**3. Results**

## *3.1. Polyphenolic Composition of BJ*

Phenolic constituents from BJ were identified by HPLC-PDA-MS analysis as reported in Section 2. The chromatographic profile displayed a suitable separation of all the polyphenol constituents in the juice after just one run, and the data analysis is reported in Table 1.


**Table 1.** Identified polyphenols and total content of each class of constituents in blueberry juice.


**Table 1.** *Cont.*

Fourteen anthocyanins were identified using UV wavelength detection at 520 nm and mass spectra in the positive ion mode, since these compounds are present as flavylium ions in the chromatographic conditions applied. Delphinidin, cyanidin, petunidin, peonidin, and malvidin were the representative aglycons, glycosylated at C-3 with the same sugars of anthocyanins, namely, glucose, galactose, and arabinose. Similarly, five flavanols were identified using UV wavelength detection at 360 nm and mass spectra in both the positive and the negative ion mode, as reported in Table 1. Aglycons were identified as quercetin and myricetin glycosylated at C-3. Two further flavonoid structures were identified, the flavan-3ols catechin and epicatechin, using the same strategy reported for the other flavonoids. Finally, ellagic acid and chlorogenic acid were unambiguously identified as simple polyphenols, by comparison of their retention times and UV and mass spectra with those of the authentic samples. Table 1 shows the total content of each group expressed as gallic acid. Anthocyanins, expressed as cyanidin-3-glucoside, represented the main constituents corresponding to about 0.63 mmol/100 mL of juice, whereas flavonol and catechin contents were about 0.017 and 0.034 mmol/100 mL, expressed as quercetin-3-glucoside and (+)-catechin equivalents, respectively. The amount of ellagic acid was ca. 0.073 mmol/100 mL juice, while that of chlorogenic acid was ca. 0.18 mmol/100 mL.

#### *3.2. E*ff*ect of BJ on Cell Viability and in Preventing Oxidative Stress Induced by BSO Treatment in SaOS-2 Cells*

Initially, we assessed cell viability during the pre-treatment (24 h in GM, day 1) before the induction of the differentiation and after two days of differentiation in the presence or not (control) of BSO and BJ (Table 2).



Viability was measured using the CellTiter-Glo Luminescent Cell Viability Assay in cells treated or not (control) with 40 μM BSO and/or BJ containing 15 <sup>μ</sup>g·mL−<sup>1</sup> of total soluble polyphenols. The data expressed as a percentage of the respective controls are means ± standard error of the mean (SEM) of four independent experiments.

In particular, BSO treatment was performed at the 40 μM concentration that induces oxidative stress due to significant decrease of GSH/GSSG ratio, an index of the intracellular oxidative status, as previously reported in SaOS-2 cells [13]. BJ containing two different concentrations of TSP (7.5 or 15 <sup>μ</sup>g·mL−1), which previously demonstrated antioxidant activity in starved osteocytes [33], was used. Table 2 shows that neither BSO nor the highest concentration of BJ, used alone or together with BSO, significantly altered cell viability as compared to control cells.

Figure 1 reports the BJ antioxidant effect on BSO-treated cells. In fact, BSO pre-treatment (24 h in GM, day 1) was able to significantly increase ROS levels as compared with the control values measured in untreated cells (Figure 1).

**Figure 1.** BJ effect on intracellular reactive oxygen species (ROS) production in BSO-treated SaOS-2 cells. Intracellular ROS, detected by measuring the fluorescence intensity of the probe <sup>2</sup>,7-dichlorodihydrofluorescein diacetate (H2DCFDA), were measured in SaOS-2 cells treated or untreated (C; control) with BSO or BSO + BJ containing 7.5 or 15 <sup>μ</sup>g·mL−<sup>1</sup> total soluble polyphenols (TSP) for one day in growth medium (GM) and subsequently treated or untreated (C; control) for two or six days in osteogenic medium (OM), as reported in Section 2. ROS data, normalized on total protein content, are expressed as a percentage of the C values, and they are the means ± SEM of four experiments repeated in triplicate; \* *p* ≤ 0.05, \*\* *p* ≤ 0.001 compared to C cells; - *p* ≤ 0.05, -- *p* ≤ 0.001 compared to BSO-treated cells; -- *p* ≤ 0.001 compared to BSO-treated cells for one day in GM.

