TGF-β 1-Dependent Downregulation of HDAC 9 Inhibits Maturation of Human Osteoblasts

Transforming growth factor β (TGF-β) is a key regulator of bone density. Recently, we have shown that TGF-β1 effectively blocks bone morphogenetic protein-induced maturation of human osteoblasts (hOBs) in a histone deacetylase (HDAC)-dependent manner. To better understand the underlying mechanisms and to identify possible therapeutic targets, the current study aimed at characterizing the expression changes of different HDACs in hOBs following recombinant human TGF-β1 treatment and investigating the effect of the altered HDACs on both the proliferation and maturation of hOBs and osteogenic cell lines. As expected from our previous work, exposure to rhTGF-β1 induced the expression of HDACs (HDAC1, -2, -3, -6). However, to our surprise, rhTGF-β1 treatment strongly suppressed the expression of HDAC9 during osteogenic differentiation. HDAC9 is reported to suppress osteoclastogenesis; however, little is known about the role of HDAC9 in osteogenesis. Chemical inhibition of HDAC9 with TMP269 increased cell numbers of hOBs, but significantly decreased their osteogenic function (alkaline phosphatase activity and matrix mineralization). In osteogenic cell lines (MG-63, CAL-72 and SAOS-2), the expression of HDAC9 negatively correlates with their proliferation capacity and positively correlates with their osteogenic differentiation potential. Being able to boost osteoclasts while inhibiting osteoblasts makes HDAC9 an interesting therapeutic target to support fracture healing and bone metabolisms.


Introduction
Transforming growth factor β (TGF-β) is the most abundant cytokine in bone matrix, affecting both bone cell growth and differentiation [1].Bone cells secrete three different isoforms of TGF-β (TGF-β 1 , -β 2 , and -β 3 ), which are incorporated in the bone matrix in their inactive form.Upon proteolytic activation, TGF-β 1-3 bind to a common set of serine/threonine kinase receptors and mediate their signaling via Smad transcription factors [1,2].
In the past years, the mechanisms of how HDACs regulate osteogenesis have attracted increasing interest.Chemical inhibition of HDAC activity could improve osteogenic differentiation, suggesting both a direct acetylation-dependent [16,18], as well as a MAPKinase-dependent mechanism [19].There is evidence that several HDACs affect the expression and activity of the key osteogenic transcription factor Runx2. Runx2 expression is shown to be inhibited by HDAC1 and HDAC3 [20][21][22].Runx2 transcriptional activity is reduced by the binding of class II HDACs (HDAC4, -5, -6, and -7) to Runx2 [23][24][25][26].In addition, enhanced expression of HDAC6 causes deacetylation of α-tubulin, causing a shortening and distortion of the cells primary cilia (mechanosensors) and thus impairing mechanically driven osteogenesis [27].The expression of HDAC9, which plays a crucial role in immunity, is reported to negatively correlate with osteoclastogenesis [28].
In order to investigate how rhTGF-β 1 affects osteoblast maturation, the current study aimed at determining the expression profile of different HDAC isoforms in hOBs during osteogenic differentiation in the presence or absence of rhTGF-β 1 , and investigating the effect of the altered HDACs on both the proliferation and osteogenic differentiation of hOBs and osteogenic cell lines.

Materials
Culture media and chemicals were obtained from Sigma-Aldrich (Munich, Germany) if not stated differently.Fetal calf serum (FCS) was purchased from Thermo Fisher Scientific (Darmstadt, Germany).PCR primers were obtained from Eurofins Genomics (Ebersberg, Germany).

Ethics Statement
All human studies were performed in accordance with the 1964 Declaration of Helsinki.Isolation of hOBs and all following experiments were in accordance with the ethical vote (ID 364/2012BO) approved (7 August 2012) by the ethics committee of the medical faculty of the Eberhard-Karls-Universität and University clinic Tübingen.In accordance with the ethical vote, informed consent (signature) was obtained from each patient donating bone samples.The donors' average age was 65.4 ± 13.2 years (seven male and 17 female).Potential tumor patients or patients with viral or bacterial infections were excluded from this study.

