Maqui Berry and Ginseng Extracts Reduce Cigarette Smoke-Induced Cell Injury in a 3D Bone Co-Culture Model

Cigarette smoking-induced oxidative stress has harmful effects on bone metabolism. Maqui berry extract (MBE) and ginseng extract (GE) are two naturally occurring antioxidants that have been shown to reduce oxidative stress. By using an osteoblast and osteoclast three-dimensional co-culture system, we investigated the effects of MBE and GE on bone cells exposed to cigarette smoke extract (CSE). The cell viability and function of the co-culture system were measured on day 14. Markers of bone cell differentiation and oxidative stress were evaluated at gene and protein levels on day 7. The results showed that exposure to CSE induced osteoporotic-like alterations in the co-culture system, while 1.5 µg/mL MBE and 50 µg/mL GE improved CSE-impaired osteoblast function and decreased CSE-induced osteoclast function. The molecular mechanism of MBE and GE in preventing CSE-induced bone cell damage is linked with the inhibition of the NF-κB signaling pathway and the activation of the Nrf2 signaling pathway. Therefore, MBE and GE can reduce CSE-induced detrimental effects on bone cells and, thus, prevent smoking-induced alterations in bone cell homeostasis. These two antioxidants are thus suitable supplements to support bone regeneration in smokers.


Introduction
Due to constant contact with the environment, free radicals are generated in the human body through factors such as respiration (oxidative reaction), external pollution, and radiation exposure [1]. Numerous human diseases, including cancer, metabolic bone disease, and aging, are closely associated with excessive production of free radicals [2,3]. Free radical-induced oxidative stress damage can be attenuated effectively by antioxidants [4]. Therefore, natural antioxidant intake is popular worldwide for the promotion of healthy life. However, there is a lack of information about the effects of several plant-based extracts with antioxidant properties on bone metabolism. To evaluate the influence of these promising extracts on bone homeostasis, we need an effective and stable model representing bone metabolism in vitro. Bone metabolism depends on the mutual regulation between osteoblasts and osteoclasts. The imbalance between bone formation and bone resorption impairs bone homeostasis and, consequently, decreases bone mineral content [5]. In a Under stress conditions, Nrf2 is liberated from the Keap1-Nrf2 complex and translocated to the nucleus, where it binds the antioxidant response element (ARE) sequences in the promoter regions on antioxidant genes [34]. Antioxidants, such as N-acetyl cysteine (NAC) and MBE, have been shown to protect bone-like cells against ROS and related cell injury by activating Nrf2 signaling [13,17].
Nuclear factor kappa-B (NF-κB) is a member of the pleiotropic transcriptional factor family and comprises NF-κB1 (P50), NF-κB2 (P52), RelA (P65), RelB, and c-Rel [35]. NF-κB plays an important role in bone metabolism, inflammation, and the immune response, while the action of NF-κB is regulated by the inhibitor of kappa B (IκB) [36]. The regulatory effect of the NF-κB signaling pathway on osteogenesis-and osteoclastogenesis-related gene expression is critical to maintain bone cell homeostasis [37]. In osteoclastogenesis, NF-κB pathways were the first known signaling pathways triggered by RANKL stimulation regulating osteoclast formation and function [38]. Animal experiments revealed that NF-κB1/NF-κB2 double knockout mice exhibited severe osteopetrosis due to the complete lack of mature osteoclasts [39], suggesting that activation of NF-κB is essential for osteoclast differentiation. In addition, silencing of IκB kinase (which enables NF-κB nuclear translocation) in mouse osteoblasts increased the bone mass with no effect on osteoclast number [40]. Therefore, inhibition of NF-κB is considered to both suppress bone resorption and promote bone formation.
Accordingly, by establishing an effective osteoblast and osteoclast 3D co-culture system in scaffolds, this research aimed to investigate the roles of MBE and GE in preventing smoking-induced bone cell damage and elucidate the associated molecular mechanism.

Chemical Reagents, Cell Culture Medium
All the chemicals were purchased either from Carl Roth (Karlsruhe, Germany) or Sigma-Aldrich (St. Louis, MO, USA). Cell culture mediums and other supplements were obtained from Sigma-Aldrich or Gibco (Thermo Fisher Scientific, Waltham, MA, USA).

The Preparation of GEL Scaffolds
The fabrication procedure has already been described in our previous research [7]. Briefly, for construction of the GEL scaffold, gelatin from cold-water fish skin (G7041, Sigma, St. Louis, MO, USA) and hydroxyapatite (21223, Sigma, St. Louis, MO, USA) were thoroughly mixed in a 50 mL falcon tube to achieve final concentrations of 4.8% and 10%, respectively. Following the addition of 1% glutaraldehyde (3778.1 Carl Roth, Karlsruhe, Germany), the mixture solution was immediately transferred into polystyrene casting molds and frozen overnight at −18 • C. After incubation at −80 • C for 1 h, the formed matrix was cut into a cylindrical shape (height: 3 mm; diameter: 6 mm) with the help of a razor blade.
To achieve optimal sterilization, GEL scaffolds were incubated with 70% ethanol overnight on a rotating shaker. Following washing with phosphate buffer saline (PBS; D8537, Sigma, St. Louis, MO, USA) three times, the scaffolds were incubated in culture medium as a sterile control.

