|Home | About | Journals | Submit | Contact Us | Français|
Dental pulp stem cells (DPSCs) are a unique precursor population isolated from post-natal human dental pulp and have the ability to regenerate a reparative dentin-like complex. Canonical Wnt signaling plays a critical role in tooth development and stem cell self-renewal through β-catenin. In this study, the regulation of odontoblast-like differentiation of DPSCs by canonical Wnt signaling was examined. DPSCs were stably transduced with canonical Wnt-1 or the active form of β-catenin, with retrovirus-mediated infection. Northern blot analysis found that Wnt-1 strongly induced the expression of matricellular protein osteopontin, and modestly enhanced the expression of type I collagen in DPSCs. Unexpectedly, Wnt-1 inhibited alkaline phosphatase (ALP) activity and the formation of mineralized nodules in DPSCs. Moreover, over-expression of β-catenin was also sufficient to suppress the differentiation and mineralization of DPSCs. In conclusion, our results suggest that canonical Wnt signaling negatively regulates the odontoblast-like differentiation of DPSCs.
Dental pulp stem cells (DPSCs) are a unique mesenchymal stem cell (MSC) population that is present in the cell-rich zone and core of the pulp (Gronthos et al., 2002). These cells have the ability to differentiate into odontoblast-like cells, pulpal fibroblasts, adipocytes, and neural-like cells (Gronthos et al., 2002). Primary human DPSCs maintain their stem cell properties and continue to express STRO-1, even after cryopreservation and 20 rounds of passaging (Zhang et al., 2006). In the presence of dexamethasone, β-glycerophosphate (β-GP), and inorganic phosphate, DPSCs are capable of differentiating into an odontoblast-like lineage and expressing alkaline phosphatase. This transition is accompanied by deposition and mineralization of collagenous matrix (Liu et al., 2005). While DPSCs have great potential for dentin regeneration and tooth repair, molecular regulation of differentiation of DPSCs is currently poorly understood.
Growing evidence demonstrates that Wnt signaling plays a critical role in development and stem cell self-renewal (Reya and Clevers, 2005). Wnts are secreted glycoproteins that can exert their effects on neighboring cells in a paracrine manner. There are currently 19 identified Wnt family proteins, divided into two main categories, canonical and non-canonical, based on their role in cytosolic β-catenin stabilization (Reya and Clevers, 2005). Canonical Wnts transduce their signals through intracellular β-catenin. In the absence of Wnt proteins, β-catenin is associated with a cytoplasmic complex containing adenomatous polyposis coli (APC), glycogen synthase kinase-3β (GSK-3β), axin, and casein kinase 1. In this complex, GSK-3β constitutively phosphorylates β-catenin, resulting in its ubiquitination and degradation by the 26S proteasome. In the presence of Wnt stimulation, a frizzled receptor and the Wnt co-receptor LDL receptor-related protein 5 (LRP5) or LRP6 transduce signals to inhibit Axin/APC/GSK-3β activity. This inhibition leads to the accumulation of free cytosolic β-catenin. The elevated cytosolic β-catenin can translocate to the nucleus, form a complex with members of the T-cell factor (Tcf)/lymphoid enhancer factor (LEF) family of transcription factors, and activate the expression of Wnt target genes. In contrast, non-canonical Wnts transduce their signal independent of β-catenin (Reya and Clevers, 2005). Genetic studies have found that Wnt/β-catenin signaling plays an essential role in mesenchymal tissue development, including skeletal maturation and tooth formation (Kratochwil et al., 2002; Sasaki et al., 2005; Jarvinen et al., 2006). However, the effects of canonical Wnt signaling on osteogenic differentiation of MSCs and their ability to form bone remain controversial (Boland et al., 2004; de Boer et al., 2004; Gaur et al., 2005; Kang et al., 2007). In this study, we sought to determine whether canonical Wnt/β-catenin signaling regulates DPSC differentiation and mineralization.
