Bone Morphogenetic Protein Regulation of Early Osteoblast Genes in Human Marrow Stromal Cells Is Mediated by Extracellular Signal-Regulated
http://www.100md.com
内分泌学杂志 2005年第8期
Department of Biochemistry, University of Pennsylvania School of Dental Medicine, Philadelphia, Pennsylvania 19104
Address all correspondence and requests for reprints to: Dr. Anna M. Osyczka, Department of Biochemistry, University of Pennsylvania School of Dental Medicine, 240 South 40th Street, Philadelphia, Pennsylvania 19104-6030. E-mail: annamo@biochem.dental.upenn.edu.
Abstract
Bone marrow stromal cells (MSC) are the major source of osteoblasts for bone remodeling and repair in postnatal animals. Rodent MSC cultured with bone morphogenetic proteins (BMPs) differentiate into osteoblasts, but most human MSC show a poor osteogenic response to BMPs. In this study we demonstrate that BMP-induced osteogenesis in poorly responsive human MSC requires modulation of ERK and phosphatidylinositol 3-kinase (PI3-K) pathways. Either treating human MSC cultures with the MAPK/ERK kinase inhibitor PD98059 or transferring them to serum-free medium with insulin or IGF-I permits BMP-dependent increases in the expression of the early osteoblast-associated genes, alkaline phosphatase and osteopontin. Increased expression of these genes in BMP-treated, serum-free cultures correlates with increased nuclear levels of activated Smads, whereas serum-free cultures of human MSC expressing constitutively active MAPK/ERK kinase show decreased expression of early osteoblast genes and decreased nuclear translocation of BMP-activated Smads. Inhibiting ERK activity in human MSC also elevates the expression of Msx2, a transcription factor that is directly regulated by Smad-binding elements in its promoter. Therefore, growth factor stimulation leading to high levels of ERK activity in human MSC results in suppressed BMP-induced transcription of several early osteoblast genes, probably because levels of BMP-activated nuclear Smads are decreased. In contrast, inhibiting the insulin/IGF-I-activated PI3-K/AKT pathway decreases BMP-induced alkaline phosphatase and osteopontin expression in serum-free cultures of human MSC, but increases BMP activation of Smads; thus, PI3-K signaling is required for BMP-induced expression of early osteoblast genes in human MSC either downstream or independent of the BMP-activated Smad signaling pathway.
Introduction
REPAIR OF ADULT bone fracture involves bone marrow-derived mesenchymal cells, which serve as a source of osteochondral progenitors that invade the fracture site, proliferate, and differentiate into cartilage and bone. These mesenchymal progenitors [also known as mesenchymal stem cells or marrow stromal cells (MSC)] can be isolated from bone marrow and cultured under conditions that promote their in vitro differentiation into specific mesenchymal phenotypes, such as chondrocytes or osteoblasts (1, 2). To achieve the osteoblastic phenotype, progenitor cells need to express several bone-type extracellular matrix proteins, such as collagen type I, osteopontin (OP), bone sialoprotein, and osteocalcin, and produce high levels of the ecto-enzyme tissue-nonspecific alkaline phosphatase (ALP). High-level expression of ALP is required for mineralization of skeletal tissues (3, 4) and is induced early during osteoblast differentiation (5, 6). In both animal and human MSC cultures, ALP and OP serve as useful markers of early osteogenesis, and the expression of these genes usually increases by the end of the first week of culture. Among many transcription factors expressed early in osteogenesis, Runx2 (cbfa1) is noteworthy, because it is required for bone formation and is an important early indicator of osteogenic capacity of cells (7, 8).
Bone morphogenetic proteins (BMPs) are osteoinductive growth factors that can induce bone formation both in vivo and in vitro (9, 10, 11). A number of clinical studies have assessed the efficacy of recombinant human BMPs in the healing of critical-size bone defects and the acceleration of bone fracture healing in humans (9, 12). Although the latter studies have been promising, relatively high doses of BMPs were required to induce adequate bone formation compared with animal models, and large variations in response among individual patients were observed (13). In rodent cell cultures, BMPs have shown a potent osteogenic activity (14, 15, 16, 17), but their ability to induce osteogenesis of human cells is less clear (8, 18). Consistent with these studies, we have recently reported major differences in the BMP response of rodent and human MSC cultures. Although addition of BMPs to rat MSC cultures resulted in a marked stimulation of ALP mRNA and enzyme activity, 90% of human MSC cultures stimulated with BMP did not show elevated ALP despite transcriptional regulation of several other BMP-responsive genes (19, 20). Jorgenson et al. (21) recently reported similar data indicating that BMP-treated human MSC do not show elevated ALP. We have now identified culture conditions in which BMP treatment of human MSC results in high ALP expression and profound stimulation of OP.
The classic BMP signaling pathway operates by activation of the Smad family of transcription factors, and there is evidence that it can also act through a Smad-independent p38 MAPK signaling pathway (22). There have also been reports suggesting that other kinase pathways, such as ERK, c-Jun N-terminal kinase, phosphatidylinositol 3-kinase (PI3-K), Wnt, and nuclear factor-B, may substitute for, activate, or modulate BMP signaling (23, 24). In this study we present evidence that in BMP-unresponsive human MSC cultures, growth factor signaling leading to high and sustained activity of ERK negatively regulates BMP-mediated Smad signaling. Furthermore, human MSC require PI3-K/AKT activation to achieve BMP-stimulated, high-level expression of several genes associated with the early stages of osteoblast differentiation.
Materials and Methods
Chemicals
Unless otherwise stated, all chemical reagents were purchased from Sigma-Aldrich Corp. (St. Louis, MA), and tissue culture solutions were obtained from Invitrogen Life Technologies, Inc. (Grand Island, NY). ITS+ Premix, composed of insulin, transferrin, selenious acid, linoleic acid and BSA, was obtained from BD Biosciences (Bedford, MA). Ascorbate-2-phosphate was obtained from Wako Chemicals (Richmond, VA), and the ERK inhibitor UO126 was purchased from BIOMOL Research Laboratories (Plymouth Meeting, PA). Proteinase inhibitor cocktails were purchased from Calbiochem (San Diego, CA). The PI3-K inhibitor Ly294002, primary rabbit polyclonal antibodies against p44/42 MAPK (ERK); phospho-p44/42 MAPK (p-ERK); phospho-Smad1, -5, and -8; AKT; and phospho-AKT were purchased from Cell Signaling Technology (Beverly, MA). Primary rabbit antihuman MADR1/Smad1 antibodies (recognizing both human Smad1 and -5) were obtained from Upstate Biotechnology, Inc. (Lake Placid, NY). Recombinant human bone morphogenetic protein type 2 (BMP-2) was provided by Wyeth/Genetics Institute (Cambridge, MA). Adenovirus containing constitutively active MAPK/ERK kinase 1 (caMEK) was a gift from Dr. Janet Rubin, Emory University (25).
Cell culture, inhibitor studies, and adenoviral infections
Human marrow samples were isolated from the proximal medullary cavities of femurs from patients undergoing total hip replacement. Primary cultures were established as described previously (19, 20) with a seeding density of 5 x 105 cells/cm2. Cells were grown in MEM supplemented with 10% pretested fetal bovine serum (FBS; Premium Select, Atlanta Biologicals, Norcross, GA) and antibiotics at 37 C in a humidified, 5% CO2 atmosphere. Medium was changed initially on d 4, then every other day thereafter. At confluence, cells were detached using 0.25% trypsin in 1 mM tetrasodium EDTA, centrifuged, resuspended in fresh MEM/10% FBS and plated at 104 cells/cm2.
All assays were carried out on first or second passage cell cultures. One day after plating the medium was exchanged for fresh medium supplemented with 0.35 mM L-ascorbic acid-2-phosphate (ascorbate), 10–7 M dexamethasone (Dex), or 100 ng/ml BMP-2 as noted. On d 4, cells were washed with Hanks’ buffered saline solution (HBSS), and medium was changed to one of the following: 1) fresh MEM containing 10% FBS; 2) MEM supplemented with ITS+ Premix, insulin at 1.1 μM, or IGF-I at 13.1 nM (SF+I); or 3) MEM supplemented only with 1.25 mg/ml BSA. Osteogenic supplements were then replaced. Medium with 10% FBS was estimated to contain 3–6 nM insulin and, based on values in the literature, an additional 2 nM IGF-I. ERK inhibitors PD98059 and UO126, PI3-K inhibitor Ly294002, and cycloheximide or respective vehicle solution were added on d 4. Inhibitors were added 1–2 h before the addition of ascorbate, BMP-2, or Dex. For infections with adeno-caMEK or empty adenovirus, cells were switched to SF+I on d 4, infected on d 5, and analyzed by real-time PCR on d 6. Cells were infected using a multiplicity of infection of 10. At this multiplicity of infection, the proportion of adenovirus-infected MSC is usually 40–70% without adenovirus-mediated cell toxicity.
ALP assay
Immediately before cell harvest for ALP assay, the numbers of viable cells were estimated using CellTiter 96 Aqueous One Solution Cell Proliferation Assay (Promega Corp., Madison, WI). Cells in 24-well culture plates were washed once with HBSS and covered with 200 μl of a solution prepared as a 1:10 (vol/vol) dilution of the reagent containing the MTS tetrazolium salt, [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt, and the electron-coupling reagent, phenazine ethosulfate, in phenol red-free MEM. Cells were incubated for 15 min at 37 C in a humidified, 5% CO2 atmosphere, MTS-containing medium from each well was transferred to 96-well plates, and the absorbance was measured at 490 nm. After washing twice with HBSS, cells were lysed in 10% Triton X-100, and stored at 4 C until ALP assay. ALP activity was determined kinetically as described previously (14).
Isolation of total RNA, cDNA synthesis, and real-time PCR analysis
Total RNA was extracted using Tri-Reagent (Molecular Research Center, Cincinnati, OH). Aliquots of total RNA (2 μg) were reverse transcribed, and cDNA samples corresponding to 50–250 ng input RNA were amplified as described previously (19). The reaction mixtures contained 2–3 μM forward and reverse primers and 3–4 mM MgCl2. Table 1 lists the primer sequences and annealing temperatures individually optimized for each primer pair. A final melt curve from 60–95 C was performed to confirm the specificity of the PCR, and the identities of PCR products were confirmed by gel electrophoresis. For calculating mRNA levels, results from real-time RT-PCR were converted to arbitrary units of mRNA, assuming a concentration-dependent straight line for a semilog plot, with a value of 3.5 for the fold change in mRNA/cycle (slope), and the crossing point cycle number with no cDNA template as an estimate of the y-intercept.