The Preparation of Cigarette Smoke Extract (CSE) and Antioxidants
CSE was always prepared freshly on the day of the medium change. Through a peristaltic pump device, smoke from one commercial cigarette (Marlboro, New York, NY, USA) was continuously drawn and dissolved into a 25 mL plain culture medium (without FBS and osteogenic induction factors) [43]. The optical density of the CSE was normalized at 320 nm (OD 320 ), with an OD 320 of 0.65~0.7 considered 100% CSE [44]. After sterile filtration with a 0.22 µm pore filter, 100% CSE was diluted with osteogenic medium to a final concentration of 5%, which was equivalent to smoking 10 cigarettes per day [45]. Maqui berry extract (MBE), ginseng extract (GE), and N-acetylcysteine (NAC, 616-91-1, Carl Roth) were dissolved in plain culture media and sterilized with 0.22 µM pore filters. The stock concentrations of MBE, GE, and NAC were 1.5 mg/mL, 3 mg/mL, and 70 mM, respectively. At the time of use, antioxidant stock solutions were diluted with culture medium to the required concentrations. MBE and GE were produced according to the requirements of the Pharmacopoeia European monograph by Anklam Extrakt GmbH, Anklam, Germany.

Resazurin Conversion Assay
The resazurin conversion assay was used to measure the mitochondrial activity, which can reflect the number of viable cells on the scaffolds [13]. Before measurement, the scaffolds were washed with PBS and transferred to a new 48-well plate. Then, 500 µL of 0.0025% resazurin solution (199303-1G; Sigma) was added per well to cover the scaffolds. After incubation for 2 h at 37 • C, 100 µL/well solution was transferred into a 96-well plate and the fluorescence at 544 nm excitation and 590 nm emission wavelengths was measured with an Omega Plate Reader (BMG Labtech, Ortenberg, GER). The values of resazurin conversion from scaffolds without cells were subtracted as background.

Total DNA Isolation and Quantification
DNA was isolated using the alkaline lysis method. After being washed with PBS, 250 µL/well of 50 mM hot (98 • C) NaOH was added and incubated for 15 min. Then, scaffolds with NaOH solution were frozen at −80 • C overnight. The scaffolds were heated at 60 • C for 20 min until fully melted. An equal volume of 100 mM Tris buffer (pH = 8.0) was used to adjust pH values. A total of 100 µL of each sample was transferred into a V-bottom 96-well plate to remove impurities by centrifuging (1000× g, 10 min). DNA concentration was measured photometrically with an LVIS plate and Omega Plate Reader (wavelength λ = 230 nm, λ = 260 nm, and λ = 280 nm; 25 flashes) [42].

RT-PCR Analysis
Gene expression levels were determined by RT-PCR. Self-made TriFast (38% v/v phenol, 0.4 mM ammonium thiocyanate (221988, Sigma, St. Louis, MO, USA) 0.8 mM guanidine thiocyanate, 0.68 mM glycerol (55290, Otto Fischer GmbH, Waldkirch, Germany), and 0.1 M sodium acetate solution) was used to extract RNA from the scaffolds. To remove impurities, the solution was centrifuged three times at 14,000× g for 10 min before chloroform/phenol extraction. The RNA concentration was measured with the Omega Plate Reader. The cDNA synthesis kit (Thermo Fisher Scientific, Waltham, MA, USA) was used to synthesize cDNA. Then, PCR amplification was performed according to the manufacture instructions for Biozym Red HS Taq Master Mix (Vienna, Austria). Primer sequences and PCR conditions for target genes are shown in Table 1. 18S rRNA served as a housekeeping gene. A 1.8% (w/v) agarose gel with ethidium bromide was used to visualize the PCR products. The electrophoresis was carried out at 90 V for 60 min to separate the PCR products. Finally, the intensity of the bands was quantitatively analyzed by ImageJ (NIH, Bethesda, MD, USA) [47].

Western Blot
Protein expression levels were detected by Western blot (WB). Cells were lysed in 1× RIPA buffer containing protease/phosphatase inhibitors (1 µg/mL Pepstatin, 5 µg/mL Leupeptin, 1 mM phenylmethylsulfonyl fluoride, 5 mM NaF, and 1 mM Na 3 VO 4 ). Total protein lysates were then centrifuged at 14,000× g for 10 min at 4 • C to remove cell debris. The protein concentrations were determined with the micro-Lowry method [15]. As previously described [47], 25 µg of total protein was separated by 12% acrylamide gels (100 V, 180 min) and then transferred onto a nitrocellulose blotting membrane (100 mA, 180 min). Ponceaus staining was used to visually ensure adequate protein transfer. After being blocked with 5% BSA for 1 h, primary and secondary antibody incubations were performed as previously described for the dot blot (Section 2.12). The antibodies used are summarized in Table 3. The chemiluminescent signals in bands were visualized with a charge-coupled device camera and quantified in ImageJ (NIH, Bethesda, MD, USA).

