| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,
,,
,1
* Japan Society for the Promotion of Science (JSPS), Chiyoda-ku, Tokyo, Japan;
Division of Sensory and Motor System Medicine and
Center for Disease Biology and Integrative Medicine, Faculty of Medicine, the University of Tokyo, Hongo, Bunkyo-ku, Tokyo, Japan; and
Department of Genetics and Developmental Biology, University of Connecticut Health Center, Farmington, Connecticut, USA
1Correspondence: Center for Disease Biology and Integrative Medicine, Faculty of Medicine, the University of Tokyo, 7–3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan. E-mail: uichung-tky{at}umin.ac.jp
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(1) promoter. Using these cells engineered to fluoresce on osteogenic differentiation, we screened cDNA libraries and combinations of major osteogenesis-related genes. Among them, the combination of constitutively active activin receptor-like kinase 6 (caALK6) and runt-related transcription factor 2 (Runx2) was the minimal unit that induced fluorescence. The combination efficiently induced osteogenic differentiation in various cell types, including terminally differentiated nonosteogenic cells. The cooperative action of the combination occurred through protein stabilization of core binding factor beta (Cbfb), induction of Runx2-Cbfb complex formation, and its DNA binding. Furthermore, transplantation of a monolayer sheet of fibroblasts transduced with the combination achieved bone regeneration within 4 wk in mouse calvarial bone defects. Thus, we successfully identified the potent combination of genes for bone regeneration, which helped broaden cell sources.—Ohba, S., Ikeda, T., Kugimiya, F., Yano, F., Lichtler, A. C., Nakamura, K., Takato, T., Kawaguchi, H., Chung, U. I. Identification of a potent combination of osteogenic genes for bone regeneration using embryonic stem (ES) cell-based sensor.
Key Words: osteogenesis • screening • biosensor • cell sheet
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Bone grafts and prosthetic implant devices are the current strategies to repair irreversible skeletal damages. However, the bone grafts have shortcomings concerning both quantity (availability of bone graft material) and quality (donor site troubles, graft rejection, and disease transmission), and the prosthetic implants have shortcomings concerning quality (biocompatibility, function, and longevity). Regenerative medicine using the technique of tissue engineering attempts to provide solutions to such problems (2)
.
There are three components important for tissue regeneration: scaffolds, cells, and signaling pathways. Among them, autologous cell transplantation of mesenchymal stromal cells (MSCs) derived from bone marrow and adipose tissue using biodegradable scaffolds have been widely used in bone regeneration (3
4
5
6)
. However, MSCs are limited both in quantity and differentiation capacity. Their definition is vague; it is controversial whether they are real stem cells that require the ability for self-renewal and multipotency. From 10 ml of bone marrow fluid or adipose tissue, only 103 to 106 cells can be isolated (3
, 7)
. To regenerate clinical bone defects, 
109 cells may be required (3)
, but it is difficult to expand MSCs by several rounds of passages without affecting their differentiation capacity (8)
. On the other hand, ES cells and multipotent adult progenitor cells are virtually unlimited in quantity; however, it is difficult and costly to isolate these cells, and their differentiation efficiency seems restricted (9
10
11)
. In short, the use of stem cells has not yet overcome these crucial hurdles and a new strategy needs to be developed. One solution for this conundrum is to establish a method to control osteoblast differentiation independent of cell sources and to apply this method to abundant autologous adult cells such as dermal fibroblasts for in vivo bone regeneration. For this purpose, we need to identify potent signals that can induce osteogenic differentiation even in nonosteogenic, nonstem cells.
Substantial progress has been made in the basic understanding of major osteogenic signaling molecules and genes such as bone morphogenetic proteins (BMPs), Hedgehogs, Runx2, Wnts, and insulin-like growth factors (IGFs) (12
13
14
15
16)
. Each factor, however, was effective on specific cell types, including stem cells and osteoblast progenitors (12
, 17)
. In addition, although most of these individual molecules are endogenously expressed in various tissues, the region where osteogenesis occurs is restricted. These data suggest that these individual factors are not potent enough and that ideal signaling may be achieved by a new factor or combinations of factors.
