Bergamot Polyphenol Fraction Exerts Effects on Bone Biology by Activating ERK 1/2 and Wnt/β-Catenin Pathway and Regulating Bone Biomarkers in Bone Cell Cultures

Epidemiological studies show that fruit consumption may modulate bone mineral density. However, data regarding the effect of the Citrus bergamia Risso (Bergamot orange), a citrus fruit containing a high concentration of flavonoids, on bone health are still lacking. In this study, we investigated the effects of Bergamot polyphenols on the Wnt/β-catenin pathway in two distinct bone cell types (Saos-2 and MG63). Findings showed that exposure to 0.01 and 0.1 mg/mL doses upregulate β-catenin expression (p = 0.001), osteoblast differentiation markers (e.g., RUNX2 and COL1A), and downregulate RANKL (p = 0.028), as compared to the control. Our results highlight, for the first time, that Bergamot polyphenols act on bone cells through the β-catenin pathway. In vivo studies are necessary to fully understand Bergamot’s role against bone resorption.


Introduction
Despite the wide range of pharmacological agents that reduce fracture risks [1][2][3][4][5][6][7][8], safety concerns have arisen regarding long-term osteoporosis pharmacotherapy. The long-term use of bisphosphonates is associated with an increased risk of atypical fractures [9], upper gastrointestinal adverse events [10], and osteonecrosis of the jaw (ONJ) [11]. A recent report has demonstrated that denosumab can also induce ONJ [12]. The use of hormone therapy (HT) for the prevention of osteoporosis in elderly postmenopausal women is currently under review, due to the increased risk of breast cancer and thromboembolic disease [13,14]. Because of the occurrence of osteosarcoma in toxicity studies, the use of teriparatide is limited to 24 months of treatment [15]. Finally, strontium ranelate has been associated with rare cases of hypersensitivity drug reaction with eosinophilia and systemic symptoms (DRESS) [16]. There is general consensus that the benefits of treatment far outweigh any risks associated with long-term treatment [17], but it is evident that there is an urgent need to find new and cost-efficient ways to prevent the development of osteoporosis.
Fruits and vegetables are rich in a variety of bioactive compounds, with antioxidant properties, that may effectively improve bone health [18]. Citrus fruits are important due to their high polyphenol content. However, to date, the effects of polyphenols, as nutraceuticals, on bone cell metabolism have scarcely been studied. Additionally, little research has focused on the potential effects of the polyphenols from Citrus bergamia Risso (Bergamot orange) on bone. Accordingly, we investigated the role of the Bergamot polyphenol fraction (BPF) in vitro on the intracellular pathways, cell proliferation, differentiation, and expression of the bone extracellular matrix proteins, in two distinct bone cell models (Saos-2 and MG63). Because the Wnt/β-catenin pathway is a key target for developing drugs against bone loss [19], we mainly focused on verifying whether BPF acts at this level.

Western Blotting and β-Catenin Knockdown
Saos-2 cells were seeded at a density of 200,000 cells/well in 6-well dishes, and 500,000 cells/well in 100 mm culture dishes. MG63 cells were seeded at a density of 100,000 cells/well in 6-well dishes. Cells were grown in serum-free medium and incubated with BPF 0.001; 0.001; 0.1 mg/mL for 10 min or 24 h. BPF, as previously prepared and characterized for polyphenol content [20], was provided by Herbal and Antioxidant Derivatives srl. (Polistena, RC, Italy). The main flavonoids identified in BPF were neoeriocitrin (370 ppm), naringin (520 ppm), and neohesperidin (310 ppm).
Moreover, Saos-2 cells were transfected with β-catenin small interfering RNA (siRNA) or with negative control (scramble) siRNA (Ambion-Life Technologies by Thermo Fisher Scientific, Rockford, IL, USA), with Lipofectamine 3000 Reagent, according to the manufacturer's instructions. The optimum concentration for silencing β-catenin expression was 10 nM. Maximum reduction of catenin expression occurred at 48 h post-transfection.

