Angiocrine Bmp2 signaling in murine liver controls normal iron homeostasis

Abstract

Microvascular endothelial cells (ECs) display a high degree of phenotypic and functional heterogeneity among different organs. Organ-specific ECs control their tissue microenvironment by angiocrine factors in health and disease. Liver sinusoidal endothelial cells (LSECs) are uniquely differentiated to fulfill important organ-specific functions in development, under homeostatic conditions, and in regeneration and liver pathology. Recently, Bmp2 has been identified by us as an organ-specific angiokine derived from LSECs. To study angiocrine Bmp2 signaling in the liver, we conditionally deleted Bmp2 in LSECs using EC subtype-specific Stab2-Cre mice. Genetic inactivation of hepatic angiocrine Bmp2 signaling in Stab2-Cre;Bmp2^fl/fl (Bmp2^LSECKO) mice caused massive iron overload in the liver and increased serum iron levels and iron deposition in several organs similar to classic hereditary hemochromatosis. Iron overload was mediated by decreased hepatic expression of hepcidin, a key regulator of iron homeostasis. Thus, angiocrine Bmp2 signaling within the hepatic vascular niche represents a constitutive pathway indispensable for iron homeostasis in vivo that is nonredundant with Bmp6. Notably, we demonstrate that organ-specific angiocrine signaling is essential not only for the homeostasis of the respective organ but also for the homeostasis of the whole organism.

Figure 1.

Angiocrine Bmp2 signaling in the liver controls tissue iron content and distribution. (A) Comparative qRT-PCR analysis of Bmp2 mRNA expression in LSECs, hepatocytes (HC), Kupffer cells (KC), and stellate cells (HSC) isolated from livers of wild-type (WT) adult C57Bl/6 mice (n = 5). β-2-Microglobulin was used as housekeeping gene. *P < .05; ***P < .001. (B) Bmp2 mRNA in situ hybridization of liver sections of Bmp2^LSECKO mice in comparison with WT controls (n = 3). Scale bar, 100 μm (×60). CV, central vein. (C) Hematoxylin and eosin (H.E.) and Sirius red staining of liver sections of Bmp2^LSECKO mice in comparison with WT controls (n = 6). Scale bar, 100 μm (×60). (D) Coimmunofluorescence of GS and arginase in liver (n = 5). Scale bar, 100 μm (×20). (E) Immunohistochemistry of LSEC markers in livers of Bmp2^LSECKO and control mice (n = 5). Scale bar, 100 μm (×60). (F) Prussian blue staining demonstrating iron deposition in liver, spleen, and pancreas of Bmp2^LSECKO (female, n = 5; male, n = 5). Scale bar, 100 μm (first column, ×10; third column, ×40; second and fourth columns, ×60). PF, portal field.

Figure 2.

Angiocrine Bmp2 signaling in the liver controls tissue and serum iron concentrations via regulation of hepatocyte hepcidin expression. (A-E) Iron concentration in (A) liver, (B) serum, (C) spleen, (D) heart, and (E) pancreas of Bmp2^LSECKO and control mice (female, n = 5; male, n = 5). *P < .05; **P < .01; ****P < .0001. (F) qRT-PCR of Hamp, Id1, Id2, Id3, Endoglin (Eng), and Atoh8 in liver of Bmp2^LSECKO and control mice (female, n = 5; male, n = 5). β-Actin was used as housekeeping gene. *P < .05; ***P < .001; ****P < .0001. (G) Serum hepcidin levels of Bmp2^LSECKO and control mice as measured by enzyme-linked immunosorbent assay (female, n = 5; male, n = 5). ***P < .001. (H) Gene expression heat map of murine hepatocytes stimulated with Bmp2 for 24 hours (left panel). Bmp2-dependent expression of Hamp, Id1, Id2, Id3, and Atoh8 was quantified by qRT-PCR (right panel) of hepatocytes stimulated with Bmp2 for 24 hours and of murine LSECs (mLSECs) stimulated with Bmp2 for 48 hours. β-Actin was used as housekeeping gene. *P < .05; **P < .01; ***P < .001; ****P < .0001.