A neuro-lymphatic communication guides lymphatic development by CXCL12 and CXCR4 signaling

Abstract

Lymphatic vessels grow through active sprouting and mature into a vascular complex that includes lymphatic capillaries and collecting vessels that ensure fluid transport. However, the signaling cues that direct lymphatic sprouting and patterning remain unclear. In this study, we demonstrate that chemokine signaling, specifically through CXCL12 and CXCR4, plays crucial roles in regulating lymphatic development. We show that LEC-specific Cxcr4-deficient mouse embryos and CXCL12 mutant embryos exhibit severe defects in lymphatic sprouting, migration and lymphatic valve formation. We also discovered that CXCL12, originating from peripheral nerves, directs the migration of dermal lymphatic vessels to align with nerves in developing skin. Deletion of Cxcr4 or blockage of CXCL12 and CXCR4 activity results in reduced VEGFR3 levels on the LEC surface. This, in turn, impairs VEGFC-mediated VEGFR3 signaling and downstream PI3K and AKT activities. Taken together, these data identify previously unknown chemokine signaling originating from peripheral nerves that guides dermal lymphatic sprouting and patterning. Our work identifies for the first time a neuro-lymphatics communication during mouse development and reveals a previously unreported mechanism by which CXCR4 modulates VEGFC, VEGFR3 and AKT signaling.

Keywords: CXCL12/CXCR4; Chemokine signaling; Lymphangiogenesis; Lymphatic development; Peripheral nerves; VEGFC/VEGFR3.

Fig. 1.

CXCR4 is expressed in LECs of migrating lymphatic vessels and collecting lymphatic valves. (A) Immunostaining of wild-type E11.5 transverse sections with antibodies against Lyve1, CXCR4 and Prox1. Arrows indicate CXCR4-positive migrating LECs. CV, cardinal vein; DA, dorsal aorta. (B) Whole-mount immunostaining of wild-type E14.5 skin with antibodies against Prox1, CXCR4 and Nrp2. Arrows indicate CXCR4-positive lymphatic sprouting tips. (C) Whole-mount immunostaining of wild-type E18.5 mesentery with antibodies against VEGFR3, CXCR4 and Prox1. Arrows indicate collecting lymphatic valves. Scale bars: 100 μm.

Fig. 2.

Deletion of CXCR4 in LECs results in defective lymphatic development with impaired lymphatic sprouting and migration at E14.5. (A) Protocol for TM injections into pregnant mice at E9.5 and E10.5, and the harvesting of embryos at E14.5. (B) Bright-field images show that CXCR4^ΔLEC/ΔLEC embryos developed severe edema at E14.5. Whole-mount immunostaining of dermal skins showed defective lymphatic vascular development with an increased midline distance in E14.5 CXCR4^ΔLEC/ΔLEC embryos. (C) Quantification of lymphatic midline distance (n=6 controls; 12 CXCR4^ΔLEC/ΔLEC embryos). (D) Whole-mount staining of dermal skins of control and CXCR4^ΔLEC/ΔLEC embryos with antibodies against Nrp2, Prox1 and PECAM1. Images were taken at the migrating fronts. Arrows indicate sprouting lymphatic tips. (E) Quantification of numbers of lymphatic sprouting tips and branches, and lymphatic vessel diameters and relative lymphatic area (n=6 controls; n=12 CXCR4^ΔLEC/ΔLEC embryos). (F) Whole-mount staining of skin samples of control and CXCR4^ΔLEC/ΔLEC embryos with antibodies against Nrp2, Prox1 and PECAM1. Images were taken at the lymphatic plexus regions. (G) Quantification of lymphatic vessel diameters and branch numbers. n=4 embryos per group. Control embryos are TM-treated Cre− and Cre+;Cxcr4f/+ littermates. Data are mean±s.e.m. *P<0.05, ***P<0.005, ****P<0.001, unpaired two-tailed Student's t-test. Scale bars: 2 mm (whole-embryo images in B); 100 μm (B,D,F).

Fig. 3.

CXCL12 is mainly derived from Schwann cells in peripheral nerves at E14.5. (A) Whole-mount staining of dermal skins of Cxcl12-DsRed embryos with antibodies against PECAM1, RFP and Nrp2. Enlarged image indicates that RFP-positive staining cells (white arrow) do not colocalize with blood vessels (yellow arrow). (B) Whole-mount staining of dermal skin of Cxcl12-DsRed embryos with antibodies against Nrp2, Tuj1 and RFP. White arrows indicate that CXCL12-expressing cells colocalize with Tuj1-positive nerves; yellow arrows indicate CXCL12-expressing cells surrounding nerves. (C) Whole-mount staining of dermal skins of Cxcl12-DsRed embryos with antibodies against Tuj1, NG2 and Nrp2. (D) Whole-mount staining of dermal skins of Cxcl12-DsRed embryos with antibodies against FABP7, RFP and Tuj1. White arrows indicate that CXCL12-expressing cells are FABP7-positive Schwann cells. n=3 embryos per group. Scale bars: 100 μm.

