Bone morphogenetic protein 9 (BMP9) controls lymphatic vessel maturation and valve formation

Abstract

Lymphatic vessels are critical for the maintenance of tissue fluid homeostasis and their dysfunction contributes to several human diseases. The activin receptor-like kinase 1 (ALK1) is a transforming growth factor-β family type 1 receptor that is expressed on both blood and lymphatic endothelial cells (LECs). Its high-affinity ligand, bone morphogenetic protein 9 (BMP9), has been shown to be critical for retinal angiogenesis. The aim of this work was to investigate whether BMP9 could play a role in lymphatic development. We found that Bmp9 deficiency in mice causes abnormal lymphatic development. Bmp9-knockout (KO) pups presented hyperplastic mesenteric collecting vessels that maintained LYVE-1 expression. In accordance with this result, we found that BMP9 inhibited LYVE-1 expression in LECs in an ALK1-dependent manner. Bmp9-KO pups also presented a significant reduction in the number and in the maturation of mesenteric lymphatic valves at embryonic day 18.5 and at postnatal days 0 and 4. Interestingly, the expression of several genes known to be involved in valve formation (Foxc2, Connexin37, EphrinB2, and Neuropilin1) was upregulated by BMP9 in LECS. Finally, we demonstrated that Bmp9-KO neonates and adult mice had decreased lymphatic draining efficiency. These data identify BMP9 as an important extracellular regulator in the maturation of the lymphatic vascular network affecting valve development and lymphatic vessel function.

Figure 1

Abnormal structure of lymphatic mesenteric collecting vessels in Bmp9-KO neonates at P0. Representative macroscopic views of mesenteric vessels in WT (A) and Bmp9-KO (B) pups at P0. The lymphatic vessels are filled with white chyle and constrictions along the vessels reflecting the presence of valves are shown with white arrows. Whole-mount fluorescence immunostainings of WT (C) and Bmp9-KO (D) mesenteries with CD31 (red) and Prox1 (green); arrows point to Prox1-overexpressing cells in valve areas. Higher magnification views of the mid-part of mesenteric collecting vessels in WT (E) or Bmp9-KO (F). Enlarged vessels, as outlined by the red lines, showing vessel sections are observed in mutants. (G) Quantification of the mean collecting vessel section in WT and Bmp9-KO pups. The mean diameter of mesenteric lymphatic collecting vessels was calculated as the average of 4 different measurements performed using Axiovision 4.8 software throughout the length of 4 different collecting vessels per mesentery. (H-I) Quantification of LEC number per millimeter of collecting vessel length and by square millimeter of collecting vessel surface unit in WT and Bmp9-KO pups. Prox1-positive nuclei were counted per vessel unit length and per vessel surface unit on at least 3 collecting vessel per mesentery using Axiovision 4.8 software. Values are the mean (± SE) from 8 individuals per genotype. ns, not significant; ***P ≤ .001, significantly different from WT pups by unpaired Student t test.

Figure 2

LYVE-1 expression is regulated by BMP9 both in vivo in lymphatic collecting vessels of the mesentery and in vitro in cultured LECs. (A) P0 lymphatic mesenteric collecting vessels were stained for LYVE-1 (red) and Prox1 (green). The dashed lines delineate the lymphatic vessels. Arrow indicates a valve. A nonlymphatic LYVE-1 immunoreactivity is also observed in isolated cells outside vessels, which probably correspond to macrophages. (B) Time-dependent inhibition of LYVE-1 mRNA expression in LECs treated with BMP9 (10 ng/mL); the results are presented as mRNA fold changes measured in BMP9-treated cells vs nontreated cells at each time point. Data are the mean ± SE from 6 independent experiments performed in duplicate. *P ≤ .05, significantly different from cytotoxic T lymphocyte (CTL) by Mann-Whitney U test. (C,E) Flow cytometry detection of LYVE-1 protein expression in LECs transfected for 24 hours with scramble siRNA or siRNA targeting Alk1 and then stimulated with (pink) or without (green) 10 ng/mL BMP9 for 48 hours. NS corresponds to fluorescence-activated cell sorter analysis in absence of antibodies (black). (D) LYVE-1 mRNA expression in LECs transfected for 24 hours with scrambled siRNA (siScr) or siRNA targeting Alk1 and then stimulated with or without 10 ng/mL BMP9 for 24 hours; data represent mean ± SE (n = 6). **P ≤ .01, significantly different from CTL by Mann-Whitney U test. Inset shows Alk1 mRNA downregulation by siAlk1 (mean ± SE, n = 4). *P ≤ .05, significantly different from siCTL by Mann-Whitney U test.

