NRF2 Is an Upstream Regulator of MYC-Mediated Osteoclastogenesis and Pathological Bone Erosion

Abstract

Osteoclasts are the sole bone-resorbing cells that play an essential role in homeostatic bone remodeling and pathogenic bone destruction such as inflammatory arthritis. Pharmacologically targeting osteoclasts has been a promising approach to alleviating bone disease, but there remains room for improvement in mitigating drug side effects and enhancing cell specificity. Recently, we demonstrated the crucial role of MYC and its downstream effectors in driving osteoclast differentiation. Despite these advances, upstream regulators of MYC have not been well defined. In this study, we identify nuclear factor erythroid 2-related factor 2 (NRF2), a transcription factor known to regulate the expression of phase II antioxidant enzymes, as a novel upstream regulator of MYC. NRF2 negatively regulates receptor activator of nuclear factor-κB ligand (RANKL)-induced osteoclastogenesis through the ERK and p38 signaling-mediated suppression of MYC transcription. Furthermore, the ablation of MYC in osteoclasts reverses the enhanced osteoclast differentiation and activity in NRF2 deficiency in vivo and in vitro in addition to protecting NRF2-deficient mice from pathological bone loss in a murine model of inflammatory arthritis. Our findings indicate that this novel NRF2-MYC axis could be instrumental for the fine-tuning of osteoclast formation and provides additional ways in which osteoclasts could be therapeutically targeted to prevent pathological bone erosion.

Keywords: MYC; NRF2; RANKL signaling; osteoclasts.

Figure 1

Extracellular signal-regulated protein kinase (ERK) and p38 activations are crucial for MYC expression after receptor activator of nuclear factor-κB ligand (RANKL) stimulation. Mouse osteoclast precursor cells (OCPs) were pretreated with either DMSO (vehicle), U0126 (5 μM), SP600125 (5 μM), SB203580 (10 μM) or LY294002 (5 μM) for 30 min and then stimulated with RANKL (50 ng/mL) for 6 h. (A) The mRNA expression of Myc (relative to the hypoxanthine guanine phosphoribosyl transferase (Hprt) housekeeping gene, n = 3). (B) Immunoblot of nuclear protein lysates using c-Myc and Lamin B antibodies. Lamin B served as the loading control. Data are representative of three experiments. (C) Signal intensity of the c-Myc immunoblot in B quantified using densitometry and normalized to Lamin B and to vehicle-treated RANKL control (n > 3). All data are shown as mean ± s.e.m. ** p < 0.01, *** p < 0.001 and **** p < 0.0001 using one-way ANOVA in (A,C); NS, not significant in (C).

Figure 2

Nuclear factor erythroid 2-related factor (NRF2) deficiency increases MYC transcription and protein expression. Control (wild-type (WT)) and NRF2-deficient (NRF2 knock-out (KO)) OCPs were stimulated with RANKL (50 ng/mL) for the indicated time points. (A) Immunoblot of nuclear protein lysates using NRF2, c-Myc, and Lamin B antibodies. Lamin B served as the loading control. Data are representative of three experiments. (B) The mRNA expression of Myc (relative to the Hprt housekeeping gene) after 6 h of RANKL stimulation (n = 3). (C) Immunoblot of nuclear protein lysates using phosphorylated c-Myc (S62), c-Myc, and Lamin B antibodies. Lamin B served as the loading control. Data are representative of three experiments. (D) Schematic diagram showing the primers (indicated by the black arrows) designed to detect un-spliced, premature Myc mRNA (pre-Myc). (E) Expression of pre-Myc after 6 h of RANKL stimulation (n = 6). (F,G) OCPs were stimulated with RANKL for 6 h and then treated with actinomycin D (10 μg/mL) for 0.25, 0.5, 1, and 3 h. (F) Percent expression of Myc after indicated h after actinomycin D treatment (n = 3). (G) Half-life of Myc transcript in WT and NRF2-deficient OCPs (n = 3). (H) Immunoblot of total cell protein lysates using p-ERK1/2, ERK1/2, p-JNK, p38, IκBα, and α-tubulin antibodies. α-tubulin served as the loading control. Data are representative of three experiments. All data are shown as mean ± s.e.m. * p < 0.05 and *** p < 0.001 using two-way ANOVA in (B,E); NS, not significant using two-tailed, unpaired t-test in (G).

Figure 3

Knockdown of NRF2 enhances RANKL-inducible MYC expression. Mouse OCPs were transfected with negative control (NC) or NRF2-specific small interfering RNAs (siRNAs) and stimulated with RANKL (50 ng/mL) for the indicated time points. (A) The expression of NRF2 was determined using immunoblot with nuclear lysates from transfected cells. Lamin B served as a loading control. Data are representative of two experiments from five mice. (B) The mRNA expression of Myc (relative to the Hprt housekeeping gene) at 19 h following RANKL stimulation (n = 5). (C) The expression of MYC using immunoblot with nuclear lysates at 24 h following RANKL stimulation. Lamin B served as a loading control. Data are representative of two experiments from five mice. (D) Signal intensity of the c-Myc immunoblot in C quantified using densitometry and normalized to Lamin B and to unstimulated NC control (n = 5). All data are shown as mean ± s.e.m. * p < 0.05 and ** p < 0.01 using two-way ANOVA.