Statistical Analysis
The data are presented as means ± the standard error of the mean (SEM). All the experiments were repeated at least three times with two or three technical replicates. Statistical analyses were performed using GraphPad Prism software (GraphPad Software 9.0, La Jolla, CA, USA). The data of the two groups were compared with the Mann-Whitney test. The data of multiple groups were compared with the non-parametric Kruskal-Wallis test, followed by Dunn's multiple comparison test. A two-way ANOVA test followed by Turkey's multiple comparisons was used when two independent variables were compared among groups. A p < 0.05 was considered statistically significant.

Cytotoxicity Tests of MBE, GE, and NAC
To select nontoxic concentrations of antioxidants for the 3D bone co-culture system, cells were treated with increasing concentrations of MBE (1.5, 3, 6, 12, and 60 µg/mL), GE (25, 50, 75, 100, and 200 µg/mL), or NAC (1, 3.5, 7, 14, and 28 mM) for 14 days. The mitochondrial activity and total DNA content were used to indicate cell viability. MBE at a concentration ≥ 12 µg/mL showed cytotoxic effects on the SCP-1/THP-1 co-culture system (Supplementary Figure S1a,b). A high concentration of GE (≥200 µg/mL) decreased mitochondrial activity significantly and decreased total DNA in the co-culture system (Supplementary Figure S1c,d). As for NAC, cell viability showed a remarkable reduction when the concentration exceeded 7 mM (Supplementary Figure S1e,f). Previous studies also reported that 1.5 µg/mL MBE [17] and 50 µg/mL GE [48] did not have cytotoxic effects in vitro. Therefore, based on our results and the literature, 1.5 µg/mL MBE and 50 µg/mL GE were used for subsequent experiments. Since our previous study demonstrated that 3.5 mM NAC had a protective role in CSE-induced bone cell injury [13], we used NAC as a positive control.

MBE and GE Suppressed Osteoclast Function in Bone Cells Exposed to CSE
To assess the effects of antioxidants on bone cells exposed to CSE, cell viability and functionality were measured after 14 days. The 3D co-culture system was exposed to 5% CSE in combination with 1.5 µg/mL MBE, 50 µg/mL GE, or 3.5 mM NAC. On day 14, co-cultures exposed to 5% CSE showed a significant reduction in mitochondrial activity (p < 0.0001; Figure 1a) and total DNA content (p < 0.0001; Figure 1b) relative to untreated cells. Co-incubation with MBE or GE slightly enhanced the viability (mitochondrial activity and total DNA content) of CSE-treated co-cultures on day 14. As expected, NAC caused a significant increase in mitochondrial activity (p < 0.05; Figure 1a) and total DNA content (p < 0.05; Figure 1b) in co-cultures treated with 5% CSE.
As well as cell viability, osteoclast function (CAII and TRAP activity) was measured in the co-culture systems. CAII and TRAP activities are early and late indicators of osteoclast differentiation, respectively [46]. Our CAII and TRAP activity results (Figure 1c,d) showed that osteoclast function was markedly increased in co-cultures exposed to 5% CSE, contributing to the development of an osteoporotic bone environment in vitro. Most interestingly, co-incubation with MBE or GE significantly reduced osteoclast function almost to control levels in bone cells exposed to CSE after 14 days (Figure 1c,d).
CSE in combination with 1.5 µg/mL MBE, 50 µg/mL GE, or 3.5 mM NAC. On day 14, cocultures exposed to 5% CSE showed a significant reduction in mitochondrial activity (p ˂ 0.0001; Figure 1a) and total DNA content (p ˂ 0.0001; Figure 1b) relative to untreated cells. Co-incubation with MBE or GE slightly enhanced the viability (mitochondrial activity and total DNA content) of CSE-treated co-cultures on day 14. As expected, NAC caused a significant increase in mitochondrial activity (p ˂ 0.05; Figure 1a) and total DNA content (p ˂ 0.05; Figure 1b) in co-cultures treated with 5% CSE. Figure 1. The effects of MBE and GE on SCP-1/THP-1 co-culture system exposed to CSE. The coculture system was co-incubated with 5% CSE and (or) antioxidants (1.5 µg/mL MBE, 50 µg/mL GE, and 3.5 mM NAC) for 14 days. (a) Mitochondrial activity of the co-culture system was measured on day 14. (b) Total DNA content of the co-culture system was measured on day 14. For osteoclast function, CAII activity (c) and TRAP activity (d) were compared after 14 days of co-culture. Statistical differences were determined using two-way ANOVA test followed by Turkey's multiple comparisons. Data are present as means ± SEM, and the significances are shown as * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001 vs. CSE group. (N = 3-4, n = 3).
As well as cell viability, osteoclast function (CAII and TRAP activity) was measured in the co-culture systems. CAII and TRAP activities are early and late indicators of osteoclast differentiation, respectively [46]. Our CAII and TRAP activity results ( Figure  1c,d) showed that osteoclast function was markedly increased in co-cultures exposed to 5% CSE, contributing to the development of an osteoporotic bone environment in vitro. Most interestingly, co-incubation with MBE or GE significantly reduced osteoclast function almost to control levels in bone cells exposed to CSE after 14 days (Figure 1c,d).
Absorbance intensity [fold of mitochondrial activity] Figure 1. The effects of MBE and GE on SCP-1/THP-1 co-culture system exposed to CSE. The coculture system was co-incubated with 5% CSE and (or) antioxidants (1.5 µg/mL MBE, 50 µg/mL GE, and 3.5 mM NAC) for 14 days. (a) Mitochondrial activity of the co-culture system was measured on day 14. (b) Total DNA content of the co-culture system was measured on day 14. For osteoclast function, CAII activity (c) and TRAP activity (d) were compared after 14 days of co-culture. Statistical differences were determined using two-way ANOVA test followed by Turkey's multiple comparisons. Data are present as means ± SEM, and the significances are shown as * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001 vs. CSE group. (N = 3-4, n = 3).