TABLE 1. Oligonucleotide primers for real-time RT-PCR analyses
Western blot analyses
Whole cell extracts were obtained by lysing cells in cell lysis buffer provided by Cell Signaling Technology. Nuclear extracts were obtained by scraping cells in Tris-buffered saline, centrifuging and resuspending pellets in ice-cold buffer composed of 10 mM HEPES (pH 8.0), 10 mM KCl, 1 mM EDTA (pH 8.0), 1 mM EGTA (pH 8.0), 1 mM dithiothreitol, 2 mM phenylmethylsulfonylfluoride, 0.5 mM sodium orthovanadate, 1 mM tosyl-L-phenylalanine, and 10 mM ethylmaleimide. After 15min incubation on ice, NP-40 was added and nuclei pelleted by centrifugation at 16,000 x g. The nuclear pellets were resuspended in cold buffer composed of 20 mM HEPES (pH 8.0), 1 mM EDTA, 1 mM EGTA, 0.4 M NaCl, 1 mM dithiothreitol, 2 mM phenylmethylsulfonylfluoride, 0.5 mM sodium orthovanadate, 0.1 mM N-tosyl-L-phenylaline chloromethyl ketone, and 10 mM N-ethylmaleimide. The protein concentration was determined with the MicroBCA protein assay reagent kit (Pierce Chemical Co., Rockford, IL). Forty to 80 mg protein from whole cell extract or 10–20 mg protein from nuclear extracts was separated on NuPAGE 4–12% bis-Tris gels (Invitrogen Life Technologies, Inc., Carlsbad, CA) under reducing conditions using 4-morpholinepropanesulfonic acid buffer and then transferred to immunoblot polyvinylidene difluoride membranes (Bio-Rad Laboratories, Hercules, CA). Membranes were incubated overnight with primary antibodies diluted 1:1,000 and then exposed for 2 h to horseradish peroxidase-linked antirabbit IgG (Amersham Biosciences, Piscataway, NJ) diluted 1:1250. The peroxidase-based signal was detected using Western Lightning Chemiluminescence Reagent Plus (PerkinElmer, Boston, MA) using a Kodak Image Station 440CF (Eastman Kodak, Rochester, NY).
Results
BMP induction of ALP activity in human MSC cultures requires serum withdrawal and addition of insulin or IGF-I
We have previously reported that the majority of human MSC isolates show a relatively poor osteogenic response to BMPs in vitro compared with similar rodent MSC cultures (20, 26). We therefore examined whether altering the culture conditions of the poorly responsive human MSC would elicit a more robust osteogenic response to BMP-2. The effects of BMP-2 were compared in the presence and absence of serum, with and without 1.1 μM insulin or 13.1 nM IGF-I. As shown in Fig. 1, BMP-2 increased ALP activity on d 7 approximately 4-fold when human MSC were cultured for the last 3 d in serum-free medium (SF) with IGF-I or insulin, but not when they were maintained in serum-containing medium (SC). Decreasing the serum concentration to 2% did not have any beneficial effect on BMP-2 stimulation of ALP activity (data not shown). Human MSC cultured for up to 14 d in SC failed to show BMP-induced ALP increase with BMP-2, BMP-4, or BMP-7 and with several sources of FBS (data not shown).
FIG. 1. Effects of IGF-I and insulin on BMP-2 stimulation of ALP activity in human MSC cultures grown in either SF or SC. Human MSC cultures were treated from d 1 with BMP-2 or were left untreated. On d 4, culture medium was changed to fresh SF or SC, with or without 13.1 nM IGF-I or 1.1 μM insulin, followed by appropriate treatments. ALP activity (mean ± SD) was determined on d 7. *, Significantly higher than without BMP, P < 0.01. #, Significantly lower than corresponding SF samples, P < 0.005.
Dose-dependency studies showed that at 6–11 nM, insulin and IGF-I had similar effects on human MSC in SF cultures, enhancing BMP-stimulated ALP activity approximately 2- to 4-fold (Table 2). Addition of either insulin or IGF-I slightly increased the number of cells in SF human MSC cultures, but adding BMP-2 to insulin or IGF-I-treated cultures decreased cell number 10–20% (Table 3). The net effect of insulin or IGF-I in combination with BMP-2 was to increase ALP levels without a significant increase in proliferation. Therefore, the effect of adding insulin or IGF-I to BMP-treated cultures was to increase ALP activity per cell, rather than produce greater numbers of ALP-positive cells. Similar data were obtained using SF supplement ITS+ Premix, which contains 1.1 μM insulin. Unless stated otherwise, additional studies were performed using SF in which insulin was supplied as ITS+ Premix.
TABLE 2. ALP activity of human MSC continuously cultured in serum or after transfer to SF with insulin or IGF-I
TABLE 3. Numbers of viable human MSC continuously cultured in serum or transferred for 3 d to SF with insulin or IGF-I
Switching human MSC cultures to SF+I promotes BMP-2 induction of ALP, OP, and Msx2 mRNAs
Comparisons of gene expression in 7-d cultures grown for the last 3 d in SF+I or continuously cultured in SC are presented in Fig. 2. Transfer to SF+I had a marked effect on BMP-2 stimulation of genes expressed early in osteoblast differentiation. ALP mRNA, which was unaffected by BMP treatment in the presence of serum, was increased by BMP-2 if cultures were grown for the last 3 d in SF+I. Previous studies (26) demonstrated that serum-containing cultures of human MSC showed BMP-induced increases in mRNA for the matrix protein osteopontin and the transcription factor Msx2. Studies with cyclohexamide treatment of human MSC cultures indicated that new protein synthesis was required for BMP-2 induction of ALP and osteopontin mRNAs, but not for Msx2 mRNA (data not shown). This is consistent with the assumption that BMP induction of Msx2 expression in human MSC is a direct result of Smad activation. In the present work, both osteopontin and Msx2 mRNA levels were modestly stimulated with BMP-2 in serum-containing MSC cultures, but BMP-2 had a far greater effect if cells were switched to SF+I. Analyses at increasing times after transfer to SF+I revealed that levels of ALP mRNA additionally increased with longer exposure to SF+I, and that at least 10 h of exposure to BMP-2 were required to obtain enhanced ALP mRNA expression in SF+I. This relatively long exposure time is consistent with evidence from cyclohexamide studies that BMP does not regulate ALP simply by binding of BMP-activated Smads to a regulatory region of the ALP gene.
FIG. 2. mRNA levels of osteoblast genes in human MSC cultured in SF+I or SC (SERUM). Cells were treated from d 1 of culture with BMP-2 or Dex or were left untreated (CTR). On d 4, culture medium was exchanged to SF+I or SERUM, followed by appropriate treatment. RNA was extracted from cells on d 5 for Msx2 or d 7 for other genes and analyzed by real-time RT-PCR. Values are the mean ± SD, expressed in mRNA arbitrary units. *, P < 0.03 compared with corresponding untreated; **, P < 0.001 compared with corresponding untreated; #, P < 0.03 compared with SF+I.
Analyses of mRNA levels of other early osteoblast genes revealed that changing from SC to SF conditions did not affect collagen type I mRNA expression (data not shown), whereas levels of Runx-2 mRNA were increased 2- to 3-fold by BMP-2 in both SC and SF+I cultures. Analyses of late differentiation markers indicated that bone sialoprotein mRNA was increased by BMP-2 in SC cultures, but transfer to SF+I medium significantly decreased BSP mRNA expression with and without inducers. BMP-2 treatment did not alter osteocalcin mRNA levels regardless of whether serum was present; however, it should be noted that vitamin D3 was not added (27). Dex, which increased ALP expression of human MSC cultured in serum, did not increase ALP mRNA in SF+I conditions.
High ERK activity negatively regulates BMP-2 stimulation of ALP, OP, and Msx2 expression
Because SF lacks growth factors that activate the ERK pathway, we examined whether sustained activation of ERK would prevent BMP stimulation of gene expression in SF+I cultures of human MSC, using a caMEK adenoviral construct that phosphorylates ERK. As shown in Fig. 3, A–C, expressing caMEK in SF+I cultures markedly decreased ALP, OP, and Msx2 mRNA levels. Conversely, inhibiting ERK signaling with 50 μM PD98059 significantly increased ALP, OP, and Msx2 mRNA expression in BMP-stimulated SC cultures (Fig. 3, D–F). Although addition of insulin or IGF-I to SC human MSC cultures did not permit BMP stimulation of ALP (Fig 1A), adding either insulin or IGF-I in the presence of 50 μM PD98059 slightly enhanced BMP induction of ALP activity in these cultures (data not shown).
FIG. 3. mRNA levels of early osteoblast genes under conditions modulating ERK activation. A–C, Human MSC were transferred to SF+I medium on d 4 and infected with either empty adenovirus or adenovirus containing caMEK1 on d 5, and RNA was harvested on d 6. Values from real-time RT-PCR are expressed as the ratio of experimental mRNA to that for glyceraldehyde-3-P dehydrogenase for the same sample. This experiment was repeated twice with two different cell isolates, with similar results. D–F, Human MSC were cultured in 10% FBS (SERUM) and treated with 50 μM PD98059 on d 4, and RNA was extracted on d 6. Values, expressed in arbitrary units of mRNA, are the mean ± SD for RNA derived from at least three different MSC isolates. *, Significantly higher than without BMP, P < 0.01; #, significantly higher than without PD98059 inhibitor, P = 0.001.
Western blot assays for phosphorylated ERK, using whole cell extracts harvested at 5, 15, 45, and 90 min, showed that ERK activation peaked 15 min after medium change in both SC and SF culture conditions. At this time point, ERK activation was the highest in human MSC cultured in SC without ERK inhibitor (Fig. 4), and BMP had no significant effect on ERK activation in these cultures. In SF cultures, addition of either insulin or BMP-2 caused a slight increase in ERK activation at 15 min, but the levels of phosphorylated ERK in these culture conditions were markedly lower than those in SC cultures. Similar results were obtained using an in vitro kinase assay to measure 32P incorporation into the ERK substrate Elk-1 (data not shown).
FIG. 4. ERK activation in human MSC cultures. Cells were treated from d 1 of culture with BMP-2 or were left untreated (Control). On d 4, medium was replaced with SC or SC supplemented with 50 μM PD98059 (SC+PD) or was changed to SF+I) or SF without insulin (SF–), followed by appropriate treatments. Whole cell extracts were prepared 15 min after medium changes and analyzed by Western blot. A, Western blot analysis of phosphorylated p44/42 (p-ERK) and total ERK. B, Densitometry quantification of activated ERK, expressed as a ratio to total ERK.
Human MSC with low ERK activity have high levels of phospho-Smad1, -5, and -8 and greater Smad nuclear localization
To obtain additional insights into whether SF culture conditions promote BMP-mediated Smad signaling, Smad1, -5, and -8 activation was analyzed in whole cell and nuclear extracts using a polyclonal antibody that recognizes Smad1, -5, and -8 phosphorylated at the C-terminal BMP activation domain. Western blot analysis of whole cell extracts showed that BMP-2 increased phosphorylation of Smad1, -5, and -8 in both SC and SF culture conditions, but the activation of BMP-regulated Smads was markedly greater in SF cultures (Fig. 5, A and B). Switching cells to SF conditions also increased levels of activated Smad1, -5, and -8 in the nucleus (Fig. 5, C and D), but adding insulin had no additional effect on the amounts of activated Smads. These results suggested that Smad activation was promoted by serum removal, and the insulin/IGF-I requirement was unrelated to Smad activation. Increasing ERK signaling in SF+I cultures by overexpressing caMEK did not affect the whole cell levels of activated Smads (Fig. 5E), but decreased the levels of activated Smads in the nucleus (Fig. 5F).
