Lymphangiogenesis requires Ang2/Tie/PI3K signaling for VEGFR3 cell-surface expression

Abstract

Vascular endothelial growth factor C (VEGF-C) induces lymphangiogenesis via VEGF receptor 3 (VEGFR3), which is encoded by the most frequently mutated gene in human primary lymphedema. Angiopoietins (Angs) and their Tie receptors regulate lymphatic vessel development, and mutations of the ANGPT2 gene were recently found in human primary lymphedema. However, the mechanistic basis of Ang2 activity in lymphangiogenesis is not fully understood. Here, we used gene deletion, blocking Abs, transgene induction, and gene transfer to study how Ang2, its Tie2 receptor, and Tie1 regulate lymphatic vessels. We discovered that VEGF-C-induced Ang2 secretion from lymphatic endothelial cells (LECs) was involved in full Akt activation downstream of phosphoinositide 3 kinase (PI3K). Neonatal deletion of genes encoding the Tie receptors or Ang2 in LECs, or administration of an Ang2-blocking Ab decreased VEGFR3 presentation on LECs and inhibited lymphangiogenesis. A similar effect was observed in LECs upon deletion of the PI3K catalytic p110α subunit or with small-molecule inhibition of a constitutively active PI3K located downstream of Ang2. Deletion of Tie receptors or blockade of Ang2 decreased VEGF-C-induced lymphangiogenesis also in adult mice. Our results reveal an important crosstalk between the VEGF-C and Ang signaling pathways and suggest new avenues for therapeutic manipulation of lymphangiogenesis by targeting Ang2/Tie/PI3K signaling.

Keywords: Cardiovascular disease; Endothelial cells; Growth factors; Vascular Biology.

Figure 1. Postnatal Tie1, Ang2, or Tie1/-2 deficiency results in an abnormal cutaneous lymphatic capillary network.

(A) Representative images of VE-cadherin, Prox1, Tie1, and Tie2 immunostaining in ear skin from P21 pups. Dashed lines indicate lymphatic capillaries. Scale bar: 50 μm. (B) LYVE1 staining of ventral ear skin from control, Tie1-deleted (control: n = 10; Tie1-deleted: n = 7), Tie2-deleted (control: n = 12; Tie2-deleted n = 7), Tie1/-2–deleted (control: n = 15; Tie1/-2–deleted: n = 11), lgG-treated (n = 5), and Ang2 Ab–treated (n = 5) pups on P21. Scale bar: 1 mm. (C) Magnification of the area outlined by the dashed boxes in B. Scale bar: 200 μm. (D) Quantification of the LYVE1-positive vessel area, vessel diameter, branch points, and vessel overlaps. Data represent the mean ± SEM. *P < 0.05 and ***P < 0.001, by 2-tailed Student’s t test.

Figure 2. Double-deletion of Tie1 and Tie2 leads to reduced SMC coverage and a lack of valves in collecting lymphatic vessels.

(A and B) Representative images of P21 ear skin immunostained for VE-cadherin, Prox1, and Tie1 or Tie2. Dashed lines indicate the collecting lymphatic vessels. Scale bars: 50 μm. (C) Dorsal ear skin of Tie1-deleted (control: n = 4; Tie1-deleted: n = 3), Tie2-deleted (control: n = 3; Tie2-deleted n = 3), Tie1/-2–deleted (control: n = 4; Tie1/-2–deleted n = 3), and Ang2 Ab–treated (IgG: n = 4; Ang2 ab: n = 3) P21 pups, immunostained for podoplanin, αSMA, and integrin α9. Scale bar: 50 μm. (D) Quantification of SMC coverage per collecting lymphatic vessel region, the number of valves, the widest and narrowest vessel diameters measured from entire collecting lymphatic vessels, and the differences between the two. R1 and R2 indicate the proximal and distal lymphatic collecting vessel regions, respectively. Data represent the mean ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001, by 2-tailed Student’s t test.

Figure 3. Requirement of Tie1 and Tie2 for Ang2-induced lymphatic vessel enlargement and Ang2 for Tie1 expression on the surface of LECs.