(q)RT-PCR
Total RNA was isolated using Trifast reagent (Peqlab, Erlangen, Germany).Screening for HDAC expression was performed with the RT 2 Profiler PCR Array for human epigenetic chromatin modification enzymes (Qiagen, Hilden, Germany) as indicated by the manufacturer.The array was performed in duplicates (n = 2) with pooled (N = 16 donors) samples.phOBs without rhTGF-β 1 treatment were used as a control.Data were analyzed with the corresponding online software.

Sulforhodamine B (SRB) Staining
Total protein content was determined by SRB (Sulforhodamine B) staining.Briefly, cells were fixed for 1 h with ice cold ethanol.Cells were stained with 0.4% SRB (in 1% acetic acid) for 20 min at ambient temperature.After washing four to five times with 1% acetic acid, bound SRB was resolved with 10 mM unbuffered TRIS solution (pH~10.5)and absorption (λ = 565 nm; Omega plate reader, BMG Labtech, Ortenberg, Germany) was determined.Cells numbers were determined using specific standard curves for each cell type.

Assessing Matrix Mineralization by Alizarin Red Staining
As a late osteogenic marker, matrix mineralization was measured.Cells were fixed with ice cold ethanol for 1 h.After washing with tap water, cells were incubated with 0.5% Alizarin Red solution (pH 4.0) for 30 min at RT.After washing with tap water, the resulting staining (red) was assessed microscopically.Alizarin Red staining was quantified photometrically (λ = 562 nm; Omega plate reader) after resolving with 10% cetylpyridiumchloride solution [12,16].

Statistics
Results are expressed as a bar chart (mean ± 95% confidence interval) of at least four independent experiments (N ≥ 4) measured as triplicates or more (n ≥ 3).The exact number of biological (N) and technical (n) replicates is given in the figure legends.Datasets were compared by one-way ANOVA (Kruskall Wallis test; GraphPad Software Inc., La Jolla, CA, USA) followed by Dunn's multiple comparison test.p < 0.05 was taken as the minimum level of significance.

rhTGF-β 1 Treatment Significantly Downregulated HDAC9 Expression in hOBs
Expression of class I HDACs (HDAC1, -2 and -3) has been reported to negatively correlate with the expression the key osteogenic transcription factor Runx2 [20][21][22].In our group, we could show that the upregulation of HDAC6 results in shortened and deformed primary cilia, which impairs the mechanically driven osteogenesis [27].Besides the increased expression of these HDACs, the expression of HDAC9 was significantly reduced.HDAC9-deficient mice show extremely low bone mineral densities, however, this effect has been associated with an increased osteoclastogenesis [28].So far little is known about the role of HDAC9 in hOBs.
Thus, in the next step we aimed at confirming the effect of rhTGF-β 1 on HDAC9 expression.HOBs (N = 8) were cultured in osteogenic differentiation medium in the presence or absence of 5 ng/mL rhTGF-β 1 .After 0, 2, 4, and 7 days, the expression of HDAC9 was analyzed.HDAC9 levels continuously increased during the maturation process.The addition of 5 ng/mL rhTGF-β 1 significantly suppressed HDAC9 expression at the timepoints investigated (Figure 2a,b).

