Nrf2 signaling activation by a small molecule activator compound 16 inhibits hydrogen peroxide-induced oxidative injury and death in osteoblasts

We explored the potential activity of compound 16 (Cpd16), a novel small molecule Nrf2 activator, in hydrogen peroxide (H2O2)-stimulated osteoblasts. In the primary murine/human osteoblasts and MC3T3-E1 murine osteoblastic cells, Cpd16 treatment at micro-molar concentrations caused disassociation of Keap1-Nrf2 and Nrf2 cascade activation. Cpd16 induced stabilization of Nrf2 protein and its nuclear translocation, thereby increasing the antioxidant response elements (ARE) reporter activity and Nrf2 response genes transcription in murine and human osteoblasts. Significantly, Cpd16 mitigated oxidative injury in H2O2-stimulited osteoblasts. H2O2-provoked apoptosis as well as programmed necrosis in osteoblasts were significantly alleviated by the novel Nrf2 activator. Cpd16-induced Nrf2 activation and osteoblasts protection were stronger than other known Nrf2 activators. Dexamethasone- and nicotine-caused oxidative stress and death in osteoblasts were attenuated by Cpd16 as well. Cpd16-induced osteoblast cytoprotection was abolished by Nrf2 short hairpin RNA or knockout, but was mimicked by Keap1 knockout. Keap1 Cys151S mutation abolished Cpd16-induced Nrf2 cascade activation and osteoblasts protection against H2O2. Importantly, weekly Cpd16 administration largely ameliorated trabecular bone loss in ovariectomy mice. Together, Cpd16 alleviates H2O2-induced oxidative stress and death in osteoblasts by activating Nrf2 cascade.

Marcotte et al. have recently developed a small molecule inhibitor of Nrf2-Keap1 interaction named compound 16 (Cpd16, PubChem CID 1073725) [33]. Cpd16 binds directly to the Keap1's Kelch-DC domain at the C-terminus of Keap1 with the IC 50 of 2.7 µM [33]. It was able to increase Nrf2 response genes in cultured cells and acted as a novel Nrf2 activator [33]. Here our study reported that Cpd16 activated Nrf2 cascade and protected osteoblasts from H 2 O 2 .

RESULTS
Cpd16 activates Nrf2 signaling cascade in osteoblasts As shown, Cpd16 dose-dependently enhanced ARE luciferase reporter activity in murine osteoblasts (Fig. 1A). Further indicating Nrf2 cascade activation, NQO1 enzyme activity was remarkably increased after 1-25 μM of Cpd16 treatment (Fig. 1A). Cpd16 at 0.2 μM failed to significantly increase the NQO1 enzyme activity, showing the dose-dependent response (Fig. 1A). CCK-8 assay results found that Cpd16 (0.2-25 μM, for 24 h) failed to significantly decrease the viability in murine osteoblasts (Fig.  1A), suggesting that the compound is relatively safe to murine osteoblasts. At the two concentrations, 5 μM and 25 μM, Cpd16 robustly increased ARE reporter activity and the NQO1 enzyme activity in murine osteoblasts (Fig. 1A), they were selected for the following experiments.
Cpd16 ameliorates H 2 O 2 -provoked oxidative injury in osteoblasts By measuring the CellROX fluorescence intensity, we demonstrated that H 2 O 2 induced robust ROS production (CellROX intensity increase) in the primary murine osteoblasts ( Fig. 2A, B).

Cpd16 administration largely ameliorates trabecular bone loss in OVX mice
To examine the potential activity by Cpd16 in vivo, the mouse OVX model was utilized. The representative micro-CT images demonstrated that weekly intraperitoneal injection of Cpd16 (5 mg/kg) largely ameliorated trabecular bone loss in the OVX mice (Fig. 8A). The reductions of BV/TV (%, Fig. 8B) and BMD (Fig.  8C) in trabecular bones of OVX mice were largely alleviated with Cpd16 administration. Moreover, BMD of the cortical bones was also slightly decreased eight weeks after OVX (Fig. 8D), which was also inhibited following Cpd16 administration (Fig. 8D). Whether the antioxidant mechanism was activated by Cpd16 in vivo was determined. As shown, the SOD activity in the left tibias was significantly decreased in OVX group mice (8 weeks after OVX, Fig. 8E), while Cpd16 administration remarkably elevated it (Fig. 8E). These results showed that Cpd16 administration largely ameliorated oxidative injury and trabecular bone loss in OVX mice.