Real Time-PCR
Saos-2 cells were seeded at a density of 200,000 cells/well in 6-well dishes, and 500,000 cells/well in 100 mm culture dishes. Cells were grown in serum-free medium and incubated with BPF 0.001; 0.001; 0.1 mg/mL for 24 h. Total RNA of cells were extracted with Trizol reagent (Life technologies, UK) according to manufacturer's instructions. cDNA was synthesized from 1 µg total RNA, using High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA). mRNA expression of RANKL, osteoprotegerin, RUNX2, type I collagen, and β-actin were quantified by real time-PCR using SYBR ® Green dye (SYBR ® Green PCR Master Mix, Applied Biosystems, Foster City, CA, USA) (see Table 1).

ALP Activity
Saos-2 cells were seeded at a density of 50,000 cells/well in 24-well dishes. Cells were cultured in osteogenic medium, supplemented with b-glycerophosphate 7.5 mM (SERVA, Heidelberg, Germany), ascorbic acid 50 mg/mL (Santa Cruz, CA, USA) and dexamethasone 10 nM, and incubated with BPF 0.001; 0.001; 0.1 mg/mL for 24 h. Cells were lysed with ice cold 50 mM Tris-HCl solution with 0.05% Triton X-100. Cells were then centrifuged at 1000 g, 4 • C for 15 min. Protein concentration was determined using Bradford assay, and ALP activity was determined by p-nitrophenyl phosphate (pNPP) colorimetric method (WAKO Chemicals USA, Richmond, VA, USA).

Cell Viability Assay
To evaluate cell viability, Saos-2 cells were seeded at a density of 10,000 cells/well in 96-well plates. Cells were grown in serum-free medium and incubated with BPF 0.001; 0.001; 0.1 mg/mL for 24 h. Cell viability was determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Briefly, MTT (Sigma, St. Louis, MO, USA) solution (5 mg/mL) was added to each well, and incubated at 37 • C for 4 h. The supernatants were then removed and replaced by 100 mL DMSO. The optical density (OD) was measured at a wavelength of 570 nm.

Cell Proliferation
Saos-2 and MG63 cells were seeded at a density 200,000 and 100,000 cells/well, respectively, in 6-well dishes. Cells were grown in serum-free medium, and incubated with BPF 0.001; 0.001; 0.1 mg/mL for 24 h. Then, cell number was determined using the Nucleo counter NC-100 (Chemometec A/S, Lillerød, Denmark).

Statistical Analysis
Data are represented as mean ± standard deviation of at least three independent experiments, and analyzed using two-tailed Student's t-test and linear regression. p-values less than 0.05 were considered significant. Statistical analysis was performed with GraphPad Prism 5.0.

BPF Increases β-Catenin, RUNX2 and Intracellular COL1A Proteins and Decreases Both the RANKL and Extracellular COL1A Protein Levels in Saos-2
Osteoblastic Saos2 cells were exposed to 0.001, 0.01 and 0.1 mg/mL BPF. Only the 0.1 mg/mL dose increased RUNX2 mRNA levels, compared to the control (p = 0.031; Figure 2B).

BPF Increases β-Catenin, RUNX2 and Intracellular COL1A Proteins and Decreases Both the RANKL and Extracellular COL1A Protein Levels in Saos-2
Osteoblastic Saos2 cells were exposed to 0.001, 0.01 and 0.1 mg/mL BPF. Only the 0.1 mg/mL dose increased RUNX2 mRNA levels, compared to the control (p = 0.031; Figure 2B).
RANKL mRNA levels decreased and COL1A mRNA levels increased at 0.01 mg/mL, compared to the control (p = 0.0007 and p = 0.016, respectively; Figure 2C,D). In addition, RANKL mRNA levels decreased and COL1A mRNA did not change at 0.1 mg/mL, compared to the control (p = 0.006) ( Figure 2C,D).
Exposure to increasing doses of BPF resulted in higher protein levels of β-catenin and RUNX2 (p = 0.001 and p = 0.039, respectively), and lower levels of RANKL (p = 0.028), than the control ( Figure 3A-C, respectively). RANKL mRNA levels decreased and COL1A mRNA levels increased at 0.01 mg/mL, compared to the control (p = 0.0007 and p = 0.016, respectively; Figure 2C,D). In addition, RANKL mRNA levels decreased and COL1A mRNA did not change at 0.1 mg/mL, compared to the control (p = 0.006) ( Figure 2C,D).
Exposure to increasing doses of BPF resulted in higher protein levels of β-catenin and RUNX2 (p = 0.001 and p = 0.039, respectively), and lower levels of RANKL (p = 0.028), than the control ( Figure  3A-C, respectively).  Figure 4A shows that intracellular COL1A production increased at the 0.1 mg/mL dose, compared to the control (t-test, p = 0.038). Whereas Figure 4B shows that the extracellular COL1A secretion was markedly suppressed in a dose dependent manner (up to 45%; p = 0.011).
Osteoprotegerin mRNA, ALP activity, and their proteins, did not change after stimulation, with all concentrations of BPF (Supplementary Figures S1A,B and S2A,B).

