HOX gene analysis in the osteogenic differentiation of human mesenchymal stem cells

Human bone marrow-derived mesenchymal stem cells (hMSCs) have the capacity to differentiate into osteoblasts during osteogenesis. Several studies attempted to identify osteogenesis-related genes in hMSCs. Although HOX genes are known to play a pivotal role in skeletogenesis, their function in the osteogenesis of hMSCs has not yet been investigated in detail. Our aim was to characterize the expression of 37 HOXgenes by multiplex RT-PCR to identify the ones most probably involved in osteogenic differentiation. The results showed that the expression patterns of four HOX genes were altered during this process. In particular, the expression levels of HOXC13 and HOXD13were dramatically changed. Real-time PCR and Western blot analysis were performed in order to further analyze the expression of HOXC13and HOXD13. The qRT-PCR results showed that transcription of HOXC13was up-regulated by up to forty times, whereas that of HOXD13was down-regulated by approximately five times after osteogenic differentiation. The Western blot results for the HOXC13 and HOXD13 proteins also corresponded well with the real-time PCR result. These findings suggest that HOXC13 and HOXD13 might be involved in the osteogenic differentiation of hMSCs.


Introduction
Bone marrow-derived stem cells can be divided into two major types: hematopoietic stem cells and nonhematopoietic, or mesenchymal, stem cells. Human bone marrow-derived mesenchymal stem cells (hMSCs) have the capacity for self-renewal and multilineage differentiation. Under the appropriate conditions, they can also give rise to mesenchymal tissues such as muscle, bone, fat, and cartilage (Pittenger et al., 1999). Due to their ability to differentiate into osteoblasts, chondrocytes, adipocytes, tenocytes and myoblasts, hMSCs hold promise for clinical applications in regenerative medicine (Song et al., 2006).
Because osteoblastic cells play a major role in the processes of normal bone growth, remodeling and fracture repair, many researchers have used the process of osteogenesis to study the differentiation and characteristics of stem cells (Kraus and Kirker-Head, 2006). To obtain osteoblastic cells, MSCs are incubated with a mixture medium containing dexamethasone, β-glycerophosphate and ascorbic acid for a period of 2~3 weeks (Bobis et al., 2006).
HOX genes were initially identified by their homology with the Drosophila HOM genes (Levine et al., 1984;Acampora et al., 1989;Duboule and Dolle, 1989). These genes encode homeodomain transcription factors related to anterior-posterior axis patterning that takes place during embryonic development (van den Akker et al., 2001). The homeodomain contains a 180-base-pair homeobox sequence that encodes a conserved 60 amino acid region and acts as a DNA-binding domain via a helix-turn-helix motif (Gehring et al., 1994). In vertebrates, 39 HOX genes have been identified. These are distributed over four homologous HOX clusters termed HOXA, B, C, and D. These loci are located on four different chromosomal locations and are comprised of nine to eleven genes (Akin and Nazarali, 2005). It is well known that HOX proteins participate in many common developmental processes during normal embryogenesis. Several reports have indicated that HOX genes play a regulatory role in skeletogenesis (Goff and Tabin, 1997;Kanzler et al., 1998;van den Akker et al., 2001;Remacle et al., 2004).
Although HOX genes are known to play an essential role in skeletal development and bone formation, there is no report regarding the screening of HOX groups that are involved in the osteogenesis of hMSCs. Thus, in the present study, the expression profile of HOX genes during osteogenic differentiation of hMSCs was investigated by multiplex PCR and the results showed significant changes in the expression of four of them during this process. Of these four genes, the expression of HOXC13 and HOXD13 showed the most dramatic changes. Therefore, the expression levels of HOXC13 and HOXD13 were evaluated by qRT-PCR and Western blot analysis, and the results were similar to the multiplex PCR result. This suggests that the other two genes (HOXA1 and HOXC11) are also involved in osteogenesis.

Research protocol
The research protocol was reviewed and approved by the human ethical care committee at St. Mary's Hospital, Catholic University in Daejeon, Republic of Korea. The hMSCs were isolated from the bone marrow of six individuals, as described below (Choi et al., 2006). All experiments were performed with hMSCs obtained after the third cell passage.
Flow cytometric analysis (FACS) of hMSCs hMSCs were analyzed by FACS-Calibur (Becton Dickinson, San Jose, CA) as previously described (Choi et al., 2006). FACS analysis was performed using fluorescein isothiocyanate (FITC)-conjugated anti-CD11b, CD29, CD34, CD45, CD73 and CD105 antibodies (BD Bioscience, San Diego, CA) to confirm that the phenotype of the hMSCs was maintained after expansion in the culture. The samples were incubated with antibodies against each surface marker for 30 min, and this treatment was followed by FACS.