FIG. 5. Levels of BMP-activated Smads (P-Smads) in human MSC cultured in SC or SF. Cells were treated from d 1 of culture with BMP-2 or were left untreated (Control). On d 4, medium was changed to SC, SF+I, or SF without insulin (SF–), followed by appropriate treatments. Proteins were extracted 1 or 3 h later and analyzed by Western blot. A, Western blot analysis of activated and total Smads in whole cell extracts 1 h after medium changes. B, Densitometric quantification of activated Smads in whole cell extracts, expressed as a ratio to total Smads. C, Western blot analysis of activated and total Smads in nuclear extracts 3 h after medium changes. D, Densitometric quantification of activated Smads in nuclear extracts, expressed as a ratio to total Smads. E and F, Cells were infected with either empty adenovirus or caMEK1 (CA-MEK) on d 4, and 6 h later were transferred to SF+I medium with or without BMP-2. E, Western blot analysis of activated and total Smads in whole cell extracts obtained at the indicated time intervals after medium changes. F, Western blot analysis of activated and total Smads in nuclear extracts prepared 3 h after medium change.
PI3-K activation is necessary for the expression of early osteoblast genes, but not Msx2
The finding that insulin or IGF-I was necessary for BMP induction of ALP in SF cultures suggested that the PI3-K/AKT pathway might be implicated in BMP-stimulated osteogenesis. The PI3-K inhibitor Ly294002, at 5–10 μM, decreased both BMP-stimulated ALP activity and ALP mRNA expression in SF+I cultures (Fig. 6, A and B). A similar inhibitory effect of Ly294002 was seen for BMP-stimulated OP mRNA expression (Fig. 6C), but not for Msx2 mRNA expression (Fig. 6D). The efficacy of the Ly294002 inhibitor was confirmed by Western blot analysis of AKT phosphorylation, which is mediated by PI3-K (Fig. 7, A and C). To determine whether PI3-K inhibitor affects BMP-mediated Smad signaling in SF+I cultures, activation of Smad1, -5, and -8 was analyzed in whole cell extracts from SF+I cultures treated with Ly294002. As shown in Fig. 7, C and D, activation of Smad1, -5, and -8 increased with BMP-2 and was further elevated when Ly29004 inhibitor was added.
FIG. 6. Effect of PI3-K inhibitor on BMP induction of early osteoblast genes. On d 4 of culture, human MSC were transferred to SF+I with and without 10 μM Ly294002, followed by appropriate treatment. RNA was prepared on d 5 or 6 and analyzed by real-time RT-PCR. Values (mean ± SD for three different MSC isolates) are expressed in arbitrary units. *, Significantly higher than without BMP, P < 0.01; **, significantly higher than without BMP, P = 0.02; #, significantly lower than corresponding sample without Ly294002 inhibitor, P < 0.01.
FIG. 7. Effect of PI3-K inhibitor on BMP activation of Smads. On d 4 of culture, human MSC were transferred to SF+I with and without 10 μM Ly294002, followed by appropriate treatment. Whole cell extracts were prepared at the indicated time intervals after medium change and analyzed by Western blot. A, Western blot analysis of activated (P-AKT) and total AKT. B, Western blot analysis of activated (P-Smads) and total Smads. C, Densitometric quantification of activated AKT, expressed as a ratio to total AKT. D, Densitometric quantification of activated Smads, expressed as a ratio to total Smads.
Discussion
We have previously reported differences in the BMP-mediated osteogenic response of human and rodent bone marrow-derived mesenchymal stem cells (19, 20, 26). Among 26 bone marrow cell samples obtained from the femoral cavity of adult human patients, 90% of isolates failed to increase ALP expression in response to 100 ng/ml BMP-2. In contrast, rat and mouse MSC cultures significantly increase expression of ALP and other osteoblast-related genes with as little as 30 ng/ml BMP-2 (5, 6, 17, 26, 28). In this study we show that these poorly responsive human MSC markedly elevate both ALP and OP expression when cultures are stimulated with BMP-2 in the presence of ERK inhibitor or in SF containing either insulin or IGF-I. Our examination of signaling pathways implicated in BMP regulation of human MSC osteogenesis indicates not only a negative role of ERK, but also a positive role of PI3-K/AKT signaling in the regulation of ALP and OP. Analyses of other osteoblast-related markers suggest that these kinase pathways may modulate the transcription of some genes expressed early in osteoblast differentiation without affecting other osteoblast phenotypic markers. This is consistent with the assumption that osteoblast differentiation is not regulated by a single set of factors coordinately controlling all osteoblast genes and that it is not necessary for progenitor cells to express all genes seen early in osteoblast differentiation to up-regulate some late osteoblast markers (29).
The role of ERK activation in osteoblast differentiation has been the subject of many studies, but no clear picture has emerged. In BMP-treated human and mouse osteoblastic cells, ERK activation is required for increased expression of the late osteoblast gene, osteocalcin (23, 30); however, activated ERK decreases ALP expression (31, 32, 33). This is consistent with our results indicating that high levels of ERK activation have a negative effect on BMP induction of early osteoblast genes in human MSC. In C3H10T1/2 cells, an embryonic murine cell line frequently considered equivalent to MSC, BMP stimulates both ERK expression and ERK activity, and reducing ERK activation decreased BMP-induced ALP activity (24). In contrast, our studies with human MSC showed no significant BMP-2 effect on total ERK levels either in the presence of serum or after transfer of cultures to SF+I. In SF cultures, BMP-2, insulin, or both slightly increased phosphorylation of ERK, but the major factor increasing ERK phosphorylation was the presence of serum. These results combined with the observed effects of caMEK in SF cultures and ERK inhibitors in the presence of serum argue that in human MSC lower levels of activated ERK are associated with improved BMP-2 stimulation of ALP and OP expression.
A positive effect of ERK activation on osteoblast differentiation has been observed with studies using other osteogenic stimuli. In dexamethasone-treated human MSC cultures, there is sustained ERK activation throughout osteogenic differentiation (34). Oscillatory fluid flow increases OP mRNA in MC3T3-E1 cells, and this stimulation is abolished with ERK inhibitors, suggesting that ERK activation plays a positive role in osteopontin induction (35). These reports suggest that the negative effect ERK activation has on the expression of early osteoblast genes in human MSC may be limited to osteogenesis induced by BMP. In epithelial cells overexpressing Smad1, ERK directly regulates BMP-activated Smads by phosphorylation at serine residues within the Smad linker region; this phosphorylation inhibited both nuclear accumulation of Smad1 and its transcriptional activity (36). Recent evidence indicates that linker phosphorylation of Smad1 is important in vivo (37). Several other Smads can be phosphorylated by ERK in the linker region, and whether it results in positive or negative consequences for Smad activity appears to be cell specific (38, 39). Using a p-Smad1, -5, and -8 antibody that detects only BMP-regulated phosphorylation, we have shown that increased ERK activation resulted in decreased nuclear levels of BMP-activated Smads. It is therefore plausible that in human MSC, ERK phosphorylation of BMP-related Smads in the linker region limits the ability of BMP-activated Smads to function as transcription factors. This hypothesis is supported by the increased Msx2 mRNA levels seen with BMP-treated human MSC cultured either in SF conditions or with ERK inhibitors in serum. BMP-2 has been shown to directly regulate Msx2 expression by Smad binding to its promoter (40, 41), and we have found that BMP stimulation of Msx2 in human MSC does not require new protein synthesis. The fact that BMPs directly regulate the Msx2 promoter by activated Smads together with our observation that BMP-stimulated Msx2 expression is modulated by ERK activation imply that levels of activated ERK are modulating BMP-activated Smad function in human MSC.
In contrast to ERK, activation of the PI3-K/AKT pathway by insulin or IGF-I seems to positively regulate BMP-2-induced ALP and OP in human MSC cultures. Both insulin and IGF-I are highly specific for their respective receptors, and a 100-fold excess of the incorrect ligand is required to displace the correct one (42). The fact that either insulin or IGF-I, at roughly equimolar concentrations, is capable of permitting a BMP response implies that activation of PI3-K signaling by either ligand is sufficient. This assumption is reinforced by the finding that blocking PI3-K signaling abolishes the BMP osteogenic response. These results support and extend previous data showing that PI3-K and AKT serine/threonine kinase are required for BMP-induced osteogenesis in 2T3 cells (43) and fetal rat calvaria cells (44). Blocking PI3-K/AKT signaling has been reported to also inhibit Runx2-induced enhancement of ALP activity and mineralization in several murine osteogenic cell lines (45). It is therefore likely that the insulin- or IGF-I-stimulated PI3-K pathway is important for osteoblast differentiation regardless of the source of osteogenic cells, and that the levels present in FBS can activate the PI3-K pathway in SC cultures of human MSC.
Our results are consistent with the hypothesis that PI3-K activation is needed to regulate ALP and OP expression in human MSC at a site downstream or independent of BMP-mediated Smads. This would be in contrast with studies using the murine BMP-responsive osteogenic cell line 2T3, which suggest a positive role for PI3-K/AKT in Smad signaling, with dominant negative AKT abrogating Smad-dependent transcription of BMP-2 (43). It is not obvious why the PI3-K inhibitor increased Smad activation but did not increase Msx-2 mRNA, a gene directly regulated by BMP-activated Smads. It may be that the rate of Msx-2 transcription was already maximal without the inhibitor, or that the increased Smad activation was too transient to affect Msx2 levels 24 h later.
Our data indicating a negative role of ERK signaling and requirement of insulin/IGF-I signaling should prove useful in understanding the limitations in BMP-based therapies with human MSC (12, 46). They also suggest that human MSC may differ from several commonly studied osteogenic cell lines and rodent MSC culture models in their responses to BMPs and protein kinase signaling. It is also noteworthy that MSC from human, rat, and mouse, when cultured in standard SC, are dissimilar in their response to osteogenic inducers. In rat MSC, ALP is induced with either BMPs or Dex (14, 26); in mouse MSC, it is induced with BMPs, but not Dex (47, 48); and in most isolates of human MSC, ALP is induced with Dex, but not BMPs (19, 20). These observations together with our new findings on BMP regulation of early osteoblast genes in human MSC cultures imply that generalizations about mechanisms of osteoblast differentiation should be made with great caution.
Acknowledgments
We are grateful to Wyeth/Genetics Institute for providing rhBMP2, and Dr. Janet Rubin (Emory University, Atlanta, GA) for her gift of adenovirus containing caMEK. We also thank Andrew Chen, Lauran Madden, Geeta Bhargave, and Eleanor Golden for their excellent technical assistance, and Drs. David L. Diefenderfer and Jonathan Garino for providing human bone marrow cells.