(A) Images of LYVE1 staining of ventral ear skin in control (n = 5), Tie1-deleted (Tie1^iΔLEC, n = 3), Ang2-overexpressing (Ang2^EC, n = 2), and Tie1^iΔLEC Ang2^EC (n = 2) p21 pups. Scale bar: 500 μm. Graph shows quantification of the average lymphatic capillary diameter. (B) Images of podoplanin staining of collecting lymphatic vessels in dorsal ear skin from the pups indicated in A on P21. Scale bar: 50 μm. Graph shows quantification of the widest collecting lymphatic vessel diameter. (C) Images of podoplanin staining of collecting lymphatic vessels in dorsal ear skin in control (n = 7), Tie2-deleted (Tie2^iΔLEC, n = 3), Ang2-overexpressing (Ang2^EC, n = 3), and Tie2^iΔLEC Ang2^EC (n = 2) pups on P21. Scale bar: 50 μm. Graph shows quantification of the widest collecting lymphatic vessel diameter. (D) Images of Tie1 staining and quantification in lymphatic capillaries and collecting vessels in ear skin from IgG- (n = 4) and Ang2 Ab–treated (n = 4) pups on P21, normalized to control. Scale bars: 50 μm. Magnification: 1.87. Data represent the mean ± SEM. **P < 0.01 and ***P < 0.001, by 1-way ANOVA with Bonferroni’s post hoc test for multiple comparisons (A–C) and 2-tailed Student’s t test (D).

Figure 4. Lack of Ang2 or Tie receptors leads to loss of VEGFR3 from the LEC surface.

(A) Images of lymphatic capillaries in ear skin from control (n = 8), Tie1-deleted (n = 6), IgG-treated (n = 4), and Ang2 Ab–treated (n = 4) P21 pups, immunostained for VEGFR3 and GOLPH4. Arrows point to VEGFR3 in the Golgi complex. Quantifications show a decrease in VEGFR3 immunofluorescence in capillary LECs and an increase in Golgi complex VEGFR3 immunofluorescence compared with total VEGFR3, normalized to control. Scale bars: 10 μm. (B) Cell-surface immunostaining for VEGFR3 in lymphatic capillaries of the ear skin from Tie1-deleted (control: n = 3; Tie1-deleted: n = 7) and Ang2-inhibited (n = 3 per group) P21 pups. Graphs show quantification of VEGFR3 immunofluorescence at the cell surface, normalized to control. Scale bars: 20 μm. Magnification: 0.40. (C) Immunostaining of VEGFR3 and VE-cadherin in collecting lymphatic vessels in ear skin from Tie1-deleted (control: n = 4; Tie1-deleted: n = 8), Tie2-deleted (control: n = 7; Tie2-deleted n = 6), Tie1/-2–deleted (control: n = 7; Tie1/-2–deleted n = 8), and Ang2 Ab–treated (n = 4 per group) P21 pups. Quantification of VEGFR3 immunofluorescence in collecting lymphatic vessels, normalized to control. Scale bars: 20 μm. Magnification: 1.86. Data represent the mean ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001, by 2-tailed Student’s t test.

Figure 5. Effect of Ang2 Ab and VEGF-C on the localization of VEGFR3 and Tie1 in cultured LECs.

(A) Cell-surface staining of VEGFR3 and Tie1 in LECs treated with a control IgG (n = 4) or Ang2 Ab (n = 4) for 30 minutes or stimulated with VEGF-C (n = 3) for 20 minutes. Scale bar: 50 μm. Magnification: 1.66. Graphs show quantification of VEGFR3 and Tie1 staining on the cell surface normalized to the IgG control. (B) Images showing colocalization of VEGFR3 and Tie1 in intracellular vesicles of LECs stimulated with VEGF-C for 20 minutes. Scale bar: 10 μm. Graph shows quantification of the number of VEGFR3- and Tie1-positive vesicles per nuclei (n = 3). Magnification: 2.36. (C and D) Colocalization of VEGFR3 with Tie1 and EEA1 (sorting endosomes) after 30 minutes of VEGF-C exposure (C) or RAB7 (late endosomes) after 120 minutes of VEGF-C exposure (D). The overview image shows a maximum projection. The boxed regions are shown as zoomed-in images of 1 optical slice per channel. White arrows point to some of the triple-positive (VEGFR3, Tie1, and EEA1 or RAB7) vesicles. Scale bars: 20 μm. Magnification: 3.24. (E and F) PLA immunofluorescence and quantification of PLA spots for Tie1 and VEGFR3 (E) and Ang2 and VEGFR3 (F) in permeabilized LECs (n = 5 fields of view; experiments were repeated 3 times with similar results). Scale bars: 10 μm. Data represent the mean ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001, by 1-way ANOVA with Dunnett’s post hoc test for multiple comparisons (A, E, and F) and 2-tailed Student’s t test (B).