Osteogenic differentiation
To induce osteogenic differentiation, hMSCs at the third passage were plated with Dulbecco's modified Ea-gle's medium (DMEM) containing 10% fetal bovine serum (FBS) in a 250-ml tissue culture flask (Nunc, Roskilde, Denmark). The cells were then incubated at 37°C in 5% CO 2 for 24 h. The medium was replaced with high-glucose DMEM containing 10% FBS, 0.1 μM dexamethasone, 10 mM β-glycerophosphate, and 0.3 mM ascorbic acid (Sigma, St. Louis, MO) for osteogenic differentiation. This osteogenic medium was replaced every 2 days for 21 days.

Alkaline phosphatase (ALP) staining
About 3 x 10 5 cells were seeded onto each well of a 6-well plate. After incubation for 12 h at 37°C in 5% CO 2 , the medium was replaced with osteogenic differentiation medium, replaced again every 2 days for periods of 10 and 21 days. The 10-day and 21-day differentiated and undifferentiated hMSCs were washed twice with ice-cold PBS (phosphate buffered saline), fixed with 2% paraformaldehyde/ 0.1 M sodium cacodylate for 10 min, and washed with 0.1 M cacodylic acid. The cells were incubated with ALP substrate solution (5 mg naphthol AS-TR phosphate in 25 mL water plus 10 mg Fast red TR in 24 mL of 0.1 M Tris buffer, pH 9.5) for 1 h at room temperature. Cells were photographed using a Nikon TE-300 (Tokyo, Japan) inverted light microscope.

von Kossa staining
Approximately 3 x 10 5 cells were seeded onto each well of a 6-well plate. After incubation for 12 h at 37°C in 5% CO 2 , the medium was replaced with osteogenic differentiation medium and thereafter replaced every 2 days for periods of 10 and 21 days. Day-10 and day-21 differentiated and undifferentiated hMSCs were washed with distilled water, fixed with 4% formalin, and then treated with 5% silver nitrate. Then the cells were exposed to UV light for 1 h, 5% thiosulfate was added, and the cells were placed at room temperature after a washing step with distilled water. The samples were photographed with a Nikon TE-300 (Tokyo, Japan) inverted light microscope.

RNA isolation and reverse transcriptase-polymerase chain reaction (RT-PCR)
RT-PCR analysis was performed as described by Jee et al. (2006). After the induction of osteogenic differentiation for 21 days, total RNA was isolated from the cells using an RNeasy Mini Kit (QIAGEN, Valencia, CA). Two micrograms of total RNA were reverse-transcribed in order to synthesize cDNA, using an AccuPower RTPReMix kit (Bioneer, Inc., Rockville, MD). The subsequent PCR amplification was performed with 1 μL of RT reaction mixture, using the following thermocycling profile: 1 cycle at 94°C for 5 min, followed by 30 cycles of 92°C for 1 min, 52°C for 1 min and 72°C for 1 min, and a final cycle at 72°C for 10 min. The primer sequences used are listed in Table 1. The GAPDH (glyceraldehyde-3-phosphate dehydrogenase) gene was used as an internal control. The PCR products were run on a 1% agarose gel and then analyzed under UV light after staining with ethidium bromide. The gel was photographed and then quantitatively measured by scanning densitometry. The experiments were performed with three different RNA samples.

Multiplex PCR of HOX genes
Multiplex PCR was performed using the GeneXP Human HOX Assay Kit (Seegene, Seoul, Republic of Korea). After the induction of osteogenic differentiation for 21 days, total RNA was isolated from the cells using an RNeasy Mini Kit (QIAGEN, Valencia, CA). Two micrograms of total RNA were reverse-transcribed in order to synthesize cDNA using an AccuPower RTPReMix kit (Bioneer, Inc., Rockville, MD). The synthesized cDNAs were used as templates for multiplex PCR,according to the manufacturer's instructions (http://www.seegene.com). PCR was carried out under the following conditions: 1 cycle at 94°C for 15 min, followed by 40 cycles of 94°C for 0.5 min, 63°C for 1.5 min, and 72°C for 1.5 min, and a final cycle at 72°C for 10 min. GAPDH was used as an internal control. Electrophoresis was carried out on a 2% agarose gel. The multiplex PCR products were analyzed with the Alpha EaseFC software (Alpha Innotech, San Leandro, CA). The experiment was performed six times on each individual (Table 2).