MBE and GE Enhanced Osteogenic Differentiation and Inhibited Osteoclastic Differentiation in Bone Cells Exposed to CSE
Expression levels of osteoblastic genes (Collagen 1 and Runx2) were investigated by RT-PCR. Collagen 1 provides structural support for bone cells but also forms a network to promote bone mineralization [49]. Runx2 is a master transcription factor of osteoblasts that can promote osteogenic differentiation of MSCs [50]. After 7 days, the expression of Collagen 1 and Runx2 decreased by 54.2% and 35.6% in cells exposed to 5% CSE, although those results were not statistically significant (Figure 2a-c). Following MBE, GE, or NAC treatment, Collagen 1 and Runx2 expression were significantly upregulated in bone cells exposed to CSE, reaching control levels (Figure 2a-c).
The osteoclastic genes (NFATc1, TRAP 5b, MMP-9) were also analyzed after 7 days of co-incubation with CSE and antioxidants. NFATc1 is an essential transcription factor in the terminal differentiation of osteoclasts [36]. TRAP 5b and MMP9 are required for the proteolytic activity of osteoclasts needed to resorb bone matrix [51]. These two genes are characteristic markers of osteoclast differentiation [51]. After 7 days of CSE exposure, the gene expression of NFATc1, TRAP 5b, and MMP-9 significantly increased (p < 0.01; Figure 2a,d-f). MBE or GE treatment showed a tendency to downregulate the expression of these osteoclastic marker genes in co-cultures exposed to CSE (Figure 2a,d,e). Similarly, osteoclastic marker gene expression was also downregulated by NAC treatment in bone cells exposed to CSE (Figure 2a,f).

MBE and GE Enhanced Osteoblast Function in Bone Cells Exposed to CSE
To evaluate the influence of MBE and GE on CSE-impaired osteogenesis, secreted AP, OCN, and PINP in the supernatant were detected by dot blot after culture of the system for 14 days. AP is produced in the early stages of osteogenic differentiation and can reflect osteoblastic activity [52]. OCN is a non-collagenous protein that is formed during the period of bone matrix mineralization [53]. Type I collagen is the major collagenous protein in the bone matrix [46]. Following the synthesis of new type I collagen, PINP is cleaved from type I procollagen by proteases outside the osteoblast. Thus, PINP could reflect the degree of bone mineralization [54]. After adding 5% CSE to the culture media, a significant decrease in AP levels (p < 0.05; Figure 3a,b), a tendency towards lower OCN levels (Figure 3a,c), and a significant decrease in PINP levels (p < 0.01; Figure 3a,d) were observed compared to untreated cells. MBE and GE treatment showed a tendency to upregulate these secreted osteoblast makers in co-cultures exposed to CSE (Figure 3a-d) after 14 days. Similar to MBE and GE, NAC treatment enhanced osteoblast markers in bone cells exposed to CSE (Figure 3a-d).