rhTGF-β1 Treatment Significantly Downregulated HDAC9 Expression in hOBs
Expression of class I HDACs (HDAC1, -2 and -3) has been reported to negatively correlate with the expression the key osteogenic transcription factor Runx2 [20][21][22].In our group, we could show that the upregulation of HDAC6 results in shortened and deformed primary cilia, which impairs the mechanically driven osteogenesis [27].Besides the increased expression of these HDACs, the expression of HDAC9 was significantly reduced.HDAC9-deficient mice show extremely low bone mineral densities, however, this effect has been associated with an increased osteoclastogenesis [28].So far little is known about the role of HDAC9 in hOBs.
Thus, in the next step we aimed at confirming the effect of rhTGF-β1 on HDAC9 expression.HOBs (N = 8) were cultured in osteogenic differentiation medium in the presence or absence of 5 ng/mL rhTGF-β1.After 0, 2, 4, and 7 days, the expression of HDAC9 was analyzed.HDAC9 levels continuously increased during the maturation process.The addition of 5 ng/mL rhTGF-β1 significantly suppressed HDAC9 expression at the timepoints investigated (Figure 2a,b).

Inhibition of HDAC9 Impairs Osteogenic Differentiation of hOBs
In order to further investigate which effects the suppression of HDAC9 has on hOBs proliferation and osteogenic function, we inhibited its activity chemically.TMP269 is reported to selectively inhibit HDAC9 activity with an IC50 as low as 23 nM [29].Our hOBs showed no toxic effects (LDH release) up to a concentration of 160 nM.To effectively block HDAC9 activity, a concentration of 40 nM TMP269 was chosen, which is well below the IC50 of the other HDACs (HDAC4, -5, and 7) that might also be blocked by this chemical [29].
hOBs were osteogenically differentiated for 14 days in the presence or absence of either 5 ng/mL rhTGF-β1 or 40 nM TMP269 (HDAC9i).As reported earlier, the addition of 5 ng/mL rhTGF-β1 significantly increased mitochondrial activity (Figure 3a) and total protein content (Figure 3b) in hOBs.Similarly, though not as pronounced, 40 nM TMP269 (HDAC9i) significantly increased mitochondrial activity (Figure 3a) and total protein content (Figure 3b) in hOBs.This indicates an increased proliferation in these cells.Contrarily, both conditions significantly suppressed hOBs maturation, as can be seen in the significantly decreased AP activity (Figure 3c) early in the differentiation process (day 7) and the resulting matrix mineralization (Figure 3d) as the endpoint of the maturation (day 21).Again, the effect of the rhTGF-β1 treatment was more pronounced as HDAC9 inhibition (TMP269 treatment) alone.

Inhibition of HDAC9 Impairs Osteogenic Differentiation of hOBs
In order to further investigate which effects the suppression of HDAC9 has on hOBs proliferation and osteogenic function, we inhibited its activity chemically.TMP269 is reported to selectively inhibit HDAC9 activity with an IC 50 as low as 23 nM [29].Our hOBs showed no toxic effects (LDH release) up to a concentration of 160 nM.To effectively block HDAC9 activity, a concentration of 40 nM TMP269 was chosen, which is well below the IC 50 of the other HDACs (HDAC4, -5, and 7) that might also be blocked by this chemical [29].
hOBs were osteogenically differentiated for 14 days in the presence or absence of either 5 ng/mL rhTGF-β 1 or 40 nM TMP269 (HDAC9i).As reported earlier, the addition of 5 ng/mL rhTGF-β 1 significantly increased mitochondrial activity (Figure 3a) and total protein content (Figure 3b) in hOBs.Similarly, though not as pronounced, 40 nM TMP269 (HDAC9i) significantly increased mitochondrial activity (Figure 3a) and total protein content (Figure 3b) in hOBs.This indicates an increased proliferation in these cells.Contrarily, both conditions significantly suppressed hOBs maturation, as can be seen in the significantly decreased AP activity (Figure 3c) early in the differentiation process (day 7) and the resulting matrix mineralization (Figure 3d) as the endpoint of the maturation (day 21).Again, the effect of the rhTGF-β 1 treatment was more pronounced as HDAC9 inhibition (TMP269 treatment) alone.

