podoplanin (scale bar 50 µm). Groups were compared pair-wise by non-parametric Kruskal Wallis followed by Dunn’s post hoc test (for immunohistochemistry) or Sidak’s multiple comparisons test (for cardiac expression analyses) *P < 0.05, **P < 0.01, ***P < 0.001 vs. healthy/sham.

Figure 5

Strain-dependent effects on cardiac inflammation, but not oedema, during pressure-overload. Cardiac expression analysis of Il1b and Il6 (A) at 8 weeks post-TAC (n = 5–10 per group) in female C57Bl/6 sham (open triangles) or TAC (closed triangles), and in female Balb/c sham (open square) or TAC (closed square). Assessment of cardiac water content (B), and pulmonary weights (C) at 8 weeks (n = 8–10 per group). Immunohistochemical analyses, at 3 weeks, of cardiac CD206⁺ macrophages (D), classical iNOS⁺ macrophages (E), and B220⁺ B cells (G) (n = 7–8 per group). Examples of cardiac macrophages (F). CD68, green, CD206, red; dapi blue. Arrows indicate CD206⁺ macrophages (scalebar, 50 µm). Analysis of cardiac Lyve-1 expression, normalized to Pdpn levels, at 3- or 8-weeks post-TAC (H). Examples of flow cytometry gating in TAC mice (I). Quantification in C57Bl/6 and Balb/c females at 8 weeks post-TAC (n = 4–5 samples per group) of cardiac-infiltrating CD45⁺ immune cells (J), CD19⁺ B cells vs CD3⁺ T cells (K), and CD4⁺ vs. CD8⁺ T cell subpopulations (L). Data is reported as cells per mg cardiac tissue. Groups were compared pair-wise by two-way ANOVA followed by Sidak’s multiple comparison tests (morphometry data and expression analyses) or non-parametric Kruskal Wallis followed Dunn’s post hoc test (for microgravimetry and immunohistochemistry) or by Mann–Whitney U test (for flow cytometry); *P < 0.05, **P < 0.01, ***P < 0.001 vs. sham.

Figure 6

Cardiac interstitial and perivascular fibrosis, and perivascular lymphangiogenesis post-TAC. Cardiac gene expression of Col1a1 and Col3a1 (A) at 8 weeks post-TAC in male C57Bl/6 sham (open circles), TAC controls (closed circles); and anti-VEGFR3-treated TAC (closed circles) (n = 6–10 per group). Quantification of interstitial collagen III density (B) at 3 and 8 weeks in male C57Bl/6 mice (n = 5–9 per group). Quantification in male C57Bl/6 (C) and examples (D) of perivascular fibrosis at 8 weeks post-TAC, evaluated as relative fibrotic area surrounding arterioles in the size range of 5–50 µm diameter (scalebar: 50 µm). Quantification of interstitial (E) and perivascular (F) fibrosis at 8 weeks in female C57Bl/6 sham (open triangles, n = 7) or TAC (closed triangles, n = 10) and in female Balb/c sham (open square, n = 7) or TAC (closed square, n = 8). Quantification of perivascular lymphatic density in female C57Bl/6 and Balb/c mice at 8 weeks (G, n = 6–8 per group). Groups were compared by non-parametric Kruskal Wallis followed by Dunn’s post hoc test (for histology) and by two-way ANOVA followed by Dunnett’s multiple comparisons test (for expression analyses). *P < 0.05, **P < 0.01, ***P < 0.001 vs. sham; ##P < 0.01, ###P < 0.001 vs. control TAC.