RNA Interference Shows that RUNX2 and RANKL Expression Is Mediated by β-Catenin
We next attempted to confirm the role of BPF on β-catenin, by silencing its expression using siRNA, and assessed its effects on RUNX2 and RANKL expression in vitro.
siRNA effectively suppressed β-catenin protein level (p = 0.005), whereas the scrambled siRNA did not influence the protein level ( Figure 5).
As expected, inhibition of β-catenin by siRNA, in the presence of BPF, reverted the effects on both the RUNX2 and RANKL protein expression, compared to the control cells transfected with a scrambled siRNA, where RUNX2 increased and RANKL decreased ( Figure 6A,B).  Figure 4A shows that intracellular COL1A production increased at the 0.1 mg/mL dose, compared to the control (t-test, p = 0.038). Whereas Figure 4B shows that the extracellular COL1A secretion was markedly suppressed in a dose dependent manner (up to 45%; p = 0.011).
Osteoprotegerin mRNA, ALP activity, and their proteins, did not change after stimulation, with all concentrations of BPF (Supplementary Figures S1A,B and S2A,B).

RNA Interference Shows that RUNX2 and RANKL Expression Is Mediated by β-Catenin
We next attempted to confirm the role of BPF on β-catenin, by silencing its expression using siRNA, and assessed its effects on RUNX2 and RANKL expression in vitro.
siRNA effectively suppressed β-catenin protein level (p = 0.005), whereas the scrambled siRNA did not influence the protein level ( Figure 5).      As expected, inhibition of β-catenin by siRNA, in the presence of BPF, reverted the effects on both the RUNX2 and RANKL protein expression, compared to the control cells transfected with a scrambled siRNA, where RUNX2 increased and RANKL decreased ( Figure 6A,B).

BPF Increases β-Catenin Protein Levels in MG63.
Since the MG63 cell line provides a useful model for studying events associated with the early osteoblastic differentiation stage [21], MG63 cells were incubated with different doses of BPF for 24 h. Then cell proliferation was determined by counting the cell number. There were no differences in the cell proliferation at all doses of BPF (Supplementary Figure S3), whereas the β-catenin protein expression significantly increased, compared to the control untreated cells (p = 0.012; Figure 7).

BPF Increases β-Catenin Protein Levels in MG63.
Since the MG63 cell line provides a useful model for studying events associated with the early osteoblastic differentiation stage [21], MG63 cells were incubated with different doses of BPF for 24 h. Then cell proliferation was determined by counting the cell number. There were no differences in the cell proliferation at all doses of BPF (Supplementary Figure S3), whereas the β-catenin protein expression significantly increased, compared to the control untreated cells (p = 0.012; Figure 7).