HDAC9 Expression Negatively Correlates with Proliferation and Positively Correlates with Maturation of Osteogenic Cell Lines
The observed data suggest that HDAC9 expression or activity negatively correlates with cell proliferation and positively correlates with osteogenic differentiation.Thus, in the next step different osteogenic cell lines with different differentiation states were tested.The osteogenic cell lines MG-63, CAL-72, and SAOS-2 were osteogenically differentiated for 14 days (each N = 4, n = 4).In order to determine the proliferation rate, mitochondrial activity and total protein content were determined on day 0 and 4 of differentiation.Within these 4 days, the mitochondrial activity and total protein content in MG-63 cells increased most (12.4-fold and 10.7-fold, respectively), followed by CAL-72 cells (5.3fold and 4.4-fold, respectively) and SAOS-2 cells (2.8-fold and 2.4-fold, respectively).These data indicate that MG-63 cells proliferate the fastest and SAOS-2 cells the slowest (Figure 4a,b).
In these cells, HDAC9 expression levels were determined in the beginning of the differentiation process (day 2).MG-63 cells showed the lowest expression for HDAC9.CAL-72 cells showed approximately 3-fold higher HDAC9 expression and SAOS-2 cells showed approximately 4.5-fold higher HDAC9 expression (Figure 4c,d).
As an early osteogenic marker, AP activity was measured on day 7 of differentiation.The highest AP activity was measured in SAOS-2 cells, followed by CAL-72 cells and MG-63 cells (Figure 4e).Similarly, the resulting matrix mineralization, as determined by Alizarin red staining, was strongest in SAOS-2 cells, followed by CAL-72 cells and MG-63 cells (Figure 4f).

HDAC9 Expression Negatively Correlates with Proliferation and Positively Correlates with Maturation of Osteogenic Cell Lines
The observed data suggest that HDAC9 expression or activity negatively correlates with cell proliferation and positively correlates with osteogenic differentiation.Thus, in the next step different osteogenic cell lines with different differentiation states were tested.The osteogenic cell lines MG-63, CAL-72, and SAOS-2 were osteogenically differentiated for 14 days (each N = 4, n = 4).In order to determine the proliferation rate, mitochondrial activity and total protein content were determined on day 0 and 4 of differentiation.Within these 4 days, the mitochondrial activity and total protein content in MG-63 cells increased most (12.4-fold and 10.7-fold, respectively), followed by CAL-72 cells (5.3-fold and 4.4-fold, respectively) and SAOS-2 cells (2.8-fold and 2.4-fold, respectively).These data indicate that MG-63 cells proliferate the fastest and SAOS-2 cells the slowest (Figure 4a,b).
In these cells, HDAC9 expression levels were determined in the beginning of the differentiation process (day 2).MG-63 cells showed the lowest expression for HDAC9.CAL-72 cells showed approximately 3-fold higher HDAC9 expression and SAOS-2 cells showed approximately 4.5-fold higher HDAC9 expression (Figure 4c,d).
As an early osteogenic marker, AP activity was measured on day 7 of differentiation.The highest AP activity was measured in SAOS-2 cells, followed by CAL-72 cells and MG-63 cells (Figure 4e).Similarly, the resulting matrix mineralization, as determined by Alizarin red staining, was strongest in SAOS-2 cells, followed by CAL-72 cells and MG-63 cells (Figure 4f).