MBE and GE Enhanced Osteogenic Differentiation and Inhibited Osteoclastic Differentiation in Bone Cells Exposed to CSE
Expression levels of osteoblastic genes (Collagen 1 and Runx2) were investigated by RT-PCR. Collagen 1 provides structural support for bone cells but also forms a network to promote bone mineralization [49]. Runx2 is a master transcription factor of osteoblasts that can promote osteogenic differentiation of MSCs [50]. After 7 days, the expression of Collagen 1 and Runx2 decreased by 54.2% and 35.6% in cells exposed to 5% CSE, although those results were not statistically significant (Figure 2a-c). Following MBE, GE, or NAC treatment, Collagen 1 and Runx2 expression were significantly upregulated in bone cells exposed to CSE, reaching control levels (Figure 2a-c).
The osteoclastic genes (NFATc1, TRAP 5b, MMP-9) were also analyzed after 7 days of co-incubation with CSE and antioxidants. NFATc1 is an essential transcription factor in the terminal differentiation of osteoclasts [36]. TRAP 5b and MMP9 are required for the proteolytic activity of osteoclasts needed to resorb bone matrix [51]. These two genes are characteristic markers of osteoclast differentiation [51]. After 7 days of CSE exposure, the gene expression of NFATc1, TRAP 5b, and MMP-9 significantly increased (p ˂ 0.01; Figure  2a,d-f). MBE or GE treatment showed a tendency to downregulate the expression of these osteoclastic marker genes in co-cultures exposed to CSE (Figure 2a,d,e). Similarly, osteoclastic marker gene expression was also downregulated by NAC treatment in bone cells exposed to CSE (Figure 2a,f). The expression levels of osteoblastic and osteoclastic genes in the co-culture system under exposure to CSE and antioxidants. SCP-1/THP-1 co-culture system was exposed to 5% CSE with or without antioxidants for 7 days. The gene expression levels were evaluated by RT-PCR. 18s rRNA served as a housekeeper gene. (a) Representative RT-PCR image showed the gene expression of Collagen 1, Runx2, NFATc1, TRAP 5b, and MMP9 in 3D co-culture system with different treatments. Expression levels of Collagen 1 (b), Runx2 (c), NFATc1 (d), TRAP 5b (e), and MMP9 (f) mRNA were measured on day 7. Statistical differences were determined using the Kruskal-Wallis test followed by Dunn's multiple comparison test. Data are presented as means ± SEM, and the significance is shown as * p < 0.05, ** p < 0.01, and **** p < 0.0001 vs. CSE group. N = 3, n = 2.  Figure 2. The expression levels of osteoblastic and osteoclastic genes in the co-culture system under exposure to CSE and antioxidants. SCP-1/THP-1 co-culture system was exposed to 5% CSE with or without antioxidants for 7 days. The gene expression levels were evaluated by RT-PCR. 18s rRNA served as a housekeeper gene. (a) Representative RT-PCR image showed the gene expression of Collagen 1, Runx2, NFATc1, TRAP 5b, and MMP9 in 3D co-culture system with different treatments. Expression levels of Collagen 1 (b), Runx2 (c), NFATc1 (d), TRAP 5b (e), and MMP9 (f) mRNA were measured on day 7. Statistical differences were determined using the Kruskal-Wallis test followed by Dunn's multiple comparison test. Data are presented as means ± SEM, and the significance is shown as * p < 0.05, ** p < 0.01, and **** p < 0.0001 vs. CSE group. N = 3, n = 2.

MBE and GE Reduced CSE-Induced Cell Injury by Downregulation of sRANKL: OPG Ratio and NF-κB Signaling Pathways
Both sRANKL and OPG are secreted by osteoblasts. sRANKL is a critical cytokine for osteoclastogenesis and its deficiency prevents the formation of multinucleated osteoclasts from monocyte/macrophage lineage cells [55]. In contrast to sRANKL, OPG inhibits osteoclastic activity by decoying sRANKL [14]. Therefore, the ratio of sRANKL and OPG reflects the modulatory ability of osteoblasts towards osteoclasts. The supernatant of the co-culture system treated with CSE and antioxidants was collected on day 14 to measure the secreted protein levels. The results showed that the amount of sRANKL secreted rose in the CSE group and fell in the MBE, GE, and NAC groups ( Figure 4a). As for the expression levels of secreted OPG, an increasing trend was also found in the MBE, GE, and NAC groups (Figure 4a). Compared with the CSE group, a dramatic reduction in the ratio of sRANKL and OPG occurred in the MBE, GE, and NAC groups (Figure 4b). This result indicates that MBE and GE prevent CSE-induced bone cell damage by downregulating the sRANKL/OPG ratio.