Discussion
This study, which is the first to directly investigate the effects of Bergamot polyphenols in bone cell models, shows that β-catenin, ERK 1/2, RUNX2, and intracellular COL1A proteins are upregulated by exposure to increasing doses of BPF. In addition, the expression of RANKL, which is a known stimuli for increasing osteoclast activity, is reduced in these cells, suggesting that BPF downregulates RANKL, and may have a potential role in bone resorption.
To identify potential differences in β-catenin expression during cell maturation, we performed our experiments in two distinct bone cell types. Saos-2 cells represent a valuable model for studying events associated with the late osteoblastic differentiation stage in human cells [21], whereas MG63 cells are associated with the early stage in human cells. We found that BPF induced β-catenin expression in both MG63 and Saos-2 cells. This is one of the key findings of this study.
It is known that a high fruit and vegetable intake is associated with a lower risk of osteoporosis [18,[22][23][24]. However, to date, the effects of citrus fruit polyphenols on bone cell metabolism have scarcely been studied. In one study, hesperidin, which is a citrus flavonoid, helped to inhibit bone loss of ovariectomized mice [25]. In other studies, because of their flavonoid content, orange and grapefruit juices positively affected serum antioxidant capacity, bone strength, and the fracture risk of ovariectomized rats [26,27]. Bergamot (Citrus bergamia Risso) differs from other citrus fruits because of its particularly high flavonoid concentration [28]. Bergamot is an endemic plant of Calabria, in southern Italy, with a unique profile of flavonoid and flavonoid glycosides in its juice, such as neoeriocitrin, neohesperidin, naringin, rutin, neodesmin, rhoifolin and poncirin. It has been demonstrated that BPF, with its antioxidant properties, has pleiotropic beneficial effects in preventing and reducing heart muscle damage [29]. Since oxidative stress is involved in the pathogenesis of osteoporosis, we investigated the effects of BPF on bone cell metabolic markers in vitro, by two distinct osteoblast models.