ROS levels increased further when BSO was subsequently added for other two days in OM as compared with control (Figure 1); from this time on (two days), BSO was no longer added, and ROS content returned to the control levels after six days after the induction of differentiation (Figure 1). In order to prevent the effect of BSO, the cells were treated simultaneously with BSO and BJ containing 7.5 or 15 <sup>μ</sup>g·mL−<sup>1</sup> TSP. Figure 1 reports that BJ at both concentrations significantly prevented ROS increase in SaOS-2 cells after just 24 h in GM, and this effect was even more marked (by about 50–70%) after two days after the induction of differentiation. The highest concentration was able to decrease ROS levels to control values. No change in the intracellular oxidative state was observed after six days after the induction of differentiation in all conditions used (Figure 1). Similarly, no change in ROS levels was observed, when only BJ at both TSP concentrations was added before and during the differentiation at all studied times (data not shown), indicating that BJ per se does not have any effect on a normal cellular redox state.

#### *3.3. BJ E*ff*ect on the Markers of Di*ff*erentiation and Osteogenic Process in SaOS-2 Cells in the Presence or Not of BSO-Induced Oxidative Stress*

In SaOS-2 cells treated or not with BSO, BJ, or BSO + BJ, as reported in Section 2, the levels of alkaline phosphatase (ALP, EC 3.1.3.1), an early biochemical marker of osteoblast differentiation and osteogenic activity, were measured [13]. This enzyme is considered an osteoblast phenotype marker in SaOS-2 cells and, therefore, an osteoblast differentiation indicator [37,43]. Figure 2 shows the time course of ALP activity in SaOS-2 cells as a percentage of the activity values measured in cells cultured in GM (controls).

**Figure 2.** BJ effect on alkaline phosphatase (ALP) activity during differentiation of BSO-treated and untreated SaOS-2 cells. ALP activity, detected by the QuantiFluo Alkaline Phosphatase Assay Kit, was measured in SaOS-2 cells treated or untreated with BJ, containing 7.5 or 15 <sup>μ</sup>g·mL−<sup>1</sup> total soluble polyphenols (TSP), or BSO or BSO + BJ for one day in growth medium (GM). Subsequently, untreated cells were cultured in GM (C; control) for two, six, 10, or 20 days and, for the same times, treated and untreated cells were cultured in osteogenic medium (OM), as reported in Section 2**.** ALP activity is expressed as a percentage of the respective C values, and these are the means ± SEM of four experiments performed in triplicate; \* *p* ≤ 0.05, \*\* *p* ≤0.001 compared to C cells; - *p* ≤ 0.05, -- *p* ≤ 0.001 compared to untreated cells; - *p* ≤ 0.05 compared to BSO-treated cells.

ALP activity significantly increased in the early phase of differentiation process (2–6 days) in untreated cells cultured in OM as compared to the respective controls. Ten days after the induction of the differentiation, the percentage of ALP increase was no longer significantly different from the control (Figure 2). The ALP trend, in cells treated with BJ containing both concentrations of TSP, showed that its activity was higher during the initial phase of differentiation with respect to the untreated cells, and it remained high even after 10 days, unlike what happened in the untreated cells (Figure 2). It is worthy to note that the BJ effect occurred in cells with normal redox state. On the contrary, BSO significantly inhibited the upregulation of ALP activity levels (Figure 2). Finally, in BSO + BJ-treated cells, BJ was able to eliminate the effect of BSO and to maintain ALP levels at the values measured in untreated cells (Figure 2).

Subsequently, under the same experimental conditions, the expression and activation of RUNX-2 by tyrosine phosphorylation were studied six days after differentiation in SaOS-2 cell lysates and in nuclear extracts by Western blot analysis (Figure 3A–C).

**Figure 3.** BJ effect on Runt-related transcription factor 2 (RUNX-2) expression and activation in BSO-treated and untreated SaOS-2 cells. RUNX-2 expression (**A**), nuclear fraction purification (**B**), and RUNX-2 activation (**C**) were detected in SaOS-2 cells treated or not with BJ at 15 <sup>μ</sup>g·mL−<sup>1</sup> total soluble polyphenols (TSP) or BSO or BSO + BJ for one day in growth medium (GM) and, subsequently, for six days in osteogenic medium (OM), in the presence or not of various treatments, as reported in Section 2. RUNX-2 expression was detected by Western blot analysis in whole cellular lysate (**A**), RUNX-2 activation was detected by phosphorylation in immunoprecipitates of nuclear extracts using the anti-RUNX-2 antibody (**C**), and β-actin and histone H3 were detected in nuclear extract (**B**). RUNX-2, p-RUNX-2, β-actin, and histone H3 were revealed with anti-RUNX-2, anti-p-tyrosine proteins, anti-β-actin, or anti-histone H3, respectively. The blots are representative of three experiments.