Figure 6. Blocking of Ang2 or deletion of Ang2 or Tie1 decreases VEGFR3 expression and signaling.

(A–C) Western blots showing VEGFR3, Prox1, and HSC70 in ear lysates from control, Ang2-inhibited (n = 3 per group) (A), Ang2-deleted (n = 4 per group) (B), and Tie1-deleted (control: n = 6; Tie1-deleted: n = 3) (C) pups. Graphs show quantification of VEGFR3 polypeptides normalized to Prox1 and the control. (D) Western blots showing p-Akt, Akt, and HSC70 detection in VEGF-C and Ang2 Ab–treated LECs. (E) Quantification of the p-Akt/Akt ratio (n = 3 per group), normalized to VEGF-C-treated samples. (F) Ang2 concentration in LEC culture medium at the indicated time points after VEGF-C stimulation (n = 3 per group). (G and H) Western blots (WB) showing Ang2 and HSC70 in VEGF-C stimulated LECs (G) and Tie2 phosphorylation in VEGF-C–stimulated (45 min) and Ang2 Ab–treated LECs (H). The experiments in G and H were performed twice with similar results. Data represent the mean ± SEM (A–C and E) and ± SD (F). **P < 0.01 and ***P < 0.001, by 2-tailed Student’s t test (A–C and F) and 1-way ANOVA with Bonferroni’s post hoc test for multiple comparisons (E).

Figure 7. PI3K signaling regulates postnatal lymphatic vessel development and VEGFR3 expression.

(A) LYVE1 staining of ventral ear skin from control and Pik3ca-deleted P21 pup (n = 3 per group). Scale bar: 1 mm and 200 μm (higher-magnification insets). Graphs show quantification of the LYVE1-positive vessel area and vessel diameter. (B) Dorsal ear skin was immunostained for podoplanin, αSMA, and integrin α9. Scale bar: 100 μm. Graphs show quantification of SMC coverage, the number of valves, and the widest and narrowest vessel diameters and the difference between the 2 in collecting lymphatic vessels. (C) Lymphatic capillaries were immunostained for VEGFR3 and GOLPH4. Arrows point to VEGFR3 in the Golgi complex. Scale bars: 10 μm. Graphs show quantification of total VEGFR3 and its fraction in the Golgi complex, normalized to control. (D) Cell-surface immunostaining for VEGFR3 in lymphatic capillaries. Graph shows quantification of VEGFR3 immunofluorescence at the cell surface, normalized to control. Scale bar: 20 μm. Magnification: 0.40. (E) Immunostaining for VEGFR3 in collecting lymphatic vessels. Scale bar: 20 μm. Graph shows quantification of VEGFR3 immunofluorescence in collecting lymphatic vessels, normalized to control. Data represent the mean ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001, by 2-tailed Student’s t test.

Figure 8. PI3K, but not Ang2, regulates VEGFR3 expression in Pik3ca^H1047R-driven LMs.