References
Prockop DJ, Gregory CA, Spees JL 2003 One strategy for cell and gene therapy: harnessing the power of adult stem cells to repair tissues. Proc Natl Acad Sci USA 100:11917–11923
Phinney DG 2002 Building a consensus regarding the nature and origin of mesenchymal stem cells. J Cell Biochem 000:7–12
Fedde KN, Blair L, Silverstein J, Coburn SP, Ryan LM, Weinstein RS, Waymire K, Narisawa S, Millán JL, MacGregor GR, Whyte MP 1999 Alkaline phosphatase knock-out mice recapitulate the metabolic and skeletal defects of infantile hypophosphatasia. J Bone Miner Res 14:2015–2026
Anderson HC, Sipe JB, Hessle L, Dhamyamraju R, Atti E, Camacho NP, Millán JL 2004 Impaired calcification around matrix vesicles of growth plate and bone in alkaline phosphatase-deficient mice. Am J Pathol 164:841–847
Rickard DJ, Kassem M, Hefferan TE, Sarkar G, Spelsberg TC, Riggs BL 1996 Isolation and characterization of osteoblast precursor cells from human bone marrow. J Bone Miner Res 11:312–324
Thies RS, Bauduy M, Ashton BA, Kurtzberg L, Wozney JM, Rosen V 1992 Recombinant human bone morphogenetic protein-2 induces osteoblastic differentiation in W-20-17 stromal cells. Endocrinology 130:1318–1324
Ducy P 2000 Cbfa1: a molecular switch in osteoblast biology. Dev Dynamics 219:461–471
Gori F, Thomas T, Hicok KC, Spelsberg TC, Riggs BL 1999 Differentiation of human marrow stromal precursor cells: BMP-2 increases OSF2/CBFA1, enhances osteoblast commitment, and inhibits late adipocyte maturation. J Bone Miner Res 14:1522–1535
Valentin-Opran A, Wozney J, Csimma C, Lilly L, Riedel GE 2002 Clinical evaluation of recombinant human bone morphogenetic protein-2. Clin Orthopaed Related Res 000:110–120
Schmitt JM, Hwang K, Winn SR, Hollinger JO 1999 Bone morphogenetic proteins: an update on basic biology and clinical relevance. J Orthopaed Res 17:269–278
Li RH, Wozney JM 2001 Delivering on the promise of bone morphogenetic proteins. Trends Biotechnol 19:255–265
Govender S, Csimma C, Genant HK, Valentin-Opran A, BESTT SG 2002 Recombinant human bone morphogenetic protein-2 for treatment of open tibial fractures: a prospective, controlled, randomized study of four hundred and fifty patients. J Bone Joint Surg 84A:2123–2134
Groeneveld EH, Burger EH 2000 Bone morphogenetic proteins in human bone regeneration. Eur J Endocrinol 142:9–21
Rickard DJ, Sullivan TA, Shenker BJ, Leboy PS, Kazhdan I 1994 Induction of rapid osteoblast differentiation in rat bone marrow stromal cell cultures by dexamethasone and BMP-2. Dev Biol 161:218–228
Yamaguchi A, Ishizuya T, Kintou N, Wada Y, Katagiri T, Wozney JM, Rosen V, Yoshiki S 1996 Effects of BMP-2, BMP-4, and BMP-6 on osteoblastic differentiation of bone marrow-derived stromal cell lines, ST2 and MC3T3–G2/PA6. Biochem Biophys Res Commun 220:366–371
Abe E, Yamamoto M, Taguchi Y, Lecka-Czernik B, O’Brien CA, Economides AN, Stahl N, Jilka RL, Manolagas SC 2000 Essential requirement of BMPs-2/4 for both osteoblast and osteoclast formation in murine bone marrow cultures from adult mice: antagonism by noggin. J Bone Miner Res 15:663–673
Hanada K, Dennis JE, Caplan AI 1997 Stimulatory effects of basic fibroblast growth factor and bone morphogenetic protein-2 on osteogenic differentiation of rat bone marrow-derived mesenchymal stem cells. J Bone Miner Res 12:1606–1614
Lecanda F, Avioli LV, Cheng SL 1997 Regulation of bone matrix protein expression and induction of differentiation of human osteoblasts and human bone marrow stromal cells by bone morphogenetic protein-2. J Cell Biochem 67:386–398
Diefenderfer DL, Osyczka AM, Reilly GC, Leboy PS 2003 BMP responsiveness in human mesenchymal stem cells. Connect Tissue Res 44:305–311
Diefenderfer DL, Osyczka AM, Garino JP, Leboy PS 2003 Regulation of BMP-induced transcription in cultured human bone marrow stromal cells. J Bone Joint Surg 85A:19–29
Jorgensen NR, Henriksen Z, Sorensen OH, Civitelli R 2004 Dexamethasone, BMP-2, and 1,25-dihydroxyvitamin D enhance a more differentiated osteoblast phenotype: validation of an in vitro model for human bone marrow-derived primary osteoblasts. Steroids 69:219–226
Derynck R, Zhang YE 2003 Smad-dependent and Smad-independent pathways in TGF-? family signalling. Nature 425:577–584
Xiao GZ, Gopalakrishnan R, Jiang D, Reith E, Benson MD, Franceschi RT 2002 Bone morphogenetic proteins, extracellular matrix, and mitogen-activated protein kinase signaling pathways are required for osteoblast-specific gene expression and differentiation in MC3T3–E1 cells. J Bone Miner Res 17:101–110
Lou JR, Tu Y, Li S, Manske PR 2000 Involvement of ERK in BMP-2 induced osteoblastic differentiation of mesenchymal progenitor cell line C3H10T1/2. Biochem Biophys Res Commun 268:757–762
Rubin J, Murphy TC, Zhu L, Roy E, Nanes MS, Fan X 2003 Mechanical strain differentially regulates endothelial nitric-oxide synthase and receptor activator of nuclear B ligand expression via ERK1/2 MAPK. J Biol Chem 278:34018–34025
Osyczka AM, Diefenderfer DL, Bhargave G, Leboy PS 2004 Different effects of BMP-2 on marrow stromal cells from human and rat bone. Cells Tissues Organs 176:109–119
Thomas GP, Bourne A, Eisman JA, Gardiner EM 2000 Species-divergent regulation of human and mouse osteocalcin genes by calciotropic hormones. Exp Cell Res 258:395–402
Pereira RC, Delany AM, Canalis E 2002 Effects of cortisol and bone morphogenetic protein-2 on stromal cell differentiation: correlation with CCAAT-enhancer binding protein expression. Bone 30:685–691
Liu F, Malaval L, Aubin JE 2003 Global amplification polymerase chain reaction reveals novel transitional stages during osteoprogenitor differentiation. J Cell Sci 116:1787–1796
Suzawa M, Tamura Y, Fukumoto S, Miyazono K, Fujita T, Kato S, Takeuchi Y 2002 Stimulation of Smad1 transcriptional activity by Ras-extracellular signal-regulated kinase pathway: a possible mechanism for collagen-dependent osteoblastic differentiation. J Bone Miner Res 17:240–248
Higuchi C, Myoui A, Hashimoto N, Kuriyama K, Yoshioka K, Yoshikawa H, Itoh K 2002 Continuous inhibition of MAPK signaling promotes the early osteoblastic differentiation and mineralization of the extracellular matrix. J Bone Miner Res 17:1785–1794
Lai CF, Cheng SL 2002 Signal transductions induced by bone morphogenetic protein-2 and transforming growth factor-? in normal human osteoblastic cells. J Biol Chem 277:15514–15522
Rodríguez JP, Ríos S, Fernández M, Santiba?ez JF 2004 Differential activation of ERK1,2 MAP kinase signaling pathway in mesenchymal stem cell from control and osteoporotic postmenopausal women. J Cell Biochem 92:745–754
Jaiswal RK, Jaiswal N, Bruder SP, Mbalaviele G, Marshak DR, Pittenger MF 2000 Adult human mesenchymal stem cell differentiation to the osteogenic or adipogenic lineage is regulated by mitogen-activated protein kinase. J Biol Chem 275:9645–9652
You J, Reilly GC, Zhen XC, Yellowley CE, Chen Q, Donahue HJ, Jacobs CR 2001 Osteopontin gene regulation by oscillatory fluid flow via intracellular calcium mobilization and activation of mitogen-activated protein kinase in MC3T3–E1 osteoblasts. J Biol Chem 276:13365–13371
Kretzschmar M, Doody J, Massagué J 1997 Opposing BMP and EGF signalling pathways converge on the TGF-? family mediator Smad1. Nature 389:618–622
Aubin J, Davy A, Soriano P 2004 In vivo convergence of BMP and MAPK signaling pathways: impact of differential Smad1 phosphorylation on development and homeostasis. Genes Dev 18:1482–1494
Funaba M, Zimmerman CM, Mathews LS 2002 Modulation of Smad2-mediated signaling by extracellular signal-regulated kinase. J Biol Chem 277:41361–41368
Hayashida T, deCaestecker M, Schnaper HW 2003 Cross-talk between ERK MAP kinase and Smad-signaling pathways enhances TGF-?; dependent responses in human mesangial cells. FASEB J 03–0037fje
Hussein SM, Duff EK, Sirard C 2003 Smad4 and ?-catenin co-activators functionally interact with lymphoid-enhancing factor to regulate graded expression of Msx2. J Biol Chem 278:48805–48814
Brugger SM, Merrill AE, Torres-Vazquez J, Wu N, Ting MC, Cho JYM, Dobias SL, Yi SE, Lyons K, Bell JR, Arora K, Warrior R, Maxson R 2004 A phylogenetically conserved cis-regulatory module in the Msx2 promoter is sufficient for BMP-dependent transcription in murine and Drosophila embryos. Development 131:5153–5165
Gutierrez J, Parrizas M, Maestro MA, Navarro I, Plisetskaya EM 1995 Insulin and IGF-I binding and tyrosine kinase activity in fish heart. J Endocrinol 146:35–44
Ghosh-Choudhury N, Abboud SL, Nishimura R, Celeste A, Mahimainathan L, Choudhury GG 2002 Requirement of BMP-2-induced phosphatidylinositol 3-kinase and Akt serine/threonine kinase in osteoblast differentiation and Smad-dependent BMP-2 gene transcription. J Biol Chem 277:33361–33368
Shoba LNN, Lee JC 2003 Inhibition of phosphatidylinositol 3-kinase and p70S6 kinase blocks osteogenic protein-1 induction of alkaline phosphatase activity in fetal rat calvaria cells. J Cell Biochem 88:1247–1255
Fujita T, Azuma Y, Fukuyama R, Hattori Y, Yoshida C, Koida M, Ogita K, Komori T 2004 Runx2 induces osteoblast and chondrocyte differentiation and enhances their migration by coupling with PI3K-Akt signaling. J Cell Biol 166:85–95
Einhorn TA 2003 Clinical applications of recombinant human BMPs: early experience and future development. J Bone Joint Surg 85A:82–88
Chen TL 2004 Inhibition of growth and differentiation of osteoprogenitors in mouse bone marrow stromal cell cultures by increased donor age and glucocorticoid treatment. Bone 35:83–95
Bellows CG, Ciaccia A, Heersche JN 1998 Osteoprogenitor cells in cell populations derived from mouse and rat calvaria differ in their response to corticosterone, cortisol, and cortisone. Bone 23:119–125(Anna M. Osyczka and Phoeb)
Address all correspondence and requests for reprints to: Dr. Anna M. Osyczka, Department of Biochemistry, University of Pennsylvania School of Dental Medicine, 240 South 40th Street, Philadelphia, Pennsylvania 19104-6030. E-mail: annamo@biochem.dental.upenn.edu.