MBE and GE Enhanced Osteoblast Function in Bone Cells Exposed to CSE
To evaluate the influence of MBE and GE on CSE-impaired osteogenesis, secreted AP, OCN, and PINP in the supernatant were detected by dot blot after culture of the system for 14 days. AP is produced in the early stages of osteogenic differentiation and can reflect osteoblastic activity [52]. OCN is a non-collagenous protein that is formed during the period of bone matrix mineralization [53]. Type I collagen is the major collagenous protein in the bone matrix [46]. Following the synthesis of new type I collagen, PINP is cleaved from type I procollagen by proteases outside the osteoblast. Thus, PINP could reflect the degree of bone mineralization [54]. After adding 5% CSE to the culture media, a significant decrease in AP levels (p ˂ 0.05; Figure 3a,b), a tendency towards lower OCN levels (Figure 3a,c), and a significant decrease in PINP levels (p ˂ 0.01; Figure 3a,d) were observed compared to untreated cells. MBE and GE treatment showed a tendency to upregulate these secreted osteoblast makers in co-cultures exposed to CSE (Figure 3a-d) after 14 days. Similar to MBE and GE, NAC treatment enhanced osteoblast markers in bone cells exposed to CSE (Figure 3a,b,c,d).  In osteoclastogenesis, NF-κB pathways were the first signaling pathways triggered by RANKL stimulation to be identified [38]. Therefore, we used WB and immunofluorescence to analyze NF-κB and p-ERK1/2 protein levels in the co-culture system on day 7. With the WB analysis, we could detect the total protein level of NF-κB in the co-culture system, while immunofluorescence staining showed the translocation of NF-κB in the nucleus. The WB results revealed that NF-κB total protein levels were similar in bone cells treated with antioxidants and/or CSE (Figure 4c,d). Interestingly, the protein expression of p-ERK1/2 was downregulated by MBE, GE, and NAC in comparison to CSE alone (Figure 4c,e). Immunofluorescence staining was further performed to investigate the translocation of NF-κB protein in the nucleus. Microscopy images showed that NF-κB protein was primarily expressed on osteoclasts rather than on osteoblasts (Supplementary Figure S2). Therefore, the translocation level of NF-κB protein in the nucleus was only analyzed for osteoclast-like cells. Compared with the absence of fluorescence signal in the control group, high expression of NF-κB protein was present after CSE exposure (Figure 4f). The addition of MBE, GE, and NAC significantly decreased the nuclear fluorescent signal intensity (Figure 4g) compared to CSE. These results indicate that MBE and GE decrease CSE-induced osteoclast differentiation by regulating the sRANKL/OPG ratio and NF-κB signaling pathways.

MBE and GE Prevented CSE-Induced Cell Injury by Activating Nrf 2 Signaling Pathway
It is well-known that CSE induces oxidative stress in bone cells and that pre-, post-, and co-incubation with antioxidants reduce bone cell stress damage [12,13,16,17]. To further explore the molecular mechanism of MBE and GE in preventing bone cell damage associated with CSE, the protein levels of p-Nrf2 and SOD1 were analyzed by Western blot. The transcription factor Nrf-2 activates the antioxidant defense system by regulating the antioxidative enzyme SOD1. After 7 days' differentiation of the system with antioxidants and (or) CSE, total protein was extracted from the co-culture systems. CSE exposure mildly increased the phosphorylation of Nrf2 and the level of SOD1 (Figure 5a). MBE and GE upregulated the expression of the activated Nrf2 and its target protein SOD1 by two-and threefold, respectively, compared to untreated systems (Figure 5b,c). As expected, NAC increased the protein levels of p-Nrf2 and SOD1 in response to CSEassociated oxidative stress. These results indicate that MBE and GE decrease CSE-induced bone cell damage by regulating Nrf2 signaling pathways. The molecular mechanism by which MBE and GE prevent CSE-induced bone cell damage is summarized in Figure 6. (g) The NF-κB protein level in the nucleus of the osteoclast. Statistical differences were determined using the Kruskal-Wallis test followed by Dunn's multiple comparison test. Data are presented as means ± SEM, and the significance is shown as * p < 0.05, ** p < 0.01 and **** p < 0.0001 vs. CSE group.

MBE and GE Prevented CSE-Induced Cell Injury by Activating Nrf 2 Signaling Pathway
It is well-known that CSE induces oxidative stress in bone cells and that pre-, post-, and co-incubation with antioxidants reduce bone cell stress damage [12,13,16,17]. To further explore the molecular mechanism of MBE and GE in preventing bone cell damage associated with CSE, the protein levels of p-Nrf2 and SOD1 were analyzed by Western blot. The transcription factor Nrf-2 activates the antioxidant defense system by regulating the antioxidative enzyme SOD1. After 7 days' differentiation of the system with antioxidants and (or) CSE, total protein was extracted from the co-culture systems. CSE exposure mildly increased the phosphorylation of Nrf2 and the level of SOD1 (Figure 5a). MBE and GE upregulated the expression of the activated Nrf2 and its target protein SOD1 by two-and threefold, respectively, compared to untreated systems (Figure 5b,c). As expected, NAC increased the protein levels of p-Nrf2 and SOD1 in response to CSE-associated oxidative stress. These results indicate that MBE and GE decrease CSE-induced bone cell damage by regulating Nrf2 signaling pathways. The molecular mechanism by which MBE and GE prevent CSE-induced bone cell damage is summarized in Figure 6.   complexes prevents the degradation of transcription factor Nrf2 and, consequently, results in the activation of Nrf2. Combined with antioxidant response element (ARE) promoter, phosphorylated Nrf2 (p-Nrf2) triggers the antioxidative transcription of SOD1. Then, the antioxidative defense system enhances osteogenesis in the SCP-1/THP-1 co-culture system. However, the slight activation of the Nrf2 signaling pathway by CSE could not inhibit the oxidative damage in osteoblasts. This led to decreased OPG secretion by osteoblasts and, thus, an increase in the RANKL: OPG ratio. Through the interaction between RANKL and RANK (RANKL receptor), increased RANKL activated the NF-κB pathway. In addition, CSE also upregulated the NF-κB pathway in the co-culture system. CSE promotes the expression of phosphorylation ERK 1/2 via the production of reactive oxygen species (ROS). ROS-induced inhibitor of kappa B kinase beta (IKKβ) results in the degradation of IκB-α and the activation of NF-κB. Transcriptional factor NF-κB promoted the expression of osteoclastrelated genes and osteoclastogenesis. The black and green arrows represent the activity of CSE and antioxidants (MBE or GE), respectively.