Figure 3

Defective lymphatic valve formation in Bmp9-KO embryos and neonates. (A) WT P0 lymphatic mesenteric collecting vessels were stained for Prox1 (green) and CD31 (red) to allow the discrimination between the different valve maturation stages; arrows point to the valve location; bars, 100 µm. (B-D) Quantitative analysis of valve formation at E18.5, P0, and P4 in WT and Bmp9-KO mice. Values correspond to the number of valves counted on 4 collecting vessels per mesentery; n = 6 (Bmp9-KO) or n = 7 (WT) at E18.5; n = 8 (Bmp9-KO) or n = 9 (WT) at P0; n = 12 (Bmp9-KO) or n = 11 (WT) at P4. ns, not significant. **P ≤ .01, ***P ≤ .001, significantly different from WT pups by unpaired Student t test.

Figure 4

BMP9-regulated genes, known to be involved in lymphatic valve development, in LECs. LECs were stimulated in serum-free medium in the absence or in the presence of 10 ng/mL BMP9 for the indicated times. Expression of HPRT was used to normalize the samples. The results are represented as mRNA fold changes measured in BMP9-treated cells vs nontreated cells at each time point. Data are the mean ± SE from 4 independent experiments performed in duplicate. *P ≤ .05, significantly different from respective controls by Mann-Whitney U test.

Figure 5

Abnormal patterning of ear lymphatic capillaries in adult Bmp9-KO mice. (A) Adult ear lymphatic capillaries were stained for LYVE-1 (green) and CD31 (red). Note that the LYVE-1 staining overlies weak CD31 staining in the lymphatic vessels. Bar represents 300 µm. (B) Quantification of the mean LYVE-1–positive area expressed as percentage of total area. The lymphatic vessel area in the inner layer of the ear of adult mice was measured using Image J software on images corresponding to 3 different fields of whole ear skin. These regions were kept constant for all samples. (C) Quantification of the mean LYVE-1–positive lymphatic vessel section expressed in micrometers. To quantify ear skin capillary mean lymphatic size, 5 horizontal lines were evenly laid on the images, and the diameters of lymphatic vessels that crossed these lines with an angle above 45° were measured using Image J according to Zhou et al. Values are mean ± SE; n = 8 for each genotype. ***P ≤ .001, significantly different from WT pups by unpaired Student t test.

Figure 6

Impairment of lymphatic drainage in adult Bmp9-KO mice. (A) Representative fluorescence images of WT and Bmp9-KO hind limbs obtained 3 and 15 minutes after injection of DiD-lipidots. The extremity of the paw was hidden not to saturate the images by the fluorescence signal at the injection site. The black arrowhead outlines lymphatic collectors; the white arrow indicates the popliteal lymph node. The values of the scale bars for fluorescence intensity were adjusted to normalize each image series at comparable values. (B) Quantitative analysis of dye accumulation into the popliteal lymph node. Signal over noise ratio was measured. Values are means ± SE; n = 6 (Bmp9-KO) or n = 9 (WT). *P ≤ .05, significantly different by Mann-Whitney U statistical test.

Figure 7

Working model for BMP9 regulation of lymphatic vessel maturation and valve development. In LECs, BMP9 acting via ALK1 inhibits LYVE-1 expression. In parallel, BMP9 also via Alk1, induces the expression of Foxc2, Cx37, ephrinB2, and Nrp-1, which have all been involved in lymphatic valve formation. Flow has been demonstrated to induce Foxc2 expression and to act in concert with Prox1 to regulate early steps of lymphatic valve morphogenesis. ALK1 expression has been reported to be regulated by flow in zebrafish. Therefore, we propose that flow could increase ALK1 expression, which would raise BMP9 signaling to favor the downstream cascade of lymphatic vessel development.