Abstract
Bone marrow stromal cells (MSC) are the major source of osteoblasts for bone remodeling and repair in postnatal animals. Rodent MSC cultured with bone morphogenetic proteins (BMPs) differentiate into osteoblasts, but most human MSC show a poor osteogenic response to BMPs. In this study we demonstrate that BMP-induced osteogenesis in poorly responsive human MSC requires modulation of ERK and phosphatidylinositol 3-kinase (PI3-K) pathways. Either treating human MSC cultures with the MAPK/ERK kinase inhibitor PD98059 or transferring them to serum-free medium with insulin or IGF-I permits BMP-dependent increases in the expression of the early osteoblast-associated genes, alkaline phosphatase and osteopontin. Increased expression of these genes in BMP-treated, serum-free cultures correlates with increased nuclear levels of activated Smads, whereas serum-free cultures of human MSC expressing constitutively active MAPK/ERK kinase show decreased expression of early osteoblast genes and decreased nuclear translocation of BMP-activated Smads. Inhibiting ERK activity in human MSC also elevates the expression of Msx2, a transcription factor that is directly regulated by Smad-binding elements in its promoter. Therefore, growth factor stimulation leading to high levels of ERK activity in human MSC results in suppressed BMP-induced transcription of several early osteoblast genes, probably because levels of BMP-activated nuclear Smads are decreased. In contrast, inhibiting the insulin/IGF-I-activated PI3-K/AKT pathway decreases BMP-induced alkaline phosphatase and osteopontin expression in serum-free cultures of human MSC, but increases BMP activation of Smads; thus, PI3-K signaling is required for BMP-induced expression of early osteoblast genes in human MSC either downstream or independent of the BMP-activated Smad signaling pathway.
Introduction
REPAIR OF ADULT bone fracture involves bone marrow-derived mesenchymal cells, which serve as a source of osteochondral progenitors that invade the fracture site, proliferate, and differentiate into cartilage and bone. These mesenchymal progenitors [also known as mesenchymal stem cells or marrow stromal cells (MSC)] can be isolated from bone marrow and cultured under conditions that promote their in vitro differentiation into specific mesenchymal phenotypes, such as chondrocytes or osteoblasts (1, 2). To achieve the osteoblastic phenotype, progenitor cells need to express several bone-type extracellular matrix proteins, such as collagen type I, osteopontin (OP), bone sialoprotein, and osteocalcin, and produce high levels of the ecto-enzyme tissue-nonspecific alkaline phosphatase (ALP). High-level expression of ALP is required for mineralization of skeletal tissues (3, 4) and is induced early during osteoblast differentiation (5, 6). In both animal and human MSC cultures, ALP and OP serve as useful markers of early osteogenesis, and the expression of these genes usually increases by the end of the first week of culture. Among many transcription factors expressed early in osteogenesis, Runx2 (cbfa1) is noteworthy, because it is required for bone formation and is an important early indicator of osteogenic capacity of cells (7, 8).
Bone morphogenetic proteins (BMPs) are osteoinductive growth factors that can induce bone formation both in vivo and in vitro (9, 10, 11). A number of clinical studies have assessed the efficacy of recombinant human BMPs in the healing of critical-size bone defects and the acceleration of bone fracture healing in humans (9, 12). Although the latter studies have been promising, relatively high doses of BMPs were required to induce adequate bone formation compared with animal models, and large variations in response among individual patients were observed (13). In rodent cell cultures, BMPs have shown a potent osteogenic activity (14, 15, 16, 17), but their ability to induce osteogenesis of human cells is less clear (8, 18). Consistent with these studies, we have recently reported major differences in the BMP response of rodent and human MSC cultures. Although addition of BMPs to rat MSC cultures resulted in a marked stimulation of ALP mRNA and enzyme activity, 90% of human MSC cultures stimulated with BMP did not show elevated ALP despite transcriptional regulation of several other BMP-responsive genes (19, 20). Jorgenson et al. (21) recently reported similar data indicating that BMP-treated human MSC do not show elevated ALP. We have now identified culture conditions in which BMP treatment of human MSC results in high ALP expression and profound stimulation of OP.
The classic BMP signaling pathway operates by activation of the Smad family of transcription factors, and there is evidence that it can also act through a Smad-independent p38 MAPK signaling pathway (22). There have also been reports suggesting that other kinase pathways, such as ERK, c-Jun N-terminal kinase, phosphatidylinositol 3-kinase (PI3-K), Wnt, and nuclear factor-B, may substitute for, activate, or modulate BMP signaling (23, 24). In this study we present evidence that in BMP-unresponsive human MSC cultures, growth factor signaling leading to high and sustained activity of ERK negatively regulates BMP-mediated Smad signaling. Furthermore, human MSC require PI3-K/AKT activation to achieve BMP-stimulated, high-level expression of several genes associated with the early stages of osteoblast differentiation.
Materials and Methods
Chemicals
Unless otherwise stated, all chemical reagents were purchased from Sigma-Aldrich Corp. (St. Louis, MA), and tissue culture solutions were obtained from Invitrogen Life Technologies, Inc. (Grand Island, NY). ITS+ Premix, composed of insulin, transferrin, selenious acid, linoleic acid and BSA, was obtained from BD Biosciences (Bedford, MA). Ascorbate-2-phosphate was obtained from Wako Chemicals (Richmond, VA), and the ERK inhibitor UO126 was purchased from BIOMOL Research Laboratories (Plymouth Meeting, PA). Proteinase inhibitor cocktails were purchased from Calbiochem (San Diego, CA). The PI3-K inhibitor Ly294002, primary rabbit polyclonal antibodies against p44/42 MAPK (ERK); phospho-p44/42 MAPK (p-ERK); phospho-Smad1, -5, and -8; AKT; and phospho-AKT were purchased from Cell Signaling Technology (Beverly, MA). Primary rabbit antihuman MADR1/Smad1 antibodies (recognizing both human Smad1 and -5) were obtained from Upstate Biotechnology, Inc. (Lake Placid, NY). Recombinant human bone morphogenetic protein type 2 (BMP-2) was provided by Wyeth/Genetics Institute (Cambridge, MA). Adenovirus containing constitutively active MAPK/ERK kinase 1 (caMEK) was a gift from Dr. Janet Rubin, Emory University (25).
Cell culture, inhibitor studies, and adenoviral infections
Human marrow samples were isolated from the proximal medullary cavities of femurs from patients undergoing total hip replacement. Primary cultures were established as described previously (19, 20) with a seeding density of 5 x 105 cells/cm2. Cells were grown in MEM supplemented with 10% pretested fetal bovine serum (FBS; Premium Select, Atlanta Biologicals, Norcross, GA) and antibiotics at 37 C in a humidified, 5% CO2 atmosphere. Medium was changed initially on d 4, then every other day thereafter. At confluence, cells were detached using 0.25% trypsin in 1 mM tetrasodium EDTA, centrifuged, resuspended in fresh MEM/10% FBS and plated at 104 cells/cm2.
All assays were carried out on first or second passage cell cultures. One day after plating the medium was exchanged for fresh medium supplemented with 0.35 mM L-ascorbic acid-2-phosphate (ascorbate), 10–7 M dexamethasone (Dex), or 100 ng/ml BMP-2 as noted. On d 4, cells were washed with Hanks’ buffered saline solution (HBSS), and medium was changed to one of the following: 1) fresh MEM containing 10% FBS; 2) MEM supplemented with ITS+ Premix, insulin at 1.1 μM, or IGF-I at 13.1 nM (SF+I); or 3) MEM supplemented only with 1.25 mg/ml BSA. Osteogenic supplements were then replaced. Medium with 10% FBS was estimated to contain 3–6 nM insulin and, based on values in the literature, an additional 2 nM IGF-I. ERK inhibitors PD98059 and UO126, PI3-K inhibitor Ly294002, and cycloheximide or respective vehicle solution were added on d 4. Inhibitors were added 1–2 h before the addition of ascorbate, BMP-2, or Dex. For infections with adeno-caMEK or empty adenovirus, cells were switched to SF+I on d 4, infected on d 5, and analyzed by real-time PCR on d 6. Cells were infected using a multiplicity of infection of 10. At this multiplicity of infection, the proportion of adenovirus-infected MSC is usually 40–70% without adenovirus-mediated cell toxicity.
ALP assay
Immediately before cell harvest for ALP assay, the numbers of viable cells were estimated using CellTiter 96 Aqueous One Solution Cell Proliferation Assay (Promega Corp., Madison, WI). Cells in 24-well culture plates were washed once with HBSS and covered with 200 μl of a solution prepared as a 1:10 (vol/vol) dilution of the reagent containing the MTS tetrazolium salt, [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt, and the electron-coupling reagent, phenazine ethosulfate, in phenol red-free MEM. Cells were incubated for 15 min at 37 C in a humidified, 5% CO2 atmosphere, MTS-containing medium from each well was transferred to 96-well plates, and the absorbance was measured at 490 nm. After washing twice with HBSS, cells were lysed in 10% Triton X-100, and stored at 4 C until ALP assay. ALP activity was determined kinetically as described previously (14).
Isolation of total RNA, cDNA synthesis, and real-time PCR analysis
Total RNA was extracted using Tri-Reagent (Molecular Research Center, Cincinnati, OH). Aliquots of total RNA (2 μg) were reverse transcribed, and cDNA samples corresponding to 50–250 ng input RNA were amplified as described previously (19). The reaction mixtures contained 2–3 μM forward and reverse primers and 3–4 mM MgCl2. Table 1 lists the primer sequences and annealing temperatures individually optimized for each primer pair. A final melt curve from 60–95 C was performed to confirm the specificity of the PCR, and the identities of PCR products were confirmed by gel electrophoresis. For calculating mRNA levels, results from real-time RT-PCR were converted to arbitrary units of mRNA, assuming a concentration-dependent straight line for a semilog plot, with a value of 3.5 for the fold change in mRNA/cycle (slope), and the crossing point cycle number with no cDNA template as an estimate of the y-intercept.