Discussion
External stimuli (such as drugs, ultraviolet light, cigarette smoke, and ionizing radiation) and endogenous free radicals can directly or indirectly damage cellular components, such as proteins, lipids, and DNA [56]. To defend against these adverse effects, the body activates a complex oxidative stress response system to mitigate cellular damage. As a key transcription factor regulating antioxidant stress, Nrf2 plays an important role in inducing the cellular antioxidant response [57]. Keap1-Nrf2 is a classical regulatory pathway that controls oxidative stress levels. Keap1 is a substrate adaptor protein of the E3 ubiquitin ligase complex that can assemble with Cullin3 (Cul3) into a functional E3 ubiquitin ligase complex (Keap1-Cul3-E3) [58]. In homeostatic conditions, the Nrf-2-Keap1-Cul3-E3 complex is degraded by the proteasome system. When cells are exposed to oxidative stress, electrophilic stress induces a conformational change in the Keap1-Cul3-E3 ubiquitin ligase. This conformational change results in the inhibition of Nrf2 ubiquitylation and disruption of the Keap1-Nrf2 complex [59]. Following release, Nrf2 is activated by phosphorylation and binds with the antioxidant response element in the nucleus, leading to the activation of cytoprotective genes, such as superoxide dismutase (SOD), catalase (CAT), glutathione reductase (GR), and glutathione peroxidase (GPX) [60,61]. Our results showed that CSE with and without antioxidants activated the Nrf2-mediated antioxidant system. However, CSE showed negative effects on osteoblast survival and differentiation and, thus, the levels of Nrf2 and SOD1 after CSE exposure may not efficiently reduce the oxidative stress induced by CSE. Therefore, high oxidative stress-induced cytotoxicity inhibits osteogenic differentiation and mineralization. In contrast, MBE and GE promoted osteogenic differentiation and function by activating the Nrf2-mediated antioxidant system. We speculate that this led to a reduction in the expression of RANKL, an increase in the expression of OPG, and, thus, a drop in the ratio of RANKL/OPG.
In normal bone tissue, old bone is continuously destroyed by osteoclasts, while osteoblasts generate new bone to maintain bone homeostasis [62]. The interaction of osteoblasts and osteoclasts is involved in maintaining the integrity and mechanical strength of bone tissue [63]. Coupled bone cells contribute to osteogenesis and osteoclastogenesis through direct contact and indirectly through cytokine secretion [64]. First, gap junctions between osteoblasts and osteoclasts affect their growth, proliferation, differentiation, and migration. Second, osteoclasts can enhance osteoblast differentiation by releasing transforming growth factor-β (TGF-β), bone morphogenetic proteins (BMPs), and insulin-like growth factors (IGFs) [65]. Similarly, M-CSF and RANKL secreted by osteoblasts can activate osteoclast precursors and promote them to differentiate into mature osteoclasts [66]. By producing OPG and RANKL, osteoblasts regulate osteoclastogenesis and osteoclast function. OPG is an important receptor for RANKL, which inhibits the maturation of osteoclasts [67]. Overexpression of OPG inhibits the production of osteoclasts, leading to osteosclerosis, whereas deficiency of OPG enhances bone resorption and leads to osteoporosis [68]. Therefore, RANKL and OPG can affect bone resorption and bone density by regulating the generation and function of osteoclasts. The increased osteoblast function in MBE and GE groups could inhibit osteoclast function via the RANKL/OPG ratio. In addition, we noticed that MMP9 gene expression was lower in the NAC group than in the MBE and GE groups. As a strong oxidant scavenger, the inhibitory effect of NAC on MMP9 was also reported in myofibroblasts in previous research and attributed to docking to Zn 2+ at the active site [69].
Since NF-κB is also a key molecule for osteoclastogenesis, we further analyzed the expression of the NF-κB pathway in the co-culture system. During osteoclast differentiation, NF-κB is the first signaling pathway activated by RANKL [70]. NF-κB knockout mice showed severe ossification due to the complete lack of osteoclasts, suggesting that the activation of NF-κB is essential for osteoclast differentiation [40]. In addition, NF-κB is an oxidative stress-sensitive transcription factor that is directly activated by ROS [71]. On the other hand, NF-κB also induces oxidative stress, indicating that there is a vicious cycle between NF-κB and oxidative stress [72]. Interestingly, we found that there were no differences in total NF-κB protein among the different groups. However, when we focused on the proteins in the nucleus, NF-κB protein expression was strongly decreased in the antioxidant group (MBE and GE) compared to the CSE group. These results show that although antioxidants do not influence the total protein NF-κB levels, antioxidants promote the nuclear translocation of the transcriptional factor. Overall, CSE indirectly regulates the expression of NF-κB through RANKL and directly activates the NF-κB signaling pathway through ROS, thereby promoting the expression of osteoclast-related genes, while MBE and GE can downregulate the NF-κB signaling pathway by reducing oxidative stress and enhancing osteoblast activity. Nagaoka and colleagues demonstrated that MBE also downregulated the nuclear translocation of NF-κB in MC3T3-E1 cells treated with lipopolysaccharide [73]. Additionally, several studies showed that GE also inhibited NF-κB nuclear translocation associated with inflammation in cancer cell lines [74,75].