DPSCs isolated from human adult third molars were kindly provided by Dr. Songtao Shi at The University of Southern California (Gronthos et al., 2000). DPSCs were cultured in α-modified Eagle’s medium (α-MEM) (Invitrogen, Carlsbad, CA, USA) supplemented with 15% fetal calf serum (FCS; Invitrogen), 10 mM L-ascorbic acid 2-phosphate (AA), 2 mM L-glutamate, 100 units/mL penicillin, and 100 μg/mL streptomycin at 37° C and 5% CO2, as described previously (Gronthos et al., 2000). To generate DPSCs stably expressing Wnt-1 or β-catenin, we generated vector-containing retroviruses in 293T cells (Chen et al., 2001; You et al., 2002). DPSCs were infected with the retroviruses expressing hemagglutinin (HA) HA.11 epitope tagged Wnt-1, HA-β-catenin, or vector control in the presence of 6 μg/mL polybrene (Sigma-Aldrich, St. Louis, MO, USA). Thirty-six hrs after infection, cells were selected with neomycin (600 μg/mL) for 10 days. The resistant clones were pooled, and cells expressing Wnt-1 or β-catenin were confirmed by Western blot analysis. For growth rate analysis, 100,000 cells were plated onto a 15 × 60 mm plate. Plates were then trypsinized at 32, 56, 84, 122, and 172 hrs, and triplicate cell counts were performed visually with the use of a hemacytometer, with trypan blue staining for viable cells.
DPSCs were induced to differentiate in α-MEM supplemented with 10 mM β-glycerophosphate, 10 nM dexamethasone, and 1.8 mM KH2PO4. Induced cells were rinsed with phosphate-buffered saline (PBS) and fixed in 70% ethanol. One mL per well of substrate solution was added; the plate was covered with foil and incubated at 37° C for 30 min. Substrate solution contained the following in 15 mL of Tris-HCl (pH 9.6): 12 mg Fast Blue BB Salt, 4 mg naphthol AS-TR phosphate, and 0.15 mL DMF. The naphthol was first dissolved in the DMF before being mixed with the Tris-HCl. The mix was filtered, and a 50-μL quantity of MgCl2 was added. Quantification of the staining density was performed with Scion Image software (Scion Corp., Frederick, MD, USA).
Induced cells were rinsed with PBS and fixed in 70% ethanol for 1 hr at 4° C. Cells were stained with 40 mM Alizarin red, pH 4.2, at room temperature for 10 min on a shaker. Cells were rinsed 5x with water to remove unbound Alizarin red. PBS was added for an additional 15 min for further reduction of non-specific staining. Quantification of the staining intensity was done with Scion Image software.
DPSCs were harvested and underwent lysis in RIPA buffer (10 mM Tris-HCl, 1 mM EDTA, 1% SDS, 1% Nonidet P-40, 1:100 proteinase inhibitor cocktail, 50 mM β-GP, 50 mM sodium fluoride). We determined extracted lysate concentrations by measuring absorbance at 595 nm using a protein assay solution (Bio-Rad, Hercules, CA, USA). From 20- to 50-μg aliquots of cell lysates per sample were run on a 7.5% SDS-PAGE gel and then transferred to a PVDF membrane (Bio-Rad). Western blot analysis was performed as described previously (Yang et al., 2004). Primary antibodies used were as follows: monoclonal anti-HA.11 (Covance, Berkeley, CA, USA), monoclonal anti-Flag (Sigma, St. Louis, MO, USA), monoclonal anti-human β-catenin (BD Biosciences, San Jose, CA, USA), monoclonal anti-α-tubulin (Sigma), and polyclonal anti-TFIIB (Santa Cruz Biotech, Santa Cruz, CA, USA). Cytoplasmic and nuclear extracts were prepared as described previously (Chen et al., 2001; You et al., 2002).
RNA was harvested with TRIzol reagent (Invitrogen, Carlsbad, CA, USA). A 10-μg quantity of each sample was loaded onto a 1.4% MOPS gel and transferred to a nylon membrane (Bio-Rad). The membrane was cross-linked with UV. The human full-length cDNA probes for OPN, type I collagen, and osteonectin (ON) were obtained from Dr. L. Fisher at the National Institute of Dental and Craniofacial Research (Bethesda, MD, USA). A random-primed labeling kit was used to generate cDNA probes (Amersham Biosciences, Piscataway, NJ, USA), which were then labeled with [α-32P]dCTP (MP Biomedicals, Irvine, CA, USA) and purified on a micro-G-50 Sephadex column (Amersham Biosciences). Hybridization was performed as described previously (You et al., 2001).