Discussion
Although it is generally accepted that TGF-β signaling plays a crucial role in bone remodeling, its regulatory role in terms of bone mass maintenance is not yet clearly characterized [3][4][5][6][7].We have previously shown that rhTGF-β1 effectively blocks rhBMP2-and rhBMP7-dependent maturation of hOBs, in an HDAC-dependent manner [16].This raised the question of which HDACs play a role in this mechanism.In the present study, we first aimed at identifying changes in HDAC expression induced by rhTGF-β1 treatment during osteogenic differentiation.In hOBs, the strongest basal gene expression levels were observed for class I HDACs (HDAC1, -2, -3, and -8) and class IIa HDACs (HDAC4, -5, -7, and -9).HDACs of class IIb (HDAC6 and -10) and class IV (HDAC11) were hardly expressed.Gene expression levels of the different HDACs did not alter significantly during osteogenic differentiation.However, the addition of 5 ng/mL rhTGF-β1 strongly induced the expression of HDAC6 (>20-fold) and doubled the expression of HDAC1, -3, -5, and -7.This is in line with our earlier observation that rhTGF-β1 treatment significantly induces HDAC activity [16].Thus, it is astonishing that the expression of HDAC9 almost disappears upon treatment with rhTGF-β1.
In the promoter of HDAC9 there are several binding sites for MEF2A and MEF2C, which have been reported to regulate the expression of HDAC9 [30,31].Liu and colleagues could show that TGFβ suppresses MEF2 function by the direct interaction of the transcription factor Smad3 with MEF2C [32].This could explain the observed drop in HDAC9 expression in rhTGF-β1-treated cells.Thus, patients with chronically elevated TGF-β1-levels (e.g., diabetics, patients with chronic inflammation, etc.), frequently suffering from delayed fracture healing, are at a high risk for decreased HDAC9 levels.
HDAC9 is described to be a critical regulator in immune response; however, little is known about the role of HDAC9 in bone metabolism.HDAC9 knockout mice have significantly lower bone mineral densities, as a consequence of increased osteoclastogenesis [28].HDAC9 is a target of miRNA188, being upregulated with age.In miRNA188 knockout mice, which have high HDAC9