Molecular interactions between the NF-κB and Nrf2 pathways have been reported [36]. Aside from its regulatory effect in the Nrf2 pathway, Keap1 has been confirmed to regulate the NF-κB signaling pathway via IKKβ [58]. IKKβ is an important component of the IKK kinase complex in the classical activation pathway of NF-κB. The enhanced activity of IKKβ could effectively upregulate the NF-κB signaling pathway [70]. Through binding with IKKβ, Keap1 induced the ubiquitination and degradation of IKKβ, thereby negatively regulating the activity of NF-κB. There is some evidence that Keap1 is involved in the functional interaction between the Nrf2 and NF-κB pathways [33,76]. Knockdown of the Keap1 gene in human umbilical vein endothelial cells not only upregulated the Nrf2-mediated antioxidant system but also suppressed the activation of the NF-κB signaling pathway [76]. Keap1 could be a key regulator of the interactions controlling cellular redox status and responses to stress. However, the interactions between Nrf2 and the NF-κB signaling pathway in the osteoblast and osteoclast co-culture system need to be further confirmed.
Delphinol ® is a standardized commercial extract from maqui berry. It contains a minimum of 25% delphinidins and 35% total anthocyanins. GE is extracted from ginseng roots, which consist of 10% total ginsenosides (natural glycosylated triterpenes also known as saponins). Those active components contained in MBE or GE have been described as possessing antioxidant properties and protective effects in several systems [46][47][48][49]. Therefore, we can hypothesize that the positive effects on bone cells were related to the active components of the extracts used in the study.
Several in vitro and in vivo studies have demonstrated that NAC reduces extracellular cystine to cysteine, serving to support intracellular GSH biosynthesis and provide sulphydryl groups that enhance glutathione synthesis and glutathione-S-transferase activity [77,78]. Moreover, NAC could also directly scavenge hydroxyl radicals through thiol groups. However, thiol radicals could be generated by the interaction of NAC with reactive radicals, implying a pro-oxidant role for NAC. Generation of thiyl radicals can happen in several ways and is also influenced by different factors, such as oxygen, the presence of metals, and the origin of the sulfur component; therefore, the context could influence the antioxidant or pro-oxidant properties of NAC [79]. During CSE treatment, bone cells increase ROS production, and this oxidative stress context may induce NAC to protect cells from free radical damage. However, under physiological oxidative stress conditions, NAC can also have negative effects on bone cells due to pro-oxidative properties. Additionally, it is known that a lower amount of free radicals is necessary to induce bone progenitor cell osteogenic differentiation; therefore, reducing bone cells from free radicals in physiological conditions could negatively influence bone homeostasis [80].
Some limitations of this study need to be recognized. Bone healing is a dynamic process that is influenced by various factors, including other organs, vascular systems, and biochemical signals; our scaffold could not simulate all those biological conditions observed in vivo. Additionally, although SCP-1 and THP-1 cell lines stably proliferate and display the physiological function of bone cells, there are still some differences in gene expression and cell function between the cell line and primary cells. In our study, we explored the effects of MBE or GE in combination with CSE; however, the effect of the co-administration of MBE and GE on CSE-induced bone cell damage remains to be investigated. We observed that the therapeutic effects of antioxidants (MBE and GE) were more pronounced regarding functionality as opposed to viability. Similar results were achieved with NAC. Further investigations will focus on whether there are other mechanisms by which antioxidants reduce CSE-induced cell damage; for example, involving the BMP signaling pathway, Wnt/β-catenin signaling pathway, MAPK signaling pathway, and TGF-β signaling pathway. Although the CSE used in the study was an aqueous extract that better mimics the absorption of molecular species into the blood in smokers' lungs and was sterile-filtered before application to the system such that particles with a size >0.2 µm were removed (comparable to the barrier function in the lung), our system could not represent the exposure of bone tissue in smokers. The addition of endothelial cells, immune cells, lung cells, or even liver cells would improve the model and better represent the molecular species or additional metabolites that are transported into the bloodstream and arrive at the bone tissue.

Conclusions
By activating the Nrf2 signaling pathway and reducing the RANKL: OPG ratio, MBE and GE can improve CSE-impaired osteogenesis and decrease CSE-induced osteoclastogenesis, thus preventing the disruption of bone homeostasis. MBE and GE mitigated CSE damage to bone cells to an extent comparable to NAC (precursor of glutathione). Therefore, MBE and GE show promise as potential bone health supplements for orthopedic patients who smoke.