TABLE 1. Oligonucleotide primers for real-time RT-PCR analyses
Western blot analyses
Whole cell extracts were obtained by lysing cells in cell lysis buffer provided by Cell Signaling Technology. Nuclear extracts were obtained by scraping cells in Tris-buffered saline, centrifuging and resuspending pellets in ice-cold buffer composed of 10 mM HEPES (pH 8.0), 10 mM KCl, 1 mM EDTA (pH 8.0), 1 mM EGTA (pH 8.0), 1 mM dithiothreitol, 2 mM phenylmethylsulfonylfluoride, 0.5 mM sodium orthovanadate, 1 mM tosyl-L-phenylalanine, and 10 mM ethylmaleimide. After 15min incubation on ice, NP-40 was added and nuclei pelleted by centrifugation at 16,000 x g. The nuclear pellets were resuspended in cold buffer composed of 20 mM HEPES (pH 8.0), 1 mM EDTA, 1 mM EGTA, 0.4 M NaCl, 1 mM dithiothreitol, 2 mM phenylmethylsulfonylfluoride, 0.5 mM sodium orthovanadate, 0.1 mM N-tosyl-L-phenylaline chloromethyl ketone, and 10 mM N-ethylmaleimide. The protein concentration was determined with the MicroBCA protein assay reagent kit (Pierce Chemical Co., Rockford, IL). Forty to 80 mg protein from whole cell extract or 10–20 mg protein from nuclear extracts was separated on NuPAGE 4–12% bis-Tris gels (Invitrogen Life Technologies, Inc., Carlsbad, CA) under reducing conditions using 4-morpholinepropanesulfonic acid buffer and then transferred to immunoblot polyvinylidene difluoride membranes (Bio-Rad Laboratories, Hercules, CA). Membranes were incubated overnight with primary antibodies diluted 1:1,000 and then exposed for 2 h to horseradish peroxidase-linked antirabbit IgG (Amersham Biosciences, Piscataway, NJ) diluted 1:1250. The peroxidase-based signal was detected using Western Lightning Chemiluminescence Reagent Plus (PerkinElmer, Boston, MA) using a Kodak Image Station 440CF (Eastman Kodak, Rochester, NY).
Results
BMP induction of ALP activity in human MSC cultures requires serum withdrawal and addition of insulin or IGF-I
We have previously reported that the majority of human MSC isolates show a relatively poor osteogenic response to BMPs in vitro compared with similar rodent MSC cultures (20, 26). We therefore examined whether altering the culture conditions of the poorly responsive human MSC would elicit a more robust osteogenic response to BMP-2. The effects of BMP-2 were compared in the presence and absence of serum, with and without 1.1 μM insulin or 13.1 nM IGF-I. As shown in Fig. 1, BMP-2 increased ALP activity on d 7 approximately 4-fold when human MSC were cultured for the last 3 d in serum-free medium (SF) with IGF-I or insulin, but not when they were maintained in serum-containing medium (SC). Decreasing the serum concentration to 2% did not have any beneficial effect on BMP-2 stimulation of ALP activity (data not shown). Human MSC cultured for up to 14 d in SC failed to show BMP-induced ALP increase with BMP-2, BMP-4, or BMP-7 and with several sources of FBS (data not shown).
FIG. 1. Effects of IGF-I and insulin on BMP-2 stimulation of ALP activity in human MSC cultures grown in either SF or SC. Human MSC cultures were treated from d 1 with BMP-2 or were left untreated. On d 4, culture medium was changed to fresh SF or SC, with or without 13.1 nM IGF-I or 1.1 μM insulin, followed by appropriate treatments. ALP activity (mean ± SD) was determined on d 7. *, Significantly higher than without BMP, P < 0.01. #, Significantly lower than corresponding SF samples, P < 0.005.
Dose-dependency studies showed that at 6–11 nM, insulin and IGF-I had similar effects on human MSC in SF cultures, enhancing BMP-stimulated ALP activity approximately 2- to 4-fold (Table 2). Addition of either insulin or IGF-I slightly increased the number of cells in SF human MSC cultures, but adding BMP-2 to insulin or IGF-I-treated cultures decreased cell number 10–20% (Table 3). The net effect of insulin or IGF-I in combination with BMP-2 was to increase ALP levels without a significant increase in proliferation. Therefore, the effect of adding insulin or IGF-I to BMP-treated cultures was to increase ALP activity per cell, rather than produce greater numbers of ALP-positive cells. Similar data were obtained using SF supplement ITS+ Premix, which contains 1.1 μM insulin. Unless stated otherwise, additional studies were performed using SF in which insulin was supplied as ITS+ Premix.
TABLE 2. ALP activity of human MSC continuously cultured in serum or after transfer to SF with insulin or IGF-I
TABLE 3. Numbers of viable human MSC continuously cultured in serum or transferred for 3 d to SF with insulin or IGF-I
Switching human MSC cultures to SF+I promotes BMP-2 induction of ALP, OP, and Msx2 mRNAs
Comparisons of gene expression in 7-d cultures grown for the last 3 d in SF+I or continuously cultured in SC are presented in Fig. 2. Transfer to SF+I had a marked effect on BMP-2 stimulation of genes expressed early in osteoblast differentiation. ALP mRNA, which was unaffected by BMP treatment in the presence of serum, was increased by BMP-2 if cultures were grown for the last 3 d in SF+I. Previous studies (26) demonstrated that serum-containing cultures of human MSC showed BMP-induced increases in mRNA for the matrix protein osteopontin and the transcription factor Msx2. Studies with cyclohexamide treatment of human MSC cultures indicated that new protein synthesis was required for BMP-2 induction of ALP and osteopontin mRNAs, but not for Msx2 mRNA (data not shown). This is consistent with the assumption that BMP induction of Msx2 expression in human MSC is a direct result of Smad activation. In the present work, both osteopontin and Msx2 mRNA levels were modestly stimulated with BMP-2 in serum-containing MSC cultures, but BMP-2 had a far greater effect if cells were switched to SF+I. Analyses at increasing times after transfer to SF+I revealed that levels of ALP mRNA additionally increased with longer exposure to SF+I, and that at least 10 h of exposure to BMP-2 were required to obtain enhanced ALP mRNA expression in SF+I. This relatively long exposure time is consistent with evidence from cyclohexamide studies that BMP does not regulate ALP simply by binding of BMP-activated Smads to a regulatory region of the ALP gene.
FIG. 2. mRNA levels of osteoblast genes in human MSC cultured in SF+I or SC (SERUM). Cells were treated from d 1 of culture with BMP-2 or Dex or were left untreated (CTR). On d 4, culture medium was exchanged to SF+I or SERUM, followed by appropriate treatment. RNA was extracted from cells on d 5 for Msx2 or d 7 for other genes and analyzed by real-time RT-PCR. Values are the mean ± SD, expressed in mRNA arbitrary units. *, P < 0.03 compared with corresponding untreated; **, P < 0.001 compared with corresponding untreated; #, P < 0.03 compared with SF+I.
Analyses of mRNA levels of other early osteoblast genes revealed that changing from SC to SF conditions did not affect collagen type I mRNA expression (data not shown), whereas levels of Runx-2 mRNA were increased 2- to 3-fold by BMP-2 in both SC and SF+I cultures. Analyses of late differentiation markers indicated that bone sialoprotein mRNA was increased by BMP-2 in SC cultures, but transfer to SF+I medium significantly decreased BSP mRNA expression with and without inducers. BMP-2 treatment did not alter osteocalcin mRNA levels regardless of whether serum was present; however, it should be noted that vitamin D3 was not added (27). Dex, which increased ALP expression of human MSC cultured in serum, did not increase ALP mRNA in SF+I conditions.
High ERK activity negatively regulates BMP-2 stimulation of ALP, OP, and Msx2 expression
Because SF lacks growth factors that activate the ERK pathway, we examined whether sustained activation of ERK would prevent BMP stimulation of gene expression in SF+I cultures of human MSC, using a caMEK adenoviral construct that phosphorylates ERK. As shown in Fig. 3, A–C, expressing caMEK in SF+I cultures markedly decreased ALP, OP, and Msx2 mRNA levels. Conversely, inhibiting ERK signaling with 50 μM PD98059 significantly increased ALP, OP, and Msx2 mRNA expression in BMP-stimulated SC cultures (Fig. 3, D–F). Although addition of insulin or IGF-I to SC human MSC cultures did not permit BMP stimulation of ALP (Fig 1A), adding either insulin or IGF-I in the presence of 50 μM PD98059 slightly enhanced BMP induction of ALP activity in these cultures (data not shown).
FIG. 3. mRNA levels of early osteoblast genes under conditions modulating ERK activation. A–C, Human MSC were transferred to SF+I medium on d 4 and infected with either empty adenovirus or adenovirus containing caMEK1 on d 5, and RNA was harvested on d 6. Values from real-time RT-PCR are expressed as the ratio of experimental mRNA to that for glyceraldehyde-3-P dehydrogenase for the same sample. This experiment was repeated twice with two different cell isolates, with similar results. D–F, Human MSC were cultured in 10% FBS (SERUM) and treated with 50 μM PD98059 on d 4, and RNA was extracted on d 6. Values, expressed in arbitrary units of mRNA, are the mean ± SD for RNA derived from at least three different MSC isolates. *, Significantly higher than without BMP, P < 0.01; #, significantly higher than without PD98059 inhibitor, P = 0.001.
Western blot assays for phosphorylated ERK, using whole cell extracts harvested at 5, 15, 45, and 90 min, showed that ERK activation peaked 15 min after medium change in both SC and SF culture conditions. At this time point, ERK activation was the highest in human MSC cultured in SC without ERK inhibitor (Fig. 4), and BMP had no significant effect on ERK activation in these cultures. In SF cultures, addition of either insulin or BMP-2 caused a slight increase in ERK activation at 15 min, but the levels of phosphorylated ERK in these culture conditions were markedly lower than those in SC cultures. Similar results were obtained using an in vitro kinase assay to measure 32P incorporation into the ERK substrate Elk-1 (data not shown).
FIG. 4. ERK activation in human MSC cultures. Cells were treated from d 1 of culture with BMP-2 or were left untreated (Control). On d 4, medium was replaced with SC or SC supplemented with 50 μM PD98059 (SC+PD) or was changed to SF+I) or SF without insulin (SF–), followed by appropriate treatments. Whole cell extracts were prepared 15 min after medium changes and analyzed by Western blot. A, Western blot analysis of phosphorylated p44/42 (p-ERK) and total ERK. B, Densitometry quantification of activated ERK, expressed as a ratio to total ERK.
Human MSC with low ERK activity have high levels of phospho-Smad1, -5, and -8 and greater Smad nuclear localization
To obtain additional insights into whether SF culture conditions promote BMP-mediated Smad signaling, Smad1, -5, and -8 activation was analyzed in whole cell and nuclear extracts using a polyclonal antibody that recognizes Smad1, -5, and -8 phosphorylated at the C-terminal BMP activation domain. Western blot analysis of whole cell extracts showed that BMP-2 increased phosphorylation of Smad1, -5, and -8 in both SC and SF culture conditions, but the activation of BMP-regulated Smads was markedly greater in SF cultures (Fig. 5, A and B). Switching cells to SF conditions also increased levels of activated Smad1, -5, and -8 in the nucleus (Fig. 5, C and D), but adding insulin had no additional effect on the amounts of activated Smads. These results suggested that Smad activation was promoted by serum removal, and the insulin/IGF-I requirement was unrelated to Smad activation. Increasing ERK signaling in SF+I cultures by overexpressing caMEK did not affect the whole cell levels of activated Smads (Fig. 5E), but decreased the levels of activated Smads in the nucleus (Fig. 5F).