Discussion
Although it is generally accepted that TGF-β signaling plays a crucial role in bone remodeling, its regulatory role in terms of bone mass maintenance is not yet clearly characterized [3][4][5][6][7].We have previously shown that rhTGF-β 1 effectively blocks rhBMP2-and rhBMP7-dependent maturation of hOBs, in an HDAC-dependent manner [16].This raised the question of which HDACs play a role in this mechanism.In the present study, we first aimed at identifying changes in HDAC expression induced by rhTGF-β 1 treatment during osteogenic differentiation.In hOBs, the strongest basal gene expression levels were observed for class I HDACs (HDAC1, -2, -3, and -8) and class IIa HDACs (HDAC4, -5, -7, and -9).HDACs of class IIb (HDAC6 and -10) and class IV (HDAC11) were hardly expressed.Gene expression levels of the different HDACs did not alter significantly during osteogenic differentiation.However, the addition of 5 ng/mL rhTGF-β 1 strongly induced the expression of HDAC6 (>20-fold) and doubled the expression of HDAC1, -3, -5, and -7.This is in line with our earlier observation that rhTGF-β 1 treatment significantly induces HDAC activity [16].Thus, it is astonishing that the expression of HDAC9 almost disappears upon treatment with rhTGF-β 1 .
In the promoter of HDAC9 there are several binding sites for MEF2A and MEF2C, which have been reported to regulate the expression of HDAC9 [30,31].Liu and colleagues could show that TGF-β suppresses MEF2 function by the direct interaction of the transcription factor Smad3 with MEF2C [32].This could explain the observed drop in HDAC9 expression in rhTGF-β 1 -treated cells.Thus, patients with chronically elevated TGF-β 1 -levels (e.g., diabetics, patients with chronic inflammation, etc.), frequently suffering from delayed fracture healing, are at a high risk for decreased HDAC9 levels.
HDAC9 is described to be a critical regulator in immune response; however, little is known about the role of HDAC9 in bone metabolism.HDAC9 knockout mice have significantly lower bone mineral densities, as a consequence of increased osteoclastogenesis [28].HDAC9 is a target of miRNA188, being upregulated with age.In miRNA188 knockout mice, which have high HDAC9 levels, age-dependent switch from osteogenesis to adipogenesis in bone is delayed [33], suggesting that HDAC9 plays a direct role in osteogenesis.In our experiments, the reduced expression of HDAC9 as well as the inhibition of HDAC9 activity resulted in reduced osteogenic function of hOBs.
It has been described that HDAC9 is strongly suppressed during adipogenesis.The lack of HDAC9 increases the transcriptional activity of C/EBPα, PPARγ, and FABP4 in adipocytes [34].Osteoblast-specific overexpression of PPARγ markedly affects bone mineral density.However, the decrease in bone mineral density was associated with a decrease in OPG to RANKL ratio and thus an increased osteoclastogenesis [35].This was confirmed with the massive bone growth of mice having an osteoclast-specific knockdown of PPARγ [36].In our osteogenic cell lines, OPG expression was strongest for MG-63 and SaOS-2 cells, and RANKL expression was strongest for SaOS-2 cells.The resulting OPG to RANKL ratio was lowest for SaOS-2 cells, where HDAC9 expression was highest, suggesting that other mechanisms might play a role and that the osteoblast-specific overexpression of HDAC9 might increase bone turnover.
Furthermore, in our mono-cultures, osteoclasts were the main target of HDAC9.As the HDAC9 inhibitor TMP269 was able to suppress osteogenic differentiation, this points towards a direct HDAC9-dependent mechanism.It has been shown that HDAC9 expression is not only controlled by MEF2A and MEF2C, but can regulate the transcriptional activity of MEF2 itself [31].MEF2 transcriptional activity is crucial for osteogenic differentiation, as it regulates the expression of the key osteogenic transcription factor Runx2 and other osteogenic marker genes [37,38].This could explain why hOBs treated with the HDAC9 inhibitor (TMP269) showed suppressed osteogenic function.However, the effect was not as pronounced as for those with rhTGF-β 1 treatment.This might be due to the other HDACs that were upregulated by rhTGF-β 1 treatment.Improved osteogenic differentiation by using small chemical inhibitors for HDACs [16,18] points towards a direct acetylation-dependent mechanism.This is described for HDAC1 and -3 overexpression, which inhibits the expression of the key osteogenic transcription factor Runx2 [20][21][22].Furthermore, the overexpression of HDAC6 causes a de-acetylation of α-tubulin, resulting in damaged primary cilia and impaired mechanically driven osteogenesis [27].Conversely, HDAC4, -5, -6, -7, and -8 have been described to bind Runx2 and thus interfere with its transcriptional activity [23][24][25][26].Thus, the decrease in HDAC9 in combination with the increased expression of HDAC1, -2, -3, and -6 in our rhTGF-β 1 -treated hOBs could be responsible for the observed loss of function.
Interestingly, with the decrease in osteogenic differentiation, the hOBs proliferation increased, as could be seen in the increased mitochondrial activity and total protein content.Using different osteogenic (osteosarcoma) cell lines, we found a negative correlation of HDAC9 expression with cell proliferation as well as a positive correlation of HDAC9 expression with osteogenic maturation of these cells.This is in clear contrast to the work from Zhao, which showed that the overexpression of HDAC9 in U2OS and MG-63 osteosarcoma cells promoted their proliferation by suppressing p53 transcriptional activity [39].Thus, further experiments have to be done to investigate how HDAC9 regulates hOBs and osteosarcoma cell proliferation, or whether the expression of HDAC9 can be used to predict the osteogenic differentiation potential of these cells.

Conclusions
Summarizing our results, rhTGF-β 1 treatment significantly suppressed the expression of HDAC9 in hOBs, which resulted in increased proliferation and decreased maturation of these cells.Considering also the reports that HDAC9 knockout induced osteoclastogenesis, HDAC9 represents an interesting therapeutic target for bone, especially for patients with chronically elevated TGF-β 1 -levels (e.g., diabetics, patients with chronic inflammation, etc.), who frequently suffer from poor bone quality, increased fracture risk, and delayed fracture healing.