Discussion
This study, which is the first to directly investigate the effects of Bergamot polyphenols in bone cell models, shows that β-catenin, ERK 1/2, RUNX2, and intracellular COL1A proteins are upregulated by exposure to increasing doses of BPF. In addition, the expression of RANKL, which is a known stimuli for increasing osteoclast activity, is reduced in these cells, suggesting that BPF down-regulates RANKL, and may have a potential role in bone resorption.
To identify potential differences in β-catenin expression during cell maturation, we performed our experiments in two distinct bone cell types. Saos-2 cells represent a valuable model for studying events associated with the late osteoblastic differentiation stage in human cells [21], whereas MG63 cells are associated with the early stage in human cells. We found that BPF induced β-catenin expression in both MG63 and Saos-2 cells. This is one of the key findings of this study.
It is known that a high fruit and vegetable intake is associated with a lower risk of osteoporosis [18,[22][23][24]. However, to date, the effects of citrus fruit polyphenols on bone cell metabolism have scarcely been studied. In one study, hesperidin, which is a citrus flavonoid, helped to inhibit bone loss of ovariectomized mice [25]. In other studies, because of their flavonoid content, orange and grapefruit juices positively affected serum antioxidant capacity, bone strength, and the fracture risk of ovariectomized rats [26,27]. Bergamot (Citrus bergamia Risso) differs from other citrus fruits because of its particularly high flavonoid concentration [28]. Bergamot is an endemic plant of Calabria, in southern Italy, with a unique profile of flavonoid and flavonoid glycosides in its juice, such as neoeriocitrin, neohesperidin, naringin, rutin, neodesmin, rhoifolin and poncirin. It has been demonstrated that BPF, with its antioxidant properties, has pleiotropic beneficial effects in preventing and reducing heart muscle damage [29]. Since oxidative stress is involved in the pathogenesis of osteoporosis, we investigated the effects of BPF on bone cell metabolic markers in vitro, by two distinct osteoblast models.
In bone, the extracellular matrix (ECM) plays an important role in osteoblast function. It is known that ECM-integrin interaction leads to the activation of the MAPKs, ERK1 and ERK2, resulting in increased RUNX2 phosphorylation, and consequently the expression of osteoblast differentiation genes, such as collagen, osteocalcin, osteopontin and bone sialoprotein [30,31]. The β-catenin signaling system functions as a transcriptional activator, and plays a crucial role in stabilizing intercellular adhaerens junctions. It has been demonstrated, in one animal model, that interference with cadherin-β-catenin interactions leads to a reduced bone mass [32]. ERK is an important factor in osteoblast differentiation, and RUNX2 is a bone specific transcription factor that is tightly regulated during the late mineralization stage of osteoblast differentiation [33]. ERK inhibition results in the suppression of differentiation markers [34]. Syringetin, the main flavonoid present in red grapes, may be beneficial in stimulating osteoblastic activity, by involving the ERK1/2 signaling pathway [35]. In another study, naringin treatment enhanced both transcriptional and translational levels of the β-catenin signaling of UMR-106 cells, and improved bone development in OVX mice, by stabilizing β-catenin through AMPK and Akt signaling [36].
We used siRNAs to suppress bone matrix protein expression. As shown in previous studies [37][38][39], RNA interference, using siRNA, is a useful biological strategy to study bone loss pathogenesis, and is a novel approach to treat several diseases. As expected, the inhibition of β-catenin by siRNA, in the presence of BPF, reversed the effects on RUNX2 and RANKL protein expression.
Taken together, our experiments with RNA interference confirm the central role of Wnt/β-catenin in osteoblast gene expression, and that BPF acts through this pathway.
COL1A, the main component of a bone matrix, plays a key role in bone, in transferring stress and resisting against deformation and fractures [40]. However, in our study we found a significant reduction in the secretion of the extracellular COL1A ( Figure 4B). This finding may confirm that a negative feedback mechanism exists to preserve bone tissue homeostasis. It has been demonstrated that serum TGF-β1, which is produced by osteoblast, is reduced in osteoporotic men [41], and that TGF-β1 is downregulated by extracellular COL1 and COL2 [42]. Thus, reduced extracellular COL1, by upregulating TGF-β1, might protect against osteoarticular diseases. Our experiments also show that BPF produced higher amounts of intracellular COL1A than the control cells ( Figure 4A). We hypothesize that BPF might promote bone repair through balancing collagen synthesis. However, since posttranslational processes are strongly involved in the formation of COL1 fibers, we believe that estrogen, vitamin D, and age-related effects on collagen secretion should be investigated in detail in future studies.
In this study, the incubation time and doses used were chosen based on previous reports [29,43,44]. In addition, since some natural polyphenols produce H 2 O 2 , during the so-called autoxidation process of polyphenols, and H 2 O 2 acts as a second messenger modulating gene expression [45], we chose the short incubation time of 10 min.
Flavanone distribution within different citrus species can be quite distinctive. The active ingredients in Bergamot are mainly naringin, neohesperidin and neoeriocitrin. C. bergamia contains 2.23, 1.60 and 1.38 mg/100 mL of naringin, neohesperidin and neoeriocitrin, respectively, whereas C. aurantium (bitter orange) contains 1.97, 0.87 and 0.77 mg/100 mL of these same flavanones, respectively, and C. aurantifolia (lime) only contains 0.01 mg/100 mL of neoeriocitrin [46]. It is evident that Bergamot is characterized by its unique profile of flavanones, and the large amounts of them. Furthermore, Bergamot contains the rare brutieridin and melitidin flavonoids.
Most of the beneficial effects found for citrus flavonoids were based on animal and in vitro studies, which were crucial in explaining mechanisms of these components. Unfortunately, very limited clinical studies have been conducted on citrus flavonoids, or citrus fruit juices, in relation to possible benefits on bone. Furthermore, the amount of polyphenols, flavonoids or other bioactive components cannot be determined from these studies. There is also a significant lack of information regarding clinical studies with pure naringin, neohesperidin and neoeriocitrin. In one study, BFP was safely used at a dose of 500 and 1000 mg/day [20]. In another study, Fang et al. [47] reported plasma concentrations, in rats, of 3.8, 0.23 and 43.5 µg/mL for naringin, naringenin and naringenin glucuronide, respectively, after an oral administration of 746·7 mg/kg naringin as a pure compound. Because we tested BPF from 1 to 100 µg/mL, and in line with these previous studies, we may predict the in vivo exposure conditions. However, we recognize that, when extrapolating in vitro models to in vivo scenarios, additional evidence is needed.
Our results may have other clinical implications. Loss of Wnt/beta-catenin pathway activity may contribute to osteosarcoma development [48]. Osteosarcoma is the most common malignant bone cancer. Thus, carefully designed studies are needed to investigate the anticancer functions of BPF.

Conclusions
Our results highlight, for the first time, a potential positive effect of polyphenols from Bergamot, a citrus fruit containing a high flavonoid concentration, on osteoblast differentiation and production of collagen. BPF also reduces RANKL expression, suggesting that it may inhibit osteoclastogenesis. Further studies are necessary to confirm our findings. Author Contributions: T.M., A.P. and S.R. designed research; C.R. and S.M. conducted research; T.M., S.R. and V.M. analyzed data; T.M., C.R. and R.P. wrote the paper; C.R. and S.M. had primary responsibility for integrity of data and final content. All authors read and approved the final manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.