Chronic inflammation triggered by the NLRP3 inflammasome in myeloid cells promotes growth plate dysplasia by mesenchymal cells

Abstract

Skeletal complications are common features of neonatal-onset multisystem inflammatory disease (NOMID), a disorder caused by NLRP3-activating mutations. NOMID mice in which NLRP3 is activated globally exhibit several characteristics of the human disease, including systemic inflammation and cartilage dysplasia, but the mechanisms of skeletal manifestations remain unknown. In this study, we find that activation of NLRP3 in myeloid cells, but not mesenchymal cells triggers chronic inflammation, which ultimately, causes growth plate and epiphyseal dysplasia in mice. These responses are IL-1 signaling-dependent, but independent of PARP1, which also functions downstream of NLRP3 and regulates skeletal homeostasis. Mechanistically, inflammation causes severe anemia and hypoxia in the bone environment, yet down-regulates the HIF-1α pathway in chondrocytes, thereby promoting the demise of these cells. Thus, activation of NLRP3 in hematopoietic cells initiates IL-1β-driven paracrine cascades, which promote abnormal growth plate development in NOMID mice.

Figure 1

Constitutive activation of the NLRP3 inflammasome in osteochondro-progenitors does not cause inflammation and abnormal growth plate development. All data were obtained from 4-week old WT, Nlrp3 ^DM1, Ikk2 ^DM1 or Nlrp3 ^DM1; Ikk2 ^DM1 male mice (n = 3–10/genotype). (A) Red blood cell-depleted bone marrow cells were stained with isotype control (data not shown) or with antibodies against CD11b and Gr1. Representative flow cytometry dot plots of the CD11b⁺/Gr1⁺ myeloid cells from each genotype are shown. (B) The percentage of CD11b⁺/Gr1⁺ cells in bone marrow cells. Data are expressed as mean ± SEM. *P < 0.05; **P < 0.005. (C) IL-1β levels in conditioned media from BMM treated with 100 ng/ml LPS for 3 hours, then with 5 mM ATP for 30 minutes. Data are expressed as mean ± SD of 3 independent experiments carried out in triplicates. NS, not significant. (D) H&E staining. GP, growth plate. Scale bar, 100 µm. (E) 3D μCT reconstruction of distal femoral metaphysis. Scale bar, 1 mm.

Figure 2

Constitutive activation of the NLRP3 inflammasome in myeloid cells causes inflammation and growth plate defects independently of PARP1. All data were obtained from 2-week old WT, Nlrp3 ^LysM, Parp1 ^D214N/D214N or Nlrp3 ^LysM; Parp1 ^D214N/D214N male mice (n = 3–4/genotype). (A) Red blood cell-depleted bone marrow cells were stained with isotype control (data not shown) or with antibodies against CD11b and Gr1. Representative flow cytometry dot plots of the CD11b⁺/Gr1⁺ myeloid cells from each genotype are shown. (B) The percentage of CD11b⁺/Gr1⁺ cells in bone marrow cells. Data are expressed as mean ± SEM. **P < 0.005. NS, not significant. (C) IL-1β levels in conditioned media from BMM treated with 100 ng/ml LPS for 3 hours, then with 5 mM ATP for 30 minutes. Data are representative of at least three independent experiments and expressed as mean ± SEM. ^$P < 0.001 vs. WT + LPS. (D) H&E staining. Red circle and bracket indicate areas of hypocellularity within the enlarged center of the epiphysis and growth plate disorganization, respectively. Scale bar, 150 µm. (E) 3D μCT reconstruction of distal femoral metaphysis. Scale bar, 0.5 mm.

Figure 3

Constitutive activation of the NLRP3 inflammasome in myeloid cells causes inflammation and growth plate defects through IL-1 signaling. All data were obtained from 2-week-old WT, Nlrp3 ^LysM, Il-1r ^−/− or Nlrp3 ^LysM; Il-1r ^−/− male mice (n = 4/genotype). (A) Red blood cell-depleted bone marrow cells were stained with isotype control (data not shown) or with antibodies against CD11b and Gr1. Representative flow cytometry dot plots of the CD11b⁺/Gr1⁺ myeloid cells from each genotype are shown. (B) Safranin O staining. Red circle and bracket indicate areas of hypocellularity within the enlarged center of the epiphysis and growth plate disorganization, respectively. Scale bar, 100 µm. (C) 3D μCT reconstruction of distal femoral metaphysis. Scale bar, 0.5 mm.

Figure 4

Constitutive activation of the NLRP3 inflammasome in myeloid cells causes anemia of inflammation through IL-1 signaling. All data were obtained from 2-week old WT or mutant male mice. (A–E) Cell blood counts (n = 4 mice/genotype). WBC, white blood cell counts, RBC, red blood cell counts. (F) Bone marrow cells (n = 4 mice/genotype) were stained with isotype control (data not shown) or with antibodies against CD11b and Ter119. Representative flow cytometry dot plots of CD11b⁺ myeloid cells or Ter119⁺ erythrocytes from each genotype are shown. (G) The percentage of Ter119⁺ erythrocytes in bone marrow cells. Data are expressed as mean ± SEM. *P < 0.05; **P < 0.005. (H) Body weight. (I) Spleen weight. (J) Percent of spleen weight (spleen weight was normalized to body weight). Quantitative data (H–J) were obtained from 5–6 mice/genotype and expressed as the mean ± SEM. *P < 0.05; **P < 0.005. (K) Body weight. (L) Spleen weight. (M) Percent of spleen weight. Quantitative data (K–M) were obtained from 3–5 mice/genotype and expressed as the mean ± SEM. **P < 0.005. NS, not significant.

Figure 5

Activation of NLRP3 in myeloid cells impairs chondrocyte responses to hypoxia. (A) Femoral sections from 2-week-old WT or Nlrp3 ^LysM male mice injected with hydroxyprobe were stained with IgG or hydroxyprobe antibody. A specimen from renal medulla was used as a positive control. Staining is indicated by the brown color. HZ, hypertrophic zone. (B) qPCR analysis of mRNA isolated from mouse epiphyses. Quantitative data were obtained from 2–3 mice/genotype and expressed as the mean ± SEM. *P < 0.05; **P < 0.005.

Figure 6

Activation of NLRP3 in myeloid cells causes excessive production of IL-1β, a cytokine that promotes chondrocyte death. (A) Bone marrow cells were centrifuged at 400 × g, and the supernatants were harvested for IL-1β measurement. **P < 0.005. (B) Western blot analysis of the effects of IL-1β on PARP1 cleavage in chondrocytes. Cells were treated with IL-1β for 24 hours. β-actin was used as loading control. Cont, control. qPCR analysis of the expression of HIF-1α (C), VEGF (D), Ca9 (E), PGK1 (F), Glut1 (G) and Binp3 (H). RNA were isolated from murine chondrocytes from the ribs of 2-day old pups. Data are expressed as the mean ± SEM. *P < 0.05; **P < 0.005; Cont, control. Rel, relative. NS, not significant. ELISA analysis of IL-1β levels in conditioned media (I) and Western blot analysis of NLRP3 expression (J) from BMM treated with 100 ng/ml LPS for 3 hours, then with 5 mM ATP for 30 minutes in normoxic or hypoxic conditions. Data are representative of 2–3 independent experiments and expressed as mean ± SEM. *P < 0.05; **P < 0.005; ^$P < 0.001 vs. WT normoxia + LPS.