In this study, we developed an ES cell-based monitoring system for osteogenic differentiation that enabled us to identify a potent combination of osteogenic genes through a convenient, reliable screening method. We then investigated the molecular mechanisms underlying the cooperative action of components of the combination. Finally, we tested its in vivo relevance through transplantation of skin fibroblasts transduced with it into the mouse bone defect model.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Preparation of adenoviruses and plasmids
Adenoviral expression vectors encoding rat caSmoothened (caSmo) (18)
, myc-tagged human GLI3C’
ClaI (19)
, hemagglutinin (HA) -tagged mouse ca lymphoid enhancer factor 1 (LEF-1), HA-tagged mouse dominant negative (dn) LEF-1 (20)
, and mouse core binding factor beta (Cbfb; a generous gift from T. Komori, Nagasaki University, Nagasaki, Japan) were constructed using the AdenoX Expression System (Clontech, Palo Alto, CA, USA), according to the manufacturer’s instructions. Adenoviruses expressing mouse Runx2 and Flag-tagged mouse dnRunx2 were generous gifts from R. Nishimura (Osaka University, Osaka, Japan); adenoviruses expressing HA-tagged mouse caALK6 and Flag-tagged mouse Smad6 from K. Miyazono (the University of Tokyo, Tokyo, Japan); adenoviruses expressing HA-tagged human insulin receptor substrate 1 (IRS-1) and HA-tagged human dnIRS-1 from W. Ogawa (Kobe University, Kobe, Japan).
Cell culture
NIH3T3 and HeLa were obtained from the Riken Cell Bank (Tsukuba, Japan) and the JCRB Cell Bank (Osaka, Japan), respectively; mesenchymal stromal cells (hMSCs) and human dermal fibroblasts (hDFs) were from Cambrex (East Rutherford, NJ, USA). Wild-type (WT) calvaria cells were isolated from C57BL/6N mice as described (16)
. Cbfb–/– calvaria cells were generous gifts from T. Fujita and T. Komori (Nagasaki University, Nagasaki, Japan). These cells were maintained in high glucose Dulbecco’s modified Eagle medium (DMEM) (Sigma-Aldrich, St. Louis, MO, USA) containing 10% FBS (Sigma-Aldrich), 50 U/ml penicillin, and 50 µg/ml streptomycin (Sigma-Aldrich). Col1a1GFP-ES cells were isolated as described (21)
from blastocysts carrying the Col1a1GFP transgene obtained by mating Col1a1GFP transgenic mice with WT mice. Maintenance of isolated ES cells, formation of embryoid bodies (EBs), and induction of their subsequent differentiation were performed as described (9
, 22)
.
For functional cloning using retroviral cDNA libraries, the retroviral supernatant was used to infect Col1a1GFP ES cells and was cultured for 10 days. For screening of adenoviral vectors, each adenovirus was infected at 50 multiplicities of infection. After infection, the cells were cultured in DMEM supplemented with 1x insulin-transferin-selenium + 1 (Sigma-Aldrich) and 1% penicillin/streptomycin (serum-free DMEM) or serum-free osteogenic medium, which is serum-free DMEM supplemented with 0.1 µM dexamethasone (Sigma-Aldrich), 50 mM ß-glycerophosphate (Sigma-Aldrich), and 50 µg/ml ascorbic acid phosphate (Wako Pure Chemicals Industry, Ltd., Osaka, Japan). For an analysis of calcification, von Kossa staining was performed as described (23)
.
Real-time RT-polymerase chain reaction (real-time RT-PCR)
Total RNA was extracted using an ISOGEN Kit (Wako Pure Chemicals Industry, Ltd., Tokyo, Japan) and treated with DNase I (Qiagen, Hilden, Germany). After reverse transcription using a Takara RNA PCR Kit, AMV version 2.1 (Takara Shuzo Co., Shiga, Japan), PCR was performed with the ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, CA, USA) and QuantiTect SYBR Green PCR Master Mix (Qiagen), according to the manufacturer’s instructions. The mRNA copy number of a specific gene in the total RNA was calculated as described (23)
. All reactions were run in triplicate. The primer sequences are available upon request.
Immunoblot and coimmunoprecipitation
Preparation of whole cell lysates was performed using a radio-immunoprecipitation assay buffer as described (24)
. Separated extraction of cytoplasmic and nuclear proteins was performed using an NE-PER Kit (Pierce Chemical Co., Rockford, IL, USA). Immunoblotting was performed as described (22)
using anti-HA mouse monoclonal antibody (mAb) (1:1000; Santa Cruz Biolaboratories, Santa Cruz, CA, USA), anti-Flag rabbit antibody (1:1000; Sigma-Aldrich), anti-Myc antibody (1:1000; Upstate, Lake Placid, NY, USA), anti-Smo rabbit polyclonal antibody (pAb) (H-300, 1:200; Santa Cruz Biolaboratories), anti-Runx2 mouse mAb (1:1000; MBL, Nagoya, Japan), anti-PEBP2ß mouse mAb (1:1000; MBL), or anti-actin rabbit antibody (1:1000; Sigma-Aldrich). Secondary antibodies (HRP-conjugated goat anti-mouse IgG or goat anti-rabbit IgG; Promega, Madison, WI, USA) were used at a dilution of 1:10,000.