DISCUSSION
The transcription factor Nrf2 promotes the transcription and expression of a large number of antioxidant and/or defense genes, serving as a potential therapeutic target involved in the mitigation oxidative injury in osteoblasts [10,13,15,16,24,28,36]. Forced activation of Nrf2 signaling in osteoblasts/osteoblastic cells, using different agents or genetic strategies, was able to significantly inhibit oxidative injury by H 2 O 2 and a number of other oxidative stimuli [10,13,15,16,24,28,36].
Here in different osteoblasts, Cpd16 treatment at only micromolar concentrations induced disassociation of Keap1-Nrf2, stabilization of Nrf2 protein and following nuclear translocation, and enhanced ARE reporter activity as well as transcription of Nrf2 response genes (HO1, GCLC, and NQO1) in cultured osteoblasts/ osteoblastic cells. Significantly, Cpd16 ameliorated oxidative injury in H 2 O 2 -stimulated osteoblasts.
We found that Cpd16-induced Nrf2 activation and osteoblasts protection against H 2 O 2 were stronger than other known Nrf2 activators (SFH, 4-OI, and TBHQ). One possibility is that Cpd16 could induce Keap1 cysteine (151) alkylatation, leading to dramatic Keap1-Nrf2 disassociation and direct Nrf2 cascade activation. Indeed, we found that Keap1 Cys151S mutation abolished Cpd16-induced Nrf2 cascade activation and osteoblasts protection in primary human osteoblasts. The detailed mechanisms warrant further characterizations.
DEX can directly induce oxidative injury and osteoblast cell death, and it is a key factor for the progression of osteoporosis and osteonecrosis [60], which can be inhibited by Nrf2 activation [29,50,61]. Here DEX-caused oxidative injury and death in osteoblasts were largely attenuated by Cpd16. This novel Nrf2 small molecule activator should have promising value for the treatment of DEX-related bone injuries.
Sustained and/or high-dose nicotine exposure can significantly inhibit cell proliferation and differentiation in osteoblasts, and inhibit alkaline phosphatase (ALP) activity and collagen synthesis [51][52][53]. These changes together will eventually induce apoptosis, serving as the primary mechanism of cigarette smoke-related osteoporosis [51][52][53]. In the primary rat osteoblasts, nicotine was shown to inhibit multiple osteogenic and angiogenic genes [53]. We found that treatment with Cpd16 potently inhibited nicotine-induced oxidative injury and death of osteoblasts.
Osteoporosis seriously affects the life of the elderly people, especially postmenopausal women [62,63]. One key pathophysiological feature of osteoporosis is osteoblast dysfunction, resulting in decreased bone formation [62,63]. Oxidative stress-induced dysfunction and death of osteoblasts is the primary reason for the bone loss during the development of osteoporosis [64,65]. Therefore, reducing oxidative stress, i.e. using Nrf2 activators, can protect osteoblasts and inhibit their death, which has a promising effect on improving osteoporosis [64,65]. Here, Cpd16 inhibited H 2 O 2 -caused oxidative injury and death in cultured osteoblasts. The Nrf2 activator also largely ameliorated oxidative stress and trabecular bone loss in OVX mice. Therefore, it should have promising value for osteoporosis management.
Culture of primary murine/human osteoblasts and MC3T3-E1 murine osteoblastic cells As described previously [10,30], the trabecular bone fragments of written informed consent healthy donors were minced, washed, and digested. Thereafter, the primary human osteoblasts were obtained and cultivated in the described medium [30]. Medium was renewed twice a week. The primary murine osteoblasts were obtained and cultured as described [10,30]. The established MC3T3-E1 cells were provided by Dr. Zhou and cultivated as reported [50]. The protocols were with approval from the Ethics Board of Shanghai Jiao Tong University School of Medicine.