Figure 4

CDDO-Im (1-[2-cyano-3-,12-dioxooleana-1,9(11)-dien-28-oyl] imidazole) treatment suppresses MYC expression and osteoclast differentiation. Mouse OCPs were pre-treated with either DMSO or the indicated doses of CDDO-Im for 30 min and then stimulated with RANKL (50 ng/mL). (A) Osteoclast differentiation of OCPs in the presence of DMSO or the indicated doses of CDDO-Im. Representative images of the TRAP-stained cells are shown. Scale bar: 50 μm. TRAP-positive, multinucleated (more than three nuclei) cells were counted in triplicates from three experiments. (B) Immunoblot of nuclear protein lysates using NRF2, c-Myc, and Lamin B antibodies after 24 h of RANKL stimulation. Lamin B served as the loading control. Data are representative of three experiments. (C) The mRNA expression of Myc, pre-Myc, Hmox1, and Gclm (relative to the Hprt housekeeping gene) after 6 h of RANKL stimulation (n = 3). All data are shown as mean ± s.e.m. * p < 0.05, ** p < 0.01, and **** p < 0.0001 using one-way ANOVA in (A,C).

Figure 5

MYC is required for enhanced osteoclastogenesis by NRF2 deficiency. Control (WT), NRF2-deficient (NRF2 KO), and MYC/NRF2-deficient (DKO) OCPs were stimulated with RANKL (50 ng/mL). (A) Immunoblot of nuclear protein lysates using NRF2, c-Myc, and Lamin B antibodies. Lamin B served as the loading control. Data are representative of three experiments. (B) Expressions of Myc and Nfe2l2 (relative to the Hprt housekeeping gene) after 6 h of RANKL stimulation (n = 3). (C,D) Osteoclast differentiation in WT, NRF2 KO, and MYC/NRF2 DKO OCPs. OCPs were cultured in the presence of RANKL for (C) 2 days or (D) 3 days, after which they were fixed and stained for TRAP. Representative images of the TRAP-stained cells are shown. Scale bar: 50 μm. TRAP-positive, multinucleated (more than three nuclei) cells were counted in triplicates from three experiments. (E) Immunoblot of nuclear protein lysates of WT, NRF2 KO, and MYC/NRF2 DKO OCPs stimulated with RANKL for the indicated time points using nuclear factor of activated T cells (NFATc1) and α-tubulin antibodies. α-Tubulin served as a loading control. Data are representative of three experiments. All data are shown as mean ± s.e.m. * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001 using two-way ANOVA in (A) and one-way ANOVA in (C,D); NS, not significant.

Figure 6

Myeloid-specific deletion of MYC decreases the osteoclast enhancing effect of NRF2 deficiency in vivo. (A–C) Micro-computed tomography (μCT) analysis of 12- to 13-week-old female WT, NRF2-deficient (NRF2 KO), and myeloid-specific MYC/NRF2-deficient (MYC^ΔM/NRF2 DKO) mice. (A) Representative μCT reconstructed images of the trabecular architecture of the distal femurs. Scale bar: 100 μm. (B) μCT measurements of the indicated parameters of the trabecular bone in the distal femurs. Bone volume/tissue volume ratio (BV/TV), trabecular numbers (Tb.N), trabecular thickness, (Tb.Th), and trabecular space (Tb.Sp) were computed using μCT analysis. (C) μCT measurements of the indicated parameters of the cortical bone in the midshaft of the femurs. BV/TV and porosity were computed using μCT analysis. Data are shown as mean mean ± standard deviation of at least seven mice per group. (D,E) Histomorphometric analysis of the trabecular bone in the distal femurs from 12- to 13-week-old female WT, NRF2 KO, and MYC^ΔM/NRF2 DKO mice. (D) Representative images showing the TRAP-positive, multinucleated osteoclasts (red-purple) in the coronal sections of the distal femur. Scale bar: 100 μm. (E) Histomorphometric analysis of the trabecular bone. Osteoclast surface area per bone surface (Oc.S/BS). Osteoclast number per bone surface (Oc.N/BS). Erosion over bone surface (ES/BS). All data are shown as mean ± s.e.m. of at least five mice per group. * p < 0.05 and ** p < 0.01 using one-way ANOVA in (B,E) except Tb.Th, which was analyzed using Kruskal‒Wallis test; NS, not significant.

Figure 7

Myeloid-specific deletion of MYC mitigates in vivo osteoclast formation and bone erosion of NRF2-deficient mice in mouse arthritis model. Arthritis was induced by the K/BxN serum transfer method in 8- to 9-week-old male WT, NRF2-deficient (NRF2 KO), myeloid-specific MYC/NRF2-deficient (MYC^ΔM/NRF2 DKO) mice. (A) Schematic timeline of the experiment design. The arthritis-inducing serum was injected into the mice intraperitoneally (IP) on day 0 and 2. (B) Time course of the clinical score and swelling of joints during the progression of arthritis. Data are shown as mean ± s.e.m. of at least five mice per group. * p < 0.05 between NRF2 KO and MYC^ΔM/NRF2 DKO groups using two-way ANOVA. (C) TRAP staining of histological sections of tarsal bones from WT, NRF2 KO, or MYC^ΔM/NRF2 DKO arthritic mice. Squares show enlarged images. Left scale bar: 400 μm. Right scale bar: 100 μm. (D) Histomorphometric analysis of the tarsal bones. Erosion over bone surface (ES/BS). Osteoclast surface area per bone surface (Oc.S/BS). Osteoclast number per bone surface (Oc.N/BS). Data are shown as mean ± s.e.m of at least five mice per group. * p < 0.05 using one-way ANOVA; NS, not significant.

Figure 8

Schematic diagram of the NRF2-regulated mechanism in osteoclastogenesis. NRF2 attenuates the RANKL-induced activation of p38 and ERK-1/2, which is crucial for MYC and NFATc1 expression.