RUNX-2 is an important transcription factor for the activation of osteoblast differentiation and for regulation of bone remodeling, and it is involved in many bone diseases [44,45]. Figure 3A shows that BJ containing 15 <sup>μ</sup>g·mL−<sup>1</sup> TSP was able to upregulate RUNX-2 levels, whereas no change was detected in cells treated with BSO only. RUNX-2 activation was evaluated by immunoprecipitation of equal amounts of nuclear proteins using anti-RUNX-2 antibody. The purification of the nuclear fraction was observed in control cells by Western blot, which revealed the presence of the histone H3 protein and not β-actin (Figure 3B). Figure 3C shows the absence of RUNX-2 or P-RUNX-2 or histone H3 bands in Western blot analysis of immunoprecipitates performed with IgG (negative control) in control cells; unlike RUNX-2 or P-RUNX-2 bands, no Histone H3 bands were detected after immunoprecipitation with anti-RUNX-2 (Figure 3). These data demonstrate the absence of non-specific bands under these experimental conditions. Figure 3C also shows that BJ containing 15 <sup>μ</sup>g·mL−<sup>1</sup> TSP increased RUNX-2 nuclear levels and its phosphorylation as compared to control. On the other hand, decreases in RUNX-2 phosphorylation and no change in its nuclear levels were observed in cells treated with BSO, whereas RUNX-2 phosphorylation appeared to be restored by simultaneous treatment with BJ. Overall, these findings indicate that BJ was able to increase RUNX-2 and ALP activation, in normal redox state conditions, as well as prevent the inhibition of these factors in BSO-induced oxidative stress.

#### *3.4. BJ E*ff*ect on the Mineralization Process in SaOS-2 Cells in the Presence or Not of BSO-Induced Oxidative Stress*

The mineralization process was studied in cells treated or not (control) with BJ containing 15 <sup>μ</sup>g·mL−<sup>1</sup> TSP or with BSO or BJ + BSO. Calcium deposition was measured 12 and 24 days after the induction of differentiation by staining with Alizarin Red S (Figure 4A).

**Figure 4.** BJ effect on the mineralization process during differentiation of BSO-treated SaOS-2 cells. Calcium content, measured by Alizarin Red S assay, was detected in SaOS-2 cells treated or untreated with BJ, containing 15 <sup>μ</sup>g·mL−<sup>1</sup> total soluble polyphenols (TSP), or BSO or BSO + BJ for one day in growth medium (GM) and, subsequently, treated or untreated (C; control) for 12 or 24 days in osteogenic medium (OM), as reported in Section 2. Calcium content is expressed as mg hydroxyapatite (HA) per cm<sup>2</sup> and values are means ± SEM of three independent experiments (**A**); \* *p* ≤ 0.05, \*\* *p* ≤ 0.001 compared to the respective C cells. Representative images of calcium content in 12-well plates (**B**).

Figure 4A,B show that BSO-induced oxidative stress in the initial phase of differentiation remarkably decreased the osteogenic activity after 12 days (by about 70%) as compared to control. This effect was prevented in BJ + BSO-treated cells, but no change in calcium deposition was observed in BJ-treated cells in this early phase as compared to control (Figure 4A,B). On the other hand, after 24 days, long-term activation of the mineralization process (of about 100%) was found in BJ- and BSO + BJ-treated cells, and no change was observed in BSO-treated cells as compared to control (Figure 4A,B). These data show that BJ phytochemicals prevent the anti-osteogenic effect of BSO-induced oxidative stress, and they are able to activate the mineralization process, even if this occurs later in time.

#### *3.5. Involvement of SIRT1 on BJ-Induced Activation of Osteogenic Di*ff*erentiation and Mineralization Process in SaOS-2 Cells*

We subsequently evaluated the involvement of sirtuin type 1 (SIRT1), a class III histone deacetylase, in BJ-induced activation of osteogenic factors and the mineralization process, which occurred in the presence of a normal intracellular redox state. A possible role of SIRT1 with regard to the antioxidant activity of BJ was also studied. Indeed, SIRT1 is related to the regulation of osteogenic di fferentiation of tendon and mesenchymal stem cells [46,47], and it is a positive regulator of RUNX-2 [47,48]. Moreover, this enzyme is activated in mammals by dietary blueberry [49,50] and in osteocytes by BJ [33]. Initially, the BJ e ffect on SIRT1 levels in SaOS-2 cells in the presence or not of BSO-induced oxidative stress was studied. Figure 5A shows that BJ significantly increased SIRT1 expression as compared to control six days after di fferentiation, and a similar increase was also observed in BSO + BJ-treated cells, whereas no change in SIRT1 levels was detected in BSO-treated cells.