Genetic modifications in osteoblasts
For Nrf2 silencing, the lentiviral construct encoding short hairpin RNA (shRNA) sequence of Nrf2 [10,30] was transduced to the primary murine osteoblasts. Following puromycin-mediated selection, the stable osteoblasts were formed. CRISPR/Cas9-induced knockout (KO) of Keap1 or Nrf2 as well as the establishment of the single stable osteoblasts were described previously [10,30].

Keap1 mutation
The GV248 lentiviral Cys151S mutant Keap1 construct (no GFP) was from Dr. Liu at Jiangsu University [66] and was stably transduced to the osteoblasts. Cys151S Keap1 was checked by western blotting.
Other assays, including cell viability CCK-8 assay, the Caspase-3/-9 activity, the JC-1 fluorescence testing mitochondrial depolarization, the CellROX fluorescence staining of ROS, the cell necrosis assay by measuring medium LDH contents, the lipid peroxidation by measuring the reactive substances (TBAR) activity, NQO1 activity assay, ARE reporter activity assay, and single strand DNA (ssDNA) ELISA as well as qRT-PCR, coimmunoprecipitation (Co-IP), western blotting, Annexin V flow cytometry and nuclear TUNEL staining assays were reported in the previous studies [10,30]. Primers were provided by Dr. Jiang at Nanjing Medical University [67,68]. The uncropped blotting images were presented in Figure S1. Alternatively, the primary human osteoblasts were pretreated with Cpd16 (25 μM) for 2 h, followed by H 2 O 2 (400 μM) stimulation, and cell viability, apoptosis, and necrosis were tested by CCK-8 (D), TUNEL-nuclei staining (E), and LDH releasing F assays, respectively. # P < 0.05.

The murine ovariectomized (OVX) procedure and micro-CT analyses
The female C57/BL6 mice, at 7 weeks of age and 21-22 g of weight, were purchased from SLAC (Shanghai, China). Mice were anesthetized as described [69] and the detailed protocols of OVX were reported early [69]. Cpd16 (5 mg/kg) or PBS was intraperitoneally injected (i.p.) at the first day of each week. Micro-CT analyses were described in an early study [70]. In brief, the OVX mice and the control mice were scanned under the micro-CT equipment (Skyscan 1176, Belgium) 8 weeks after OVX procedure and the high-resolution scanogram were retrieved [70]. The dataset was reconstructed under a CT analyzer software and bone erosion was calculated using the in-house Fiji script [70]. The trabecular bone volume (BV) versus the total volume (TV), in %, was measured. The bone mineral density (BMD, g/cm 3 ) of trabecular bones and cortical bones was calculated as well [70]. After completion of micro-CT, the right tibias of the mice were collected and superoxide dismutase (SOD) activity in the fresh bone tissues was analyzed by a SOD ELISA kit (Thermo-Fisher Invitrogen, Shanghai, China). All animal experiments were conducted under protocols approved by IACUC of Soochow University.

Statistical analyses
Statistical analysis was described early [10,30]. P < 0.05 was considered as a statistically significant difference. Quantified values were mean ± standard deviation (SD). All in vitro experiments were repeated five times and similar results were obtained.

DATA AVAILABILITY
All data are available upon request. Fig. 8 Cpd16 administration largely ameliorates trabecular bone loss in OVX mice. The female C57/BL6 mice were subject to bilateral ovariectomy (OVX) procedure. Afterward, Cpd16 (at 5 mg/kg) or PBS were intraperitoneally injected (i.p.) at the first day of each week, and mice were sacrificed after 8 weeks. The representative micro-CT images of trabecular bones and cortical bones were presented (A). BV/TV (%, B) and bone mineral density (BMD, g/cm 3 , C) of trabecular bones were calculated. BMD of cortical bones was recorded as well (D). The relative SOD activity in the left tibia bone tissues in different groups was shown (E). The control group mice were orally administered with PBS ("Ctrl"). 10 mice per group. # P < 0.05.