Fig. 4.

CXCL12 mutants show reduced lymphatic vessel sprouting with enlarged lymphatic vessels at E14.5. (A) Whole-mount staining of dermal skin of CXCL12^DsRed/+ embryos and CXCL12^DsRed/DsRed embryos with antibodies against Nrp2. Arrows indicate midline distance. Scale bars: 500 μm. (B) Quantification of the midline distance. n=3 embryos per group. (C-H) Whole-mount staining of dermal skin of CXCL12^DsRed/+ embryos and CXCL12^DsRed/DsRed embryos with antibodies against Nrp2, Prox1 and PECAM1. D,E and G,H are enlarged images from the outlined areas in C and F, respectively. Scale bars: 100 μm. (I) Quantification of the numbers of sprouts and branches per field. n=3 embryos per group. Data are mean±s.e.m. *P<0.05, unpaired two-tailed Student's t-test.

Fig. 5.

Loss of CXCR4 leads to reduced surface levels of VEGFR3 in LECs. (A) Whole-mount staining of E14.5 skins of control and CXCR4^ΔLEC/ΔLEC embryos with antibodies against VEGFR3 and Prox1 in the presence of the permeabilization reagent Triton. Regions outlined are enlarged in the right panels. Arrows indicate surface VEGFR3 in LECs in control embryos. Arrowheads indicate perinuclear patterns of VEGFR3 in LECs of CXCR4^ΔLEC/ΔLEC embryos. n=4 controls; n=3 CXCR4^ΔLEC/ΔLEC embryos. Scale bars: 10 μm. (B) Whole-mount staining of E14.5 skins of control and CXCR4^ΔLEC/ΔLEC embryos with antibodies against VEGFR3 and Lyve1 without the permeabilization reagent Triton. Dashed lines indicate Lyve1-positive lymphatic vessels. Quantification of VEGFR3 intensity in LECs of control and CXCR4^ΔLEC/ΔLEC embryos are shown in the right panels of A and B. Control embryos are TM-treated Cre− and Cre+;Cxcr4f/+ littermates. n=3 embryos per group. Data are mean±s.e.m. **P<0.01, unpaired two-tailed Student's t-test. Scale bars: 50 μm.

Fig. 6.

Blockage of CXCR4 activity impairs VEGFC-induced VEGFR3 intracellular trafficking and attenuates VEGFC-induced VEGFR3/PI3K/AKT activities and lymphatic functions. (A) PLA (VEGFR3 and CXCR4 antibodies) followed by VE-Cadherin staining of LECs stimulated with or without VEGFC for 30 min. PLA with VEGFR3 antibodies (IgG served as negative control). Quantification of relative PLA signals (fold-change) is shown in the right panel. n=3 repeats. Data are mean±s.e.m. **P<0.01, unpaired two-tailed Student's t-test. (B-E) Immunostaining of LECs with antibodies against VEGFR3, CXCR4 and EEA1 after LECs were treated with VEGFC for 15 min with or without AMD3100 pre-treatment. Enlarged images and arrowheads indicate colocalization of VEGFR3 and CXCR4 in EEA1-positive early endosome in LECs. (F) Relative triple-positive vesicles (VEGFR3, CXCR4 and EEA1) in LECs among groups are quantified as fold change. (G) Profile intensity of a triple-positive vesicle in VEGFC-treated LECs. (H) Western blot analysis of LECs silenced with CXCR4 siRNA (siCXCR4) or control siRNA, and then stimulated with or without VEGFC for 15 min. (I) Quantification of total VEGFR3, p-VEGFR3, p-AKT and p-ERK activities. n=4 repeats. (J) Bright-field images of LEC sprouting (top panel), wound healing (middle panel) and tube formation (bottom panel) assays when LECs were treated overnight with or without VEGFC in the presence of AMD3100. Quantifications of the spheroid sprouting assay (n=3 repeats), wound healing (n=3 repeats) and tube formation assays (n=5 repeats) are shown on the right. Data are mean±s.e.m. *P<0.05, **P<0.01, ***P<0.005, ****P<0.001, one-way ANOVA followed by Tukey's test. Scale bars: 50 μm in A; 25 μm in B-E; 500 μm in J.