**Figure 5.** BJ e ffect on sirtuin type 1 (SIRT1) expression and SIRT1 inhibition role in BJ-induced ALP activation during di fferentiation of SaOS-2 cells. SIRT-1 expression ( **A**) and ALP activity (**B**) were detected in SaOS-2 cells treated or not with BJ at 15 <sup>μ</sup>g·mL−<sup>1</sup> total soluble polyphenols (TSP) or BSO, BSO + BJ, or BJ + EX527, for one day in growth medium (GM) and, subsequently, in osteogenic medium (OM), in the presence or not of various treatments, as reported in Section 2. SIRT1 expression was detected after six days in OM by ELISA kit according to the manufacturer's instructions, and levels are expressed as a percentage of the C values (A). ALP activity was also detected after two and 10 days in GM (control, C) and OM, as well as in transfected SIRT1 small interfering RNA (siRNA) or Scr RNA cells after two days in OM. SIRT1 expression and ALP activity values are expressed as a percentage of the respective C values, and they are the means ± SEM of three independent experiments; \* *p* ≤ 0.05 compared to C cells; - *p* ≤ 0.05 compared to BJ-treated cells.

Subsequently, the role of SIRT1 in BJ activation of osteoblast di fferentiation and mineralization was determined. To evaluate this, SIRT1 expression and activity were reduced using cells transfected with a specific SIRT1 siRNA or treated with EX527, a specific inhibitor of SIRT1, as reported in Section 2. Figure 5B reports that the downregulation of both SIRT1 and EX527, at the concentration able to inhibit SIRT1 activity [33], removed the activating e ffect of BJ on ALP activity two days after di fferentiation, and this e ffect was also observed at 10 days in EX527-treated cells.

SIRT1 involvement in BJ-induced RUNX-2 expression and activation was also studied in whole cellular lysates and in nuclear immunoprecipitates (Figure 6A,B).

**Figure 6.** Effect of SIRT1 inhibition on BJ activation of RUNX-2 and the mineralization process, as well as on BJ antioxidant action during differentiation of SaOS-2 cells. RUNX-2 expression (**A**), RUNX-2 activation (**B**), calcium content (**C**,**D**), and ROS levels (**E**) were detected in SaOS-2 cells treated or not with BJ at 15 <sup>μ</sup>g·mL−<sup>1</sup> total soluble polyphenols (TSP) or BSO, BSO + BJ, BJ + EX527, or BJ + BSO + EX527, for one day in growth medium (GM) and, subsequently, in osteogenic medium (OM), in the presence or not of various treatments, as reported in Section 2. RUNX-2 expression was detected after six days in OM by Western blot analysis of whole cellular lysates (**A**), and RUNX-2 activation was detected by phosphorylation in immunoprecipitates of nuclear extracts using the anti-RUNX-2 antibody (**B**). RUNX-2, p-RUNX-2, and β-actin were revealed with anti-RUNX-2, anti-p-tyrosine proteins, and anti-β-actin, respectively. The blots are representative of three experiments. Calcium content was measured after 24 days in OM, and it is expressed as mg hydroxyapatite per cm<sup>2</sup> (**C**). Representative images of calcium content in 12-well plates (**D**). ROS data are expressed as a percentage of the C values (**E**). The values are means ± SEM of three independent experiments; \* *p* ≤ 0.05 compared to the respective C cells; - *p* ≤ 0.05 compared to BSO + BJ-treated cells.

EX527 treatment partially prevented RUNX-2 expression (Figure 6A), as well as RUNX-2 activation (Figure 6B) induced by BJ six days after di fferentiation. The possible involvement of SIRT1 in long-term activation of the mineralization process induced by BJ is shown in Figure 6C,D. Finally, we evaluated the e ffect of SIRT1 on ROS levels in cells treated with BSO + BJ, considering that the antioxidant action of BJ could also be mediated by SIRT1. Figure 6E shows that BJ antioxidant action may only partially be mediated by SIRT1.