Total RNA was extracted from cells at specific time-points with the TRIzol reagent (Invitrogen). From 2 to 5 μg of total RNA was converted to cDNA by means of the SuperScript First Strand Synthesis kit (Invitrogen). The resulting cDNA was diluted 1:20 and used for 20-μL reactions containing CYBR green master mix, dNTPs, primers, and Platinum Taq DNA Polymerase (Invitrogen). As an internal control, 18S cDNA was simultaneously quantified. All experiments were run in triplicate and repeated at least two times to confirm results. Primers for human bone sialoprotein (BSP) were designed based on sequence NM_004967 from the NCBI database and were as follows: 5′-CAGGCCACGATATTATCTT TACA-3′ (forward) and 5′-CTCCTCTTCTTCCT CCTCCTC-3′ (reverse).
To investigate whether DPSCs respond to Wnt stimulation, we stably transduced DPSCs with retroviruses expressing HA-Wnt-1 or control neomycin vector (Fig. 1A). Western blot analysis revealed that β-catenin in the cytosolic and nuclear compartments was increased in Wnt-1-expressing DPSCs (DPSC/Wnt-1) compared with control cells (DPSC/V), suggesting that Wnt-1 stabilizes β-catenin in DPSCs (Fig. 1B). Growth rate analysis revealed that there was no significant difference in the proliferation rates between DPSC/Wnt-1 and DPSC/V cells (Fig. 1C).
To induce cell differentiation, we treated both DPSC/Wnt-1 and DPSC/V cells with dexamethasone and β-glycerophosphate. Type I collagen is the most abundant dentin matrix protein. Thus, we examined whether Wnt-1 regulated the expression of type I collagen. Northern blot analysis found that Wnt-1 slightly enhanced the gene expression of type I collagen (Fig. 2A). Interestingly, while OPN was barely detected in DPSC/V cells, OPN was significantly induced in DPSC/Wnt-1 cells. Upon the induction of differentiation, the level of OPN was slightly reduced in DPSC/Wnt-1 cells (Fig. 2B). ON was expressed at similar levels in DPSC/Wnt1 cells relative to DPSC/V cells (Fig. 2C). Unlike OPN and ON, BSP expression was strongly induced after differentiation of DPSC/V cells for 7 to 14 days, with only a transient induction noted in DPSC/Wnt-1 cells, as determined by qPCR (Fig. 2D).
ALP activity is an early marker of both odontoblastic and osteoblastic differentiation and plays an important role in mineralization. We examined whether Wnt-1 affected ALP activity in DPSCs. ALP activity was strongly detected in DPSC/V cells after 5 days of culture in differentiation medium, but not in DPSC/Wnt-1 cells (Fig. 3A). ALP activity in DPSC/Wnt-1 cells was suppressed by more than 90% when compared with that in DPSC/V cells. Wnts are secreted glycoproteins that act by binding to cell-surface receptors (Reya and Clevers, 2005). Thus, we generated co-cultures of DPSC/V and DPSC/Wnt-1 cells to determine whether secreted Wnt-1 could inhibit ALP activity in DPSC/V cells. DPSC/V cells were co-cultured with DPSC/Wnt-1 cells, with initial DPSC/V cells plated ranging from 0% to 100% of the total culture (Fig. 3B). For comparison, the intensity of the staining in the pure DPSC/V cells was set to 100%, and the intensity of the pure DPSCs/Wnt-1 cells was set to 0%. The mixture of 20% DPSC/Wnt-1 cells with 80% control DPSC/V cells significantly inhibited over 65% of the ALP activity. Similarly, in a culture with 50% DPSC/V cells and 50% DPSC/Wnt-1 cells, over 90% of the ALP activity was inhibited (Fig. 3B). These results suggest that secreted Wnt-1 can inhibit ALP activity in DPSCs.
Since ALP activity plays a critical role in mineralization, we examined whether Wnt-1 impaired mineralization of DPSCs. After the prolonged induction of differentiation, we stained cells with Alizarin red to detect the level of calcium mineral deposition. While DPSC/V cells exhibited a staining density of 0.86 ± 0.073 pixels per area, DPSCs/Wnt-1 cells demonstrated a staining density of only 0.086 ± 0.088 pixels per area (Fig. 3C). Moreover, we found that secreted Wnt-1 could also inhibit the mineral formation of DPSCs in a manner similar to that demonstrated above with ALP (data not shown).