Real-time quantitative PCR analysis (qPCR)
After the induction osteogenic differentiation for 21 days, total RNA was isolated using an RNeasy Mini Kit (QIAGEN, Valencia, CA). Two micrograms of total RNA were reverse-transcribed in order to synthesize cDNA, using an AccuPower RTPReMix kit (Bioneer, Inc., Rockville, MD). For relative quantification, the reactions were performed in a total volume of 20 μL, containing 15 μL of LightCycler ® FastStart DNA Master SYBR Green 1 (Roche Diagnostics, Mannheim, Germany), 10 ng of cDNA, and 10 pmol of each primer. Real-time quantitative PCR was carried out with specific primers, in a LightCycler Instrument (Roche Diagnostics, Mannheim, Germany). The samples were analyzed in triplicate. The primer sequences used are listed in Table 1. GAPDH was used as an internal control. For quantification, the data were analyzed using the LightCycler analysis software (Roche Diagnostics, Mannheim, Germany). Relative quantification of target gene expression was evaluated using the comparative C T method (Wang et al., 2004). The ΔC T value was determined by subtracting the target C T of each sample from its respective GAPDH C T value. Calculation of ΔC T involves using the mean ΔC T value of the control gene as an arbitrary con-HOX gene analysis in the osteogenic differentiation

Statistical analysis
To investigate differentially expressed HOX genes during osteogenic differentiation from hMSCs, the data obtained from multiplex PCR were examined by variance analysis (ANOVA) with SPSS 12.0 software for Windows (SPSS, Chicago, IL). Tukey's HSD test was used for post hoc comparisons. For all statistical tests, an error probability of p < 0.05 was regarded as significant.

Characterization of hMSCs
In an effort to explore the characterization of hMSCs, flow cytometry was used to examine the expression of the surface antigens CD11b, CD29, CD34, CD45, CD73, and CD105 in the isolated hMSCs. The isolated hMSCs were submitted to FACS analysis and found to be positive for CD29 (68 ± 2.5%), CD73 (96.9 ± 2.7%) and CD105 (91.5 ± 2.5%), and negative for CD11b, CD34 and CD45. These results show that the hMSCs were successfully iso-lated and that the culture-expanded hMSCs maintained their phenotype (Figure 1).

Osteogenic differentiation
ALP and von Kossa staining were used to examine the differentiation of hMSCs into osteoblasts in the osteogenic medium. Although ALP staining at day 10 showed a weak color signal, the intensity of ALP activity increased remarkably by day 21. The intensity of von Kossa staining also peaked at day 21 ( Figure 2A). RT-PCR was performed using osteogenic markers to confirm hMSC osteogenesis ( Table 1). The mRNA expression levels of the osteogenic markers, which included bone sialoprotein (BSP), OCN and ALP, were significantly higher at day 21 than at day 0 ( Figure 2B). Immunoblot analysis was performed using OCN and OPN in order to obtain further confirmation of osteogenesis. The results of von Kossa staining and RT-PCR were identical to the result observed with ALP and showed that the expression of the OCN and OPN proteins increased as differentiation progressed ( Figure 2C). All of the corresponding results confirmed that the hMSCs were successfully differentiated into osteoblasts.

Analysis of HOX gene expression using multiplex PCR
Multiplex PCR was used to assess the expression levels of HOX genes during osteogenic differentiation. The expres-Chae et al. HOX gene analysis in the osteogenic differentiation  Statistical analysis revealed that four HOX genes showed significant differences in expression at the transcription level. The HOXA1, HOXC11 and HOXC13 genes were found to be up-regulated. The expression of HOXC13 was unaltered between day 0 and day 10 and only increased after day 10. The expression of HOXA1 gradually increased for 21 days, but the increase in the expression of HOXC13 was more dramatic. The mRNA level of HOXC11 fluctuated during osteogenesis. The expression of HOXC11 increased during the first 10 days of osteogenic differentiation, but then decreased over the next 11 days ( Figure 3A). The expression of HOXD13 was down-regulated during the osteogenesis of hMSCs. The mRNA level of HOXD13 decreased gradually over the 21-day period ( Figure 3B).

Expression of HOXC13 and HOXD13
The expression of HOXC13 and HOXD13 showed the most dramatic change after 21 days of differentiation. The expression of HOXC13 increased by approximately 91%, whereas that of HOXD13 decreased by 50% after osteogenesis. Real-time quantitative PCR and immunoblotting analysis were carried out in order to further confirm the increased expression of HOXC13 and HOXD13. The results of qPCR showed that the expression of HOXC13 was five times higher at day 10 and forty-two times higher at day 21 than in the undifferentiated state, respectively, whereas the mRNA expression of HOXD13 showed a five-fold decrease at day 10 ( Figure 4). These qPCR results of the HOX genes were in agreement with those of multiplex PCR.
The expression levels of these two HOX genes were then submitted to immunoblot analysis to further evaluate their protein level in the osteogenic differentiation of hMSCs. The results showed increased expression of the HOXC13 protein and decreased expression of the HOXD13 protein after 21 days of differentiation ( Figure 5). This result was in agreement with those of multiplex and real-time PCR.