Figure 1 .
Figure 1.Effect of rhTGF-β1 treatment on HDAC expression during osteogenic differentiation of hOBs: The expression of HDACs was determined using the human RT 2 Profiler PCR Array for epigenetic chromatin modification enzymes (Qiagen, Hilden, Germany), in hOBs (N = 16/pooled; n = 2) osteogenically differentiated for 2 days in the presence or absence of 5 ng/mL rhTGF-β1.(a) Pie chart of the mean percental distribution of the HDAC isoforms (2 −ΔCt ) during osteogenic differentiation in the absence or presence of 5 ng/mL rhTGF-β1; (b) Expression changes of the different HDAC isoforms induced by 5 ng/mL rhTGF-β1 treatment (fold regulation).* p < 0.05 and ** p < 0.01 as compared to untreated cells.

Figure 1 .
Figure 1.Effect of rhTGF-β 1 treatment on HDAC expression during osteogenic differentiation of hOBs: The expression of HDACs was determined using the human RT 2 Profiler PCR Array for epigenetic chromatin modification enzymes (Qiagen, Hilden, Germany), in hOBs (N = 16/pooled; n = 2) osteogenically differentiated for 2 days in the presence or absence of 5 ng/mL rhTGF-β 1 .(a) Pie chart of the mean percental distribution of the HDAC isoforms (2 −∆Ct ) during osteogenic differentiation in the absence or presence of 5 ng/mL rhTGF-β 1 ; (b) Expression changes of the different HDAC isoforms induced by 5 ng/mL rhTGF-β 1 treatment (fold regulation).* p < 0.05 and ** p < 0.01 as compared to untreated cells.

Figure 2 .
Figure 2. rhTGF-β1 treatment suppresses HDAC9 expression during osteogenic differentiation of hOBs: hOBs (N = 8) were osteogenically differentiated in the presence or absence of 5 ng/mL rhTGF-β1.On day 0, 2, 4, and 7, the expression of HDAC9 was determined by RT-PCR.GAPDH was used as a housekeeping gene.(a) Representative figure of the RT-PCR products (N = 3); (b) Densitometric analysis (N = 8; n = 2) of all RT-PCR signals.hOBs without rhTGF-β1 treatment are represented as circles connected with straight lines and cells with rhTGF-β1 treatment are represented as rectangles connected with dotted lines.** p < 0.01 and *** p < 0.001 as compared to untreated cells.

Figure 4 .
Figure 4. HDAC9 expression negatively correlates with proliferation and positively correlates with maturation in osteogenic cell lines: osteogenic cell lines (MG-63, CAL-72 and SAOS-2/N = 4, n = 4) were osteogenically differentiated for 14 days.(a) Increase in mitochondrial activity (resazurin conversion) until day 4 of differentiation; (b) Increase in total protein content (SRB staining) until day 4 of differentiation; (c) Representative figure of the RT-PCR products (HDAC9 and GAPDH) on day 2 of differentiation; (d) Densitometric analysis of all RT-PCR signals; (e) AP activity on day 4 of differentiation; (f) Matrix mineralization (Alizarin Red staining) on day 14 of differentiation.** p < 0.01 and *** p < 0.001 as indicated.

Figure 4 .
Figure 4. HDAC9 expression negatively correlates with proliferation and positively correlates with maturation in osteogenic cell lines: osteogenic cell lines (MG-63, CAL-72 and SAOS-2/N = 4, n = 4) were osteogenically differentiated for 14 days.(a) Increase in mitochondrial activity (resazurin conversion) until day 4 of differentiation; (b) Increase in total protein content (SRB staining) until day 4 of differentiation; (c) Representative figure of the RT-PCR products (HDAC9 and GAPDH) on day 2 of differentiation; (d) Densitometric analysis of all RT-PCR signals; (e) AP activity on day 4 of differentiation; (f) Matrix mineralization (Alizarin Red staining) on day 14 of differentiation.** p < 0.01 and *** p < 0.001 as indicated.