Wnt-1 is a founder member of the Wnt family of proteins. Previously, studies from our group and others have demonstrated that Wnt-1 signals through β-catenin. To determine whether β-catenin was sufficient to suppress the differentiation of DPSCs, we stably transduced DPSCs with a mutant S37Aβ-catenin (Fig. 4A). The S37Aβ-catenin contains a serine-to-alanine mutation that protects the β-catenin from being phosphorylated and targeted for degradation by GSK3-β. Therefore, it is constitutively active within the cell. As was observed with DPSC/Wnt-1 cells, DPSC/β-cat cells failed to express ALP when compared with control cells after induction (Fig. 4B). While DPSC/V cells were highly mineralized, with a staining density of 0.999 pixels per area, S37Aβ-catenin-expressing DPSCs (DPSC/β-cat) showed poor mineralization and had a staining density of only 0.053 pixels per area (Fig. 4C).
The results reported here are the first demonstration that canonical Wnt signaling can inhibit the odontoblast-like differentiation of DPSCs. Recent work from several groups has elucidated the importance of the canonical Wnt/β-catenin signaling pathway in skeletal development and the post-natal maintenance of bone mass (Krishnan et al., 2006). Loss-of-function mutations in human LRP5, the canonical Wnt co-receptor, result in osteoporosis-pseudoglioma syndrome, in which persons have a low bone mineral density and skeletal fragility (Gong et al., 2001). In contrast, activating mutations in human LRP5 have resulted in increased bone density in affected persons (Boyden et al., 2002). Several elegant studies with knock-out mice have also confirmed that the canonical Wnt/β-catenin signaling pathway is essential for bone development as well as the post-natal maintenance of bone mass (Day et al., 2005; Holmen et al., 2005; Krishnan et al., 2006). These important discoveries suggest that the activation of the Wnt/β-catenin signaling pathway may help to improve bone or dental tissue engineering. Because human adult bone marrow MSCs are one of the most promising adult stem cell populations for bone regeneration and repair, several studies have examined whether the canonical Wnt proteins, such as Wnt-1 or Wnt-3a, could enhance the osteogenic differentiation of MSCs. Paradoxically, several studies have found that Wnt-1 or Wnt-3a strongly inhibited the osteogenic differentiation of MSCs (Boland et al., 2004; de Boer et al., 2004; Cho et al., 2006). Consistent with these studies, we found that canonical Wnt signaling inhibits the odontoblast-like differentiation of DPSCs. Thus, targeting the canonical Wnt/β-catenin signaling pathway to enhance mineralized tissue regeneration may require further investigation.
Odontoblast-like differentiation of DPSCs has been assumed to be similar to that observed with bone marrow MSCs. Osteogenic differentiation is characterized by an increase in ALP expression within the first 4 days (Cheng et al., 1994). This is followed by a deposition of collagenous matrix and the subsequent maturation and mineralization of the matrix (Jaiswal et al., 1997). Since Wnt-1 did not inhibit the production of extracellular matrix, it is likely that the inhibition of mineralization by Wnt-1 is due to the lack of ALP activity. Studies of crystal nucleation have suggested that OPN is a potent inhibitor of mineral formation and mineral crystal growth and proliferation at high concentrations (Boskey et al., 2002). Interestingly, Wnt-1 strongly induces OPN expression in DPSCs. Therefore, it is possible that the induction of OPN by Wnt-1 may play a role in the suppression of mineral formation in DPSCs. Interestingly, we could not find that β-catenin induced OPN expression (data not shown). Thus, it is unlikely that OPN is an essential molecule in the inhibition of DPSC differentiation mediated by the canonical Wnt signaling pathway. While Wnt-1 did not affect ON expression, we found that Wnt-1 partially inhibited BSP expression. Thus, Wnt signaling may impair mineralization by inhibiting BSP. Finally, since the Wnt/β-catenin signaling pathway regulates adult stem cell properties, it is possible that canonical Wnt signaling may help to maintain MSCs in an undifferentiated state, thereby inhibiting the odontoblast-like differentiation of DPSCs. In the future, it will be interesting to develop a model system to test whether canonical Wnt signaling regulates the stem cell properties of DPSCs.
This work was supported by grants R01DE 016513 (C.-Y. W) and T32DE 007057 (E. S.) from the National Institute of Dental and Craniofacial Research.