FIG. 5. Levels of BMP-activated Smads (P-Smads) in human MSC cultured in SC or SF. Cells were treated from d 1 of culture with BMP-2 or were left untreated (Control). On d 4, medium was changed to SC, SF+I, or SF without insulin (SF–), followed by appropriate treatments. Proteins were extracted 1 or 3 h later and analyzed by Western blot. A, Western blot analysis of activated and total Smads in whole cell extracts 1 h after medium changes. B, Densitometric quantification of activated Smads in whole cell extracts, expressed as a ratio to total Smads. C, Western blot analysis of activated and total Smads in nuclear extracts 3 h after medium changes. D, Densitometric quantification of activated Smads in nuclear extracts, expressed as a ratio to total Smads. E and F, Cells were infected with either empty adenovirus or caMEK1 (CA-MEK) on d 4, and 6 h later were transferred to SF+I medium with or without BMP-2. E, Western blot analysis of activated and total Smads in whole cell extracts obtained at the indicated time intervals after medium changes. F, Western blot analysis of activated and total Smads in nuclear extracts prepared 3 h after medium change.
PI3-K activation is necessary for the expression of early osteoblast genes, but not Msx2
The finding that insulin or IGF-I was necessary for BMP induction of ALP in SF cultures suggested that the PI3-K/AKT pathway might be implicated in BMP-stimulated osteogenesis. The PI3-K inhibitor Ly294002, at 5–10 μM, decreased both BMP-stimulated ALP activity and ALP mRNA expression in SF+I cultures (Fig. 6, A and B). A similar inhibitory effect of Ly294002 was seen for BMP-stimulated OP mRNA expression (Fig. 6C), but not for Msx2 mRNA expression (Fig. 6D). The efficacy of the Ly294002 inhibitor was confirmed by Western blot analysis of AKT phosphorylation, which is mediated by PI3-K (Fig. 7, A and C). To determine whether PI3-K inhibitor affects BMP-mediated Smad signaling in SF+I cultures, activation of Smad1, -5, and -8 was analyzed in whole cell extracts from SF+I cultures treated with Ly294002. As shown in Fig. 7, C and D, activation of Smad1, -5, and -8 increased with BMP-2 and was further elevated when Ly29004 inhibitor was added.
FIG. 6. Effect of PI3-K inhibitor on BMP induction of early osteoblast genes. On d 4 of culture, human MSC were transferred to SF+I with and without 10 μM Ly294002, followed by appropriate treatment. RNA was prepared on d 5 or 6 and analyzed by real-time RT-PCR. Values (mean ± SD for three different MSC isolates) are expressed in arbitrary units. *, Significantly higher than without BMP, P < 0.01; **, significantly higher than without BMP, P = 0.02; #, significantly lower than corresponding sample without Ly294002 inhibitor, P < 0.01.
FIG. 7. Effect of PI3-K inhibitor on BMP activation of Smads. On d 4 of culture, human MSC were transferred to SF+I with and without 10 μM Ly294002, followed by appropriate treatment. Whole cell extracts were prepared at the indicated time intervals after medium change and analyzed by Western blot. A, Western blot analysis of activated (P-AKT) and total AKT. B, Western blot analysis of activated (P-Smads) and total Smads. C, Densitometric quantification of activated AKT, expressed as a ratio to total AKT. D, Densitometric quantification of activated Smads, expressed as a ratio to total Smads.
Discussion
We have previously reported differences in the BMP-mediated osteogenic response of human and rodent bone marrow-derived mesenchymal stem cells (19, 20, 26). Among 26 bone marrow cell samples obtained from the femoral cavity of adult human patients, 90% of isolates failed to increase ALP expression in response to 100 ng/ml BMP-2. In contrast, rat and mouse MSC cultures significantly increase expression of ALP and other osteoblast-related genes with as little as 30 ng/ml BMP-2 (5, 6, 17, 26, 28). In this study we show that these poorly responsive human MSC markedly elevate both ALP and OP expression when cultures are stimulated with BMP-2 in the presence of ERK inhibitor or in SF containing either insulin or IGF-I. Our examination of signaling pathways implicated in BMP regulation of human MSC osteogenesis indicates not only a negative role of ERK, but also a positive role of PI3-K/AKT signaling in the regulation of ALP and OP. Analyses of other osteoblast-related markers suggest that these kinase pathways may modulate the transcription of some genes expressed early in osteoblast differentiation without affecting other osteoblast phenotypic markers. This is consistent with the assumption that osteoblast differentiation is not regulated by a single set of factors coordinately controlling all osteoblast genes and that it is not necessary for progenitor cells to express all genes seen early in osteoblast differentiation to up-regulate some late osteoblast markers (29).
The role of ERK activation in osteoblast differentiation has been the subject of many studies, but no clear picture has emerged. In BMP-treated human and mouse osteoblastic cells, ERK activation is required for increased expression of the late osteoblast gene, osteocalcin (23, 30); however, activated ERK decreases ALP expression (31, 32, 33). This is consistent with our results indicating that high levels of ERK activation have a negative effect on BMP induction of early osteoblast genes in human MSC. In C3H10T1/2 cells, an embryonic murine cell line frequently considered equivalent to MSC, BMP stimulates both ERK expression and ERK activity, and reducing ERK activation decreased BMP-induced ALP activity (24). In contrast, our studies with human MSC showed no significant BMP-2 effect on total ERK levels either in the presence of serum or after transfer of cultures to SF+I. In SF cultures, BMP-2, insulin, or both slightly increased phosphorylation of ERK, but the major factor increasing ERK phosphorylation was the presence of serum. These results combined with the observed effects of caMEK in SF cultures and ERK inhibitors in the presence of serum argue that in human MSC lower levels of activated ERK are associated with improved BMP-2 stimulation of ALP and OP expression.
A positive effect of ERK activation on osteoblast differentiation has been observed with studies using other osteogenic stimuli. In dexamethasone-treated human MSC cultures, there is sustained ERK activation throughout osteogenic differentiation (34). Oscillatory fluid flow increases OP mRNA in MC3T3-E1 cells, and this stimulation is abolished with ERK inhibitors, suggesting that ERK activation plays a positive role in osteopontin induction (35). These reports suggest that the negative effect ERK activation has on the expression of early osteoblast genes in human MSC may be limited to osteogenesis induced by BMP. In epithelial cells overexpressing Smad1, ERK directly regulates BMP-activated Smads by phosphorylation at serine residues within the Smad linker region; this phosphorylation inhibited both nuclear accumulation of Smad1 and its transcriptional activity (36). Recent evidence indicates that linker phosphorylation of Smad1 is important in vivo (37). Several other Smads can be phosphorylated by ERK in the linker region, and whether it results in positive or negative consequences for Smad activity appears to be cell specific (38, 39). Using a p-Smad1, -5, and -8 antibody that detects only BMP-regulated phosphorylation, we have shown that increased ERK activation resulted in decreased nuclear levels of BMP-activated Smads. It is therefore plausible that in human MSC, ERK phosphorylation of BMP-related Smads in the linker region limits the ability of BMP-activated Smads to function as transcription factors. This hypothesis is supported by the increased Msx2 mRNA levels seen with BMP-treated human MSC cultured either in SF conditions or with ERK inhibitors in serum. BMP-2 has been shown to directly regulate Msx2 expression by Smad binding to its promoter (40, 41), and we have found that BMP stimulation of Msx2 in human MSC does not require new protein synthesis. The fact that BMPs directly regulate the Msx2 promoter by activated Smads together with our observation that BMP-stimulated Msx2 expression is modulated by ERK activation imply that levels of activated ERK are modulating BMP-activated Smad function in human MSC.
In contrast to ERK, activation of the PI3-K/AKT pathway by insulin or IGF-I seems to positively regulate BMP-2-induced ALP and OP in human MSC cultures. Both insulin and IGF-I are highly specific for their respective receptors, and a 100-fold excess of the incorrect ligand is required to displace the correct one (42). The fact that either insulin or IGF-I, at roughly equimolar concentrations, is capable of permitting a BMP response implies that activation of PI3-K signaling by either ligand is sufficient. This assumption is reinforced by the finding that blocking PI3-K signaling abolishes the BMP osteogenic response. These results support and extend previous data showing that PI3-K and AKT serine/threonine kinase are required for BMP-induced osteogenesis in 2T3 cells (43) and fetal rat calvaria cells (44). Blocking PI3-K/AKT signaling has been reported to also inhibit Runx2-induced enhancement of ALP activity and mineralization in several murine osteogenic cell lines (45). It is therefore likely that the insulin- or IGF-I-stimulated PI3-K pathway is important for osteoblast differentiation regardless of the source of osteogenic cells, and that the levels present in FBS can activate the PI3-K pathway in SC cultures of human MSC.
Our results are consistent with the hypothesis that PI3-K activation is needed to regulate ALP and OP expression in human MSC at a site downstream or independent of BMP-mediated Smads. This would be in contrast with studies using the murine BMP-responsive osteogenic cell line 2T3, which suggest a positive role for PI3-K/AKT in Smad signaling, with dominant negative AKT abrogating Smad-dependent transcription of BMP-2 (43). It is not obvious why the PI3-K inhibitor increased Smad activation but did not increase Msx-2 mRNA, a gene directly regulated by BMP-activated Smads. It may be that the rate of Msx-2 transcription was already maximal without the inhibitor, or that the increased Smad activation was too transient to affect Msx2 levels 24 h later.
Our data indicating a negative role of ERK signaling and requirement of insulin/IGF-I signaling should prove useful in understanding the limitations in BMP-based therapies with human MSC (12, 46). They also suggest that human MSC may differ from several commonly studied osteogenic cell lines and rodent MSC culture models in their responses to BMPs and protein kinase signaling. It is also noteworthy that MSC from human, rat, and mouse, when cultured in standard SC, are dissimilar in their response to osteogenic inducers. In rat MSC, ALP is induced with either BMPs or Dex (14, 26); in mouse MSC, it is induced with BMPs, but not Dex (47, 48); and in most isolates of human MSC, ALP is induced with Dex, but not BMPs (19, 20). These observations together with our new findings on BMP regulation of early osteoblast genes in human MSC cultures imply that generalizations about mechanisms of osteoblast differentiation should be made with great caution.
Acknowledgments
We are grateful to Wyeth/Genetics Institute for providing rhBMP2, and Dr. Janet Rubin (Emory University, Atlanta, GA) for her gift of adenovirus containing caMEK. We also thank Andrew Chen, Lauran Madden, Geeta Bhargave, and Eleanor Golden for their excellent technical assistance, and Drs. David L. Diefenderfer and Jonathan Garino for providing human bone marrow cells.