(A) Diagram showing the induction of progressive microcystic LM and its treatment by using the soluble VEGF-C trap (AAV-VEGFR3-Ig; AAV-sR3) combined or not with the PI3K pathway inhibitor dactolisib or alpelisib (BYL719). (B) LYVE1 and VEGFR3 staining of ears from 7.5-week-old Vegfr3CreER^T2 R26-LSL-Pik3ca^H1047R mice treated with 4-OHT at 3 weeks of age, followed by treatment with dactolisib, AAV-sR3, and/or vehicle for 1.5 weeks. (C) Quantification of VEGFR3 in lymphatic vessels from Vegfr3CreER^T2 R26-LSL-Pik3ca^H1047R mice treated with AAV-Ctrl plus vehicle (n = 4), AAV-sR3 plus vehicle (n = 6), AAV-Ctrl plus dactolisib (n = 3), or AAV-sR3 plus dactolisib (n = 3), normalized to control. (D) Diagram showing the induction of progressive microcystic LM and treatment with IgG or Ang2 Ab. (E) LYVE1 and VEGFR3 staining of ears from Vegfr3CreER^T2 R26-LSL-Pik3ca^H1047R mice treated with 4-OHT at 3.5 weeks of age, followed by treatment with IgG or Ang2 Ab for 2 weeks. (F) Quantification of VEGFR3 in lymphatic vessels of Vegfr3CreER^T2 R26-LSL-Pik3ca^H1047R mice treated with IgG or Ang2 ab (n = 3 per group), normalized to control. Scale bars: 200 μm. Data represent the mean ± SEM. *P < 0.05, by 1-way ANOVA with Bonferroni’s post hoc test for multiple comparisons (C) and 2-tailed Student’s t test (F).

Figure 9. Ang2 and Tie receptors regulate VEGF-C–induced lymphangiogenesis in adult mice.

(A) VEGFR3 staining of ear skin from control and Tie1/-2–deleted mice treated with Ad-LacZ or Ad–VEGF-C for 10 days. (B) Percentage of VEGFR3-positive area and number of sprouts per field (Ad-LacZ: n = 3 ears; Tie1;Tie2-del+Ad-LacZ: n = 3 ears; Ad-VEGF-C: n = 4 ears; Tie1;Tie2-del+Ad-VEGF-C: n = 4 ears). (C) LYVE1 staining of skeletal muscle from control and Tie1/-2–deleted mice treated with AAV-empty or AAV-VEGF-C for 4 weeks. (D) Percentage of LYVE1-positive area (AAV-empty: n = 4 muscles; Tie1;Tie2-del plus AAV-Empty: n = 6 muscles; AAV-VEGF-C: n = 6 muscles; Tie1;Tie2-del plus AAV-VEGF-C: n = 6 muscles). (E) LYVE1 staining of ear skin in mice treated for 1.5 weeks with Ad-VEGF-C, Ang2 Ab, or both. (F) Percentage of LYVE1-positive area and number of sprouts per field (n = 8 ears per group). Scale bars: 100 μm. Data represent the mean ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001, by 1-way ANOVA with Bonferroni’s post hoc test for multiple comparisons and 2-tailed Student’s t test.

Figure 10. Schematic summary of how Ang2/Tie/PI3K signaling controls lymphangiogenesis via regulation of VEGFR3 cell-surface expression.

Ang2 is shown as a tetramer of 2 asymmetric dimers binding to and activating the Tie2-Tie1 cluster (, –68). Ang/Tie signaling promotes activation or PI3K and Akt, leading to inhibition of FoxO1 and its target genes, such as Angpt2 (24, 27). In LECs, VEGF-C increased Ang2 release from stimulated cells and subsequent Tie2 and Akt activation. Disruption of Ang2/Tie signaling in Ang2 Ab–treated or Tie1- or Tie1/-2–deleted pups resulted in increased Angpt2 gene expression, suggesting that Ang2 acts as an agonistic Tie2 ligand in LECs that promotes PI3K activation. VEGF-C induces internalization of VEGFR3 and its vesicular trafficking for degradation or recycling, which are regulated by PI3K (–42). VEGF-C stimulation led to cotrafficking of VEGFR3 and Tie1 into EEA1/RAB5-positive early/sorting endosomes and, subsequently, to the RAB7-positive late endosomal degradative vesicle route. Our results show that inhibition or deletion of Ang2 or PI3K, or deletion of Tie receptors, promoted loss of VEGFR3 from the LEC surface and its increased degradation, leading to decreased lymphangiogenesis.