Fig. 7.

Loss of CXCR4 in LECs leads to impaired mesenteric collecting lymphatic valve formation. (A) Diagram of TM administration and embryo harvest strategies. (B,C) Whole-mount immunostaining of E17.5 and E18.5 mesenteries of controls and CXCR4^ΔLEC/ΔLEC embryos with antibodies against VEGFR3, Prox1 and PECAM1. Quantification of the numbers of lymphatic valves in control and CXCR4^ΔLEC/ΔLEC embryos are shown in the right panels. (D) Whole-mount immunostaining of E18.5 mesenteries of controls and CXCR4^ΔLEC/ΔLEC embryos with antibodies against integrin α9, Prox1 and DAPI. Quantification of numbers of mature lymphatic valves in control and CXCR4^ΔLEC/ΔLEC embryos are shown on the right. Control embryos are TM-treated Cre− and Cre+;Cxcr4f/+ littermates. n=4 embryos per group for E17.5; n=5 controls; n=4 CXCR4^ΔLEC/ΔLEC embryos for E18.5. Arrows indicate lymphatic valve regions. Data are mean±s.e.m. ***P<0.005, ****P<0.001, unpaired two-tailed Student's t-test. Scale bars: 50 μm.

Fig. 8.

Postnatal deletion of CXCR4 does not affect lymphatic development. (A) The strategy for TM administration and harvest of mesenteries or ears at P7 or P21, respectively. (B) Whole-mount immunostaining of P21 ear skins of controls and CXCR4^ΔLEC/ΔLEC pups with antibodies against Lyve1 and PECAM1. Quantification of the percentage of lymphatic vessel area is shown on the right. n=4 controls; n=3 CXCR4^ΔLEC/ΔLEC mice per group. (C) Whole-mount staining of P7 mesenteries of control and CXCR4^ΔLEC/ΔLEC mice with antibodies against VEGFR3, Prox1 and PECAM1. Quantification of numbers of collecting lymphatic valves per vessel is shown on the right. Arrows indicate lymphatic valves. n=3 pups per group. Data are mean±s.e.m. ns, not significant, unpaired two-tailed Student's t-test. Scale bars: 100 μm.

Fig. 9.

Loss of CXCR4 in adult mice results in reduced lymphatic sprouting and VEGFR3 surface expression in response to VEGFC in vivo. (A) Whole-mount staining of adult ears of control and CXCR4^ΔLEC/ΔLEC mice treated with PBS or VEGFC for 5 days with antibodies against Lyve1. Arrows indicate sprouting tips of lymphatic vessels. Quantification of the number of lymphatic sprouts and percentage of lymphatic vessel area are shown on the right. n=3 mice per group. (B) Immunostaining of adult ears of wild-type mice treated with PBS or VEGFC with antibodies against Lyve1, CXCR4 and PECAM1. Lymphatic vessels are outlined. Arrows in enlarged images indicate CXCR4-expressing lymphatic vessels. Quantification of relative CXCR4 intensity is shown on the right. Data are mean±s.e.m. **P<0.01, unpaired two-tailed Student's t-test. (C) Immunostaining and quantification of permeabilized adult ear sections from control and CXCR4^ΔLEC/ΔLEC mice treated with PBS or VEGFC with antibodies against Lyve1, VEGFR3 and PECAM1. (D) Immunostaining and quantification of non-permeabilized adult ear sections from control and CXCR4^ΔLEC/ΔLEC mice treated with PBS or VEGFC with antibodies against Lyve1, VEGFR3 and PECAM1. n=4 mice per group. Data are mean±s.e.m. *P<0.05, **P<0.01, ***P<0.005, ****P<0.001, two-way ANOVA followed by Bonferroni's test. Scale bars: 100 μm in A,C,D; 50 μm in B.

Fig. 10.

How CXCL12/CXCR4 mediates VEGFC/VEGFR3/PI3K activities to regulate lymphatic development. CXCL12 is primarily expressed by Schwann cells in the peripheral nerves in the dermal skins of E14.5 embryos. CXCL12/CXCR4 is required for maintaining cell surface levels of VEGFR3 in LECs. VEGFC stimulation triggers CXCR4/VEGFR3 internalization and subsequently VEGFR3/PI3K/AKT signaling to regulate lymphatic development. Our results also indicate that VEGFC treatment upregulates CXCR4 expression, which might serve as a positive-feedback loop to enhance VEGFC/VEGFR3 signaling (dashed lines). Created in BioRender. Do, L. (2024) https://BioRender.com/j33w742.