References
Prockop DJ, Gregory CA, Spees JL 2003 One strategy for cell and gene therapy: harnessing the power of adult stem cells to repair tissues. Proc Natl Acad Sci USA 100:11917–11923
Phinney DG 2002 Building a consensus regarding the nature and origin of mesenchymal stem cells. J Cell Biochem 000:7–12
Fedde KN, Blair L, Silverstein J, Coburn SP, Ryan LM, Weinstein RS, Waymire K, Narisawa S, Millán JL, MacGregor GR, Whyte MP 1999 Alkaline phosphatase knock-out mice recapitulate the metabolic and skeletal defects of infantile hypophosphatasia. J Bone Miner Res 14:2015–2026
Anderson HC, Sipe JB, Hessle L, Dhamyamraju R, Atti E, Camacho NP, Millán JL 2004 Impaired calcification around matrix vesicles of growth plate and bone in alkaline phosphatase-deficient mice. Am J Pathol 164:841–847
Rickard DJ, Kassem M, Hefferan TE, Sarkar G, Spelsberg TC, Riggs BL 1996 Isolation and characterization of osteoblast precursor cells from human bone marrow. J Bone Miner Res 11:312–324
Thies RS, Bauduy M, Ashton BA, Kurtzberg L, Wozney JM, Rosen V 1992 Recombinant human bone morphogenetic protein-2 induces osteoblastic differentiation in W-20-17 stromal cells. Endocrinology 130:1318–1324
Ducy P 2000 Cbfa1: a molecular switch in osteoblast biology. Dev Dynamics 219:461–471
Gori F, Thomas T, Hicok KC, Spelsberg TC, Riggs BL 1999 Differentiation of human marrow stromal precursor cells: BMP-2 increases OSF2/CBFA1, enhances osteoblast commitment, and inhibits late adipocyte maturation. J Bone Miner Res 14:1522–1535
Valentin-Opran A, Wozney J, Csimma C, Lilly L, Riedel GE 2002 Clinical evaluation of recombinant human bone morphogenetic protein-2. Clin Orthopaed Related Res 000:110–120
Schmitt JM, Hwang K, Winn SR, Hollinger JO 1999 Bone morphogenetic proteins: an update on basic biology and clinical relevance. J Orthopaed Res 17:269–278
Li RH, Wozney JM 2001 Delivering on the promise of bone morphogenetic proteins. Trends Biotechnol 19:255–265
Govender S, Csimma C, Genant HK, Valentin-Opran A, BESTT SG 2002 Recombinant human bone morphogenetic protein-2 for treatment of open tibial fractures: a prospective, controlled, randomized study of four hundred and fifty patients. J Bone Joint Surg 84A:2123–2134
Groeneveld EH, Burger EH 2000 Bone morphogenetic proteins in human bone regeneration. Eur J Endocrinol 142:9–21
Rickard DJ, Sullivan TA, Shenker BJ, Leboy PS, Kazhdan I 1994 Induction of rapid osteoblast differentiation in rat bone marrow stromal cell cultures by dexamethasone and BMP-2. Dev Biol 161:218–228
Yamaguchi A, Ishizuya T, Kintou N, Wada Y, Katagiri T, Wozney JM, Rosen V, Yoshiki S 1996 Effects of BMP-2, BMP-4, and BMP-6 on osteoblastic differentiation of bone marrow-derived stromal cell lines, ST2 and MC3T3–G2/PA6. Biochem Biophys Res Commun 220:366–371
Abe E, Yamamoto M, Taguchi Y, Lecka-Czernik B, O’Brien CA, Economides AN, Stahl N, Jilka RL, Manolagas SC 2000 Essential requirement of BMPs-2/4 for both osteoblast and osteoclast formation in murine bone marrow cultures from adult mice: antagonism by noggin. J Bone Miner Res 15:663–673
Hanada K, Dennis JE, Caplan AI 1997 Stimulatory effects of basic fibroblast growth factor and bone morphogenetic protein-2 on osteogenic differentiation of rat bone marrow-derived mesenchymal stem cells. J Bone Miner Res 12:1606–1614
Lecanda F, Avioli LV, Cheng SL 1997 Regulation of bone matrix protein expression and induction of differentiation of human osteoblasts and human bone marrow stromal cells by bone morphogenetic protein-2. J Cell Biochem 67:386–398
Diefenderfer DL, Osyczka AM, Reilly GC, Leboy PS 2003 BMP responsiveness in human mesenchymal stem cells. Connect Tissue Res 44:305–311
Diefenderfer DL, Osyczka AM, Garino JP, Leboy PS 2003 Regulation of BMP-induced transcription in cultured human bone marrow stromal cells. J Bone Joint Surg 85A:19–29
Jorgensen NR, Henriksen Z, Sorensen OH, Civitelli R 2004 Dexamethasone, BMP-2, and 1,25-dihydroxyvitamin D enhance a more differentiated osteoblast phenotype: validation of an in vitro model for human bone marrow-derived primary osteoblasts. Steroids 69:219–226
Derynck R, Zhang YE 2003 Smad-dependent and Smad-independent pathways in TGF-? family signalling. Nature 425:577–584
Xiao GZ, Gopalakrishnan R, Jiang D, Reith E, Benson MD, Franceschi RT 2002 Bone morphogenetic proteins, extracellular matrix, and mitogen-activated protein kinase signaling pathways are required for osteoblast-specific gene expression and differentiation in MC3T3–E1 cells. J Bone Miner Res 17:101–110
Lou JR, Tu Y, Li S, Manske PR 2000 Involvement of ERK in BMP-2 induced osteoblastic differentiation of mesenchymal progenitor cell line C3H10T1/2. Biochem Biophys Res Commun 268:757–762
Rubin J, Murphy TC, Zhu L, Roy E, Nanes MS, Fan X 2003 Mechanical strain differentially regulates endothelial nitric-oxide synthase and receptor activator of nuclear B ligand expression via ERK1/2 MAPK. J Biol Chem 278:34018–34025
Osyczka AM, Diefenderfer DL, Bhargave G, Leboy PS 2004 Different effects of BMP-2 on marrow stromal cells from human and rat bone. Cells Tissues Organs 176:109–119
Thomas GP, Bourne A, Eisman JA, Gardiner EM 2000 Species-divergent regulation of human and mouse osteocalcin genes by calciotropic hormones. Exp Cell Res 258:395–402
Pereira RC, Delany AM, Canalis E 2002 Effects of cortisol and bone morphogenetic protein-2 on stromal cell differentiation: correlation with CCAAT-enhancer binding protein expression. Bone 30:685–691
Liu F, Malaval L, Aubin JE 2003 Global amplification polymerase chain reaction reveals novel transitional stages during osteoprogenitor differentiation. J Cell Sci 116:1787–1796
Suzawa M, Tamura Y, Fukumoto S, Miyazono K, Fujita T, Kato S, Takeuchi Y 2002 Stimulation of Smad1 transcriptional activity by Ras-extracellular signal-regulated kinase pathway: a possible mechanism for collagen-dependent osteoblastic differentiation. J Bone Miner Res 17:240–248
Higuchi C, Myoui A, Hashimoto N, Kuriyama K, Yoshioka K, Yoshikawa H, Itoh K 2002 Continuous inhibition of MAPK signaling promotes the early osteoblastic differentiation and mineralization of the extracellular matrix. J Bone Miner Res 17:1785–1794
Lai CF, Cheng SL 2002 Signal transductions induced by bone morphogenetic protein-2 and transforming growth factor-? in normal human osteoblastic cells. J Biol Chem 277:15514–15522
Rodríguez JP, Ríos S, Fernández M, Santiba?ez JF 2004 Differential activation of ERK1,2 MAP kinase signaling pathway in mesenchymal stem cell from control and osteoporotic postmenopausal women. J Cell Biochem 92:745–754
Jaiswal RK, Jaiswal N, Bruder SP, Mbalaviele G, Marshak DR, Pittenger MF 2000 Adult human mesenchymal stem cell differentiation to the osteogenic or adipogenic lineage is regulated by mitogen-activated protein kinase. J Biol Chem 275:9645–9652
You J, Reilly GC, Zhen XC, Yellowley CE, Chen Q, Donahue HJ, Jacobs CR 2001 Osteopontin gene regulation by oscillatory fluid flow via intracellular calcium mobilization and activation of mitogen-activated protein kinase in MC3T3–E1 osteoblasts. J Biol Chem 276:13365–13371
Kretzschmar M, Doody J, Massagué J 1997 Opposing BMP and EGF signalling pathways converge on the TGF-? family mediator Smad1. Nature 389:618–622
Aubin J, Davy A, Soriano P 2004 In vivo convergence of BMP and MAPK signaling pathways: impact of differential Smad1 phosphorylation on development and homeostasis. Genes Dev 18:1482–1494
Funaba M, Zimmerman CM, Mathews LS 2002 Modulation of Smad2-mediated signaling by extracellular signal-regulated kinase. J Biol Chem 277:41361–41368
Hayashida T, deCaestecker M, Schnaper HW 2003 Cross-talk between ERK MAP kinase and Smad-signaling pathways enhances TGF-?; dependent responses in human mesangial cells. FASEB J 03–0037fje
Hussein SM, Duff EK, Sirard C 2003 Smad4 and ?-catenin co-activators functionally interact with lymphoid-enhancing factor to regulate graded expression of Msx2. J Biol Chem 278:48805–48814
Brugger SM, Merrill AE, Torres-Vazquez J, Wu N, Ting MC, Cho JYM, Dobias SL, Yi SE, Lyons K, Bell JR, Arora K, Warrior R, Maxson R 2004 A phylogenetically conserved cis-regulatory module in the Msx2 promoter is sufficient for BMP-dependent transcription in murine and Drosophila embryos. Development 131:5153–5165
Gutierrez J, Parrizas M, Maestro MA, Navarro I, Plisetskaya EM 1995 Insulin and IGF-I binding and tyrosine kinase activity in fish heart. J Endocrinol 146:35–44
Ghosh-Choudhury N, Abboud SL, Nishimura R, Celeste A, Mahimainathan L, Choudhury GG 2002 Requirement of BMP-2-induced phosphatidylinositol 3-kinase and Akt serine/threonine kinase in osteoblast differentiation and Smad-dependent BMP-2 gene transcription. J Biol Chem 277:33361–33368
Shoba LNN, Lee JC 2003 Inhibition of phosphatidylinositol 3-kinase and p70S6 kinase blocks osteogenic protein-1 induction of alkaline phosphatase activity in fetal rat calvaria cells. J Cell Biochem 88:1247–1255
Fujita T, Azuma Y, Fukuyama R, Hattori Y, Yoshida C, Koida M, Ogita K, Komori T 2004 Runx2 induces osteoblast and chondrocyte differentiation and enhances their migration by coupling with PI3K-Akt signaling. J Cell Biol 166:85–95
Einhorn TA 2003 Clinical applications of recombinant human BMPs: early experience and future development. J Bone Joint Surg 85A:82–88
Chen TL 2004 Inhibition of growth and differentiation of osteoprogenitors in mouse bone marrow stromal cell cultures by increased donor age and glucocorticoid treatment. Bone 35:83–95
Bellows CG, Ciaccia A, Heersche JN 1998 Osteoprogenitor cells in cell populations derived from mouse and rat calvaria differ in their response to corticosterone, cortisol, and cortisone. Bone 23:119–125(Anna M. Osyczka and Phoeb)