Discussion
Many factors are known to regulate osteogenesis (Bobis et al., 2006). The important factors involved in osteogenic regulation include bone morphogenetic protein (BMP), Chae et al. transforming growth factor (TGF), insulin-like growth factor (IGF), brain-derived growth factor (BDGF), fibroblast growth factor (FGF), leptin and parathyroid hormone-related peptide (PTHrP). These proteins regulate the expression of signals needed for bone remodeling. In addition, many reports have suggested that various transcription factors participate in osteogenesis. Among them, Cbfa1/ Runx2, Osterix, ΔFosB, Fra-1, Aj18, Osf1, Msx2, Dlx5 and TWIST have been shown to play pivotal roles.
Several studies have also reported that HOX genes are involved in osteogenesis. These reports showed that HOXA2 plays several important roles in the process of skeletogenesis (Gendron-Maguire et al., 1993;Rijli et al., 1993;Kanzler et al., 1998). Another study found, using quantitative RT-PCR (Dobreva et al., 2006), that the expression of HOXA2 was up-regulated during osteogenesis. HOXA10 has been shown to contribute to osteogenic lineage determination (Hassan et al., 2007). HOXC8 was reported to be involved in the regulation of osteogenesis through bone morphogenic protein (BMP) pathways (Juan et al., 2006). However, no significant changes in the expression of HOXA2, HOXA10 and HOXC8 were observed in the present study. The differences in these results may be due to the fact that HOXA2 may have been induced during mouse embryogenesis, and HOXA10 and HOXC8 expression were likely induced by BMP. However, in the present study, mesenchymal stem cells were used, and osteogenic differentiation was induced in vitro using dexamethasone, β-glycerophosphate and ascorbic acid.
Knowledge regarding the expression patterns of the HOX genes during osteogenic differentiation may reveal the signal pathway of osteogenesis and may also help in the potential therapeutic application of hMSCs. However, a report regarding the expression profile of HOX genes during osteogenesis has not yet been published. In the present report, 37 HOX genes were investigated in order to determine their expression patterns during the osteogenesis of hMSCs. For this purpose, we performed multiplex PCR, real-time PCR and Western blot analysis. Based on the results, we suggest that four HOX genes, HOXA1, HOXC13, HOXC11 and HOXD13, might be involved in the osteogenic differentiation of hMSCs. HOXA1 is a key gene in skull development, and it is a retinoic acid (RA) direct target gene (Ijichi and Ijichi, 2002). Mice with mutations in the HOXA1 hexapeptide motif show skeletal defects (Remacle et al., 2004). Similar results were reported by Martinez-Ceballos et al. (2005), who showed that the disruption of the HOXA1 gene results in abnormal ossification of the skull. Andrews et al. (1994) reported that osteogenic protein-1 (OP-1), a member of the TGF-β superfamily, induces HOXA1. In addition, recent microarray analyses revealed that BSP and Col1a1, both key markers of osteogenesis, are the target molecules of HOXA1 (Martinez-Ceballos et al., 2005). The results of multiplex PCR showed that HOXA1 was significantly increased during osteogenesis. The results of the present study and those of previous reports suggest that HOXA1 is an important factor involved in the osteogenesis of hMSCs.
In the present study, the expression of HOXC13 showed the largest increase. However, there are no previous reports suggesting a relationship between HOXC13 and osteogenesis. Kulessa et al. (2000) reported that the over-HOX gene analysis in the osteogenic differentiation  HOXC13 and HOXD1. The data were presented as fold changes relative to day 0. The mRNA expression of HOXC13 was five times higher on day 10 and 42 times higher on day 21 compared to the expression in a control. The expression of HOXD13 decreased rapidly at day 10 and slowly increased at day 21. The real-time PCR data were normalized with GAPDH expression. Asterisk (*) indicates a significant increase between two samples (p < 0.05).
Figure 5 -Immunoblotting analysis of HOXC13 and HOXD13. Whole cell proteins obtained on day 0, day 10 and day 21 were blotted onto a nitrocellulose membrane. The protein level of HOXC13 showed a significant increase on day 21, whereas that of HOXD13 decreased gradually during osteogenic differentiation. β-actin was used as control. Asterisk (*) indicates a significant increase between two samples (p < 0.05).