Macrophage 11β-HSD-1 deficiency promotes inflammatory angiogenesis

Abstract

11β-Hydroxysteroid dehydrogenase-1 (11β-HSD1) predominantly converts inert glucocorticoids into active forms, thereby contributing to intracellular glucocorticoid levels. 11β-HSD1 is dynamically regulated during inflammation, including in macrophages where it regulates phagocytic capacity. The resolution of inflammation in some disease models including inflammatory arthritis is impaired by 11β-HSD1 deficiency or inhibition. However, 11β-HSD1 deficiency/inhibition also promotes angiogenesis, which is beneficial in mouse models of surgical wound healing, myocardial infarction or obesity. The cell types responsible for the anti-inflammatory and anti-angiogenic roles of 11β-HSD1 have not been characterised. Here, we generated Hsd11b1^MKO mice with LysM-Cre mediated deletion of Hsd11b1 to investigate whether 11β-HSD1 deficiency in myeloid phagocytes is pro-angiogenic and/or affects the resolution of inflammation. Resolution of inflammatory K/BxN-induced arthritis was impaired in Hsd11b1^MKO mice to a similar extent as in mice globally deficient in 11β-HSD1. This was associated with >2-fold elevation in levels of the endothelial marker Cdh5 mRNA, suggesting increased angiogenesis in joints of Hsd11b1^MKO mice following arthritis. A pro-angiogenic phenotype was confirmed by measuring angiogenesis in subcutaneously implanted polyurethane sponges, in which Hsd11b1^MKO mice showed 20% greater vessel density than their littermate controls, associated with higher expression of Cdh5 Thus, 11β-HSD1 deficiency in myeloid phagocytes promotes angiogenesis. Targeting 11β-HSD1 in macrophages may be beneficial in tissue repair.

Keywords: 11β-HSD1; glucocorticoid; inflammation; macrophage; steroid metabolism.

Figure 1

11β-HSD1 activity is markedly reduced in resident peritoneal macrophages from Hsd11b1^MKO mice. Resident peritoneal cells, or cells elicited to the peritoneum by i.p. injection of 0.2 mL 10% thioglycollate (TG), were harvested from Hsd11b1^MKO (MKO: white bars) and control Hsd11b1^f/f mice (Con: black bars). 11β-HSD1 activity in peritoneal cells was measured by conversion of [³H]-11-dehydrocorticosterone to corticosterone in resident peritoneal cells (A) and in cells elicited to the peritoneum 24 h (B) or 96 h (C) following thioglycollate injection. Activity was also measured in purified neutrophils (Ly6G⁺ cells) isolated from the peritoneum 24 h after thioglycollate injection (D). 11β-HSD1 activity is expressed as pmol corticosterone/h/10⁵ cells. Values are means ± s.e.m. and were analysed by unpaired t-test (n = 3–8, *P < 0.05).

Figure 2

Resolution of K/BxN serum transfer-induced arthritis is impaired in Hsd11b1^MKO mice. Arthritis was induced in Hsd11b1^Del1/Del1 (global knockout, GKO: triangular symbols, n = 3), Hsd11b1^MKO(MKO: circular symbols/white bars, n = 6) and control Hsd11b1^f/fmice (Con: square symbols/black bars, n = 7) by i.p. injection of 100 µL K/BxN serum at day 0. (A) Clinical scoring of joint inflammation over 21 days. Values are means ± s.e.m. for Hsd11b1^MKO and control Hsd11b1^f/f mice. Data were analysed by two-way repeated measurement analysis of variance (ANOVA) with Bonferroni’s multiple comparisons test; *P < 0.05, **P < 0.01, ***P < 0.001. Only the mean value is shown for Hsd11b1^Del1/Del1mice. (B) The area under the curve (AUC) was calculated for the clinical score from days 9 to 21 (the resolution phase). Values are means ± s.e.m. Data were analysed by one-way ANOVA (P < 0.01) with post-hoc Dunnett’s multiple comparisons test with the Hsd11b1^f/f group as control; *P < 0.05. (C) Representative sections of joints from Hsd11b1^MKO and Hsd11b1^f/f mice collected 21 days after injection and stained with haematoxylin and eosin showing tenosynovitis characterised by synovium hyperplasia, bone erosion and new bone formation. Scale bars, 100 µm. (D) Histopathological changes were quantified using a scoring index (see the ‘Materials and methods’ section for details) by an investigator blind to genotype. Values are means ± s.e.m. Data were analysed by Mann–Whitney test, n = 6, *P < 0.05.

Figure 3

Higher Cdh5 expression suggests greater angiogenesis in the joints of Hsd11b1^MKO mice. Arthritis was induced by i.p. injection of 100 µL K/BxN serum, and the mice were killed 21 days later. (A) Immunofluorescent staining of blood vessels in the mesenchymal tissue of Hsd11b1^MKO (MKO: top) and control mice (Con: bottom). From left to right: DAPI, isolectin B4, α-SMA. (B) RNA was extracted from hind joints and qPCR used to measure levels of Cdh5 mRNA relative to Hprt, used as an internal standard. Values are in arbitrary units (AU) and are means ± s.e.m. Data from Hsd11b1^MKO (white bar) and Hsd11b1^f/f mice (black bar) were analysed by unpaired t-test, n = 5–7, *P < 0.05.

Figure 4

Compared to littermate controls, Hsd11b1^MKO mice show greater angiogenesis in the subcutaneous sponge implantation assay. (A) Representative images (×125 magnification) of haematoxylin and eosin-stained sections of implanted sponges removed after 21 days, showing neovascularisation (blood vessels: arrows). (B) Quantification of blood vessel numbers by Chalkley counting. RNA was extracted from sponges removed 21 days after implantation, and qPCR was used to measure levels of: (C) Cdh5 mRNA, (D) Il1 mRNA, and (E) Angiopoietin-1, -2 and -4 mRNAs, relative to levels of Hprt mRNA, used as an internal standard (mRNA values in arbitrary units, AU). Values are means ± s.e.m. Data from Hsd11b1^MKO (MKO: white bars) and Hsd11b1^f/f mice (Con: black bars) were analysed by unpaired t-test (A, B, C, E) or unpaired t-test with Welch’s correction (D); *P < 0.05, n = 10–11.