Coimmunoprecipitation assays were performed using a Catch and Release kit (Upstate) according to the manufacturer’s instructions. After being treated with 100 µg of dithiobis (succinimidyl propionate, a reducible chemical cross-linker) (DSP, Pierce Chemical Co.) per milliliter for 20 min, the cell lysates were incubated with 5 µg of anti-Runx2 antibodies at 4°C for 4 h. Immune complexes were eluted and subjected to SDS-PAGE.
Chromatin immunoprecipitation (ChIP)
ChIP was performed using a Chromatin Immunoprecipitation (ChIP) Assay Kit (Upstate), 5 µg of anti-Runx2 antibodies, and 5 µg of anti-PEBP2ß antibodies according to the manufacturer’s instructions. PCR was performed to amplify the promoter region of the osteocalcin gene containing OSE2 site. The primer sequences are available upon request.
Generation of mouse dermal fibroblast (mDF) cell sheets
Skin tissues were obtained from the backs of 8-wk-old male transgenic mice expressing GFP ubiquitously (C57BL/6-TgN(act-enhanced GFP (EGFP))OsbC14-Y01-FM131, GFP transgenic mice, Riken Bioresource Center, Tsukuba, Japan). Isolation of mDFs was performed as described (25)
. Briefly, after trypsinization with 0.25% trypsin (Gibco BRL, Rockville, MD, USA) in Hank’s balanced salt solution (Gibco BRL), the dermis was manually separated from the epidermis and digested with 3.5% collagenase (Wako Pure Chemicals Industry, Ltd.) in DMEM. Isolated mDFs were maintained in DMEM containing 10% FBS and 1% penicillin/streptomycin. To generate cell sheets, the original approach (26)
was modified by using collagen films to support the cell sheets because dermal fibroblasts exhibited weak cell-cell adhesion. The mDFs were plated onto collagen films (CELLGEN; Koken, Tokyo, Japan) after adenoviral infection and cultured in serum-free osteogenic medium for 1 wk to induce calcification.
Transplantation of mDF cell sheets
Mice were anesthetized with ketamine/xylazine (80 and 5 mg/kg) solution through intraperitoneally injection and a round craniotomy defect (4 mm in diameter) was manually created as described (27)
. mDF cell sheets cut into a round shape (5 mm in diameter) were placed to cover the defects. At 2, 4, and 8 wk after the operation for analyses, mice were euthanized by asphyxiation with carbon dioxide. Because substantial spontaneous bone regeneration occurred at 8 wk but not at 4 wk, we chose to evaluate the induction of bone formation within 4 wk after surgery (28)
. To assess bone regeneration, radiological analysis, tissue preparation, H&E staining, and immunohistological analysis using a rabbit pAb against GFP (Molecular Probes, Inc., Eugene, OR, USA) were performed as described (23)
. The area of the regenerated bone detected by X-ray for each animal was measured using the NIH Image. The ratio of the regenerated bone area to the original defect area (RBA/ODA) was calculated and used as the index of bone regeneration. Animal experiments were performed according to the protocol approved by the Animal Care and Use Committee of the University of Tokyo.
Statistical analysis
The means of groups were compared by ANOVA and the significance of differences was determined by post hoc testing using Bonferroni’s method.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(1)
promoter (Col1a1GFP) (29)
; we call them Col1a1GFP ES cells hereafter. We previously reported restrictive expression of GFP in bone tissues of these transgenic mice (29)
(Supplemental Fig. 1A), suggesting that GFP fluorescence in Col1a1GFP ES cells is produced only when they differentiate into osteoblastic cells (Supplemental Fig. 1B). Indeed, expression of GFP fluorescence and osteocalcin mRNA showed the same temporal pattern in osteogenic cultures of Col1a1GFP ES cells (Supplemental Fig. 1C, D). These findings indicated that the site of insertion of Col1a1GFP did not interfere with its expression and that GFP fluorescence controlled by Col1a1 promoter reliably reflected osteogenic differentiation, suggesting that this system might allow us to monitor osteogenic differentiation easily, precisely, and noninvasively without analyzing differentiation markers or staining cells. Using this system, we planned to identify much more potent genes for bone regeneration than conventional ones. For identification, we set three criteria: such genes have to induce osteogenic differentiation (Fig. 1
A) within a week in serum-free medium (Fig. 1B
) and in various cell types (Fig. 1C
), including nonosteogenic cells. The first two criteria were set in order to find osteogenic stimuli that were rapid and potent enough for clinical applications and to avoid the confounding influences of cytokines and the potential contamination of pathogens contained in the serum (30)
. We set the third in order to broaden the range of the cell types that respond to the osteogenic stimuli.
|
We first tried to clone a single potent osteogenic gene through functional screening of the retroviral cDNA libraries. We chose cDNA libraries isolated from mouse embryos at embryonic day 13.5 or 15.5, because bone formation started around embryonic day 14.0 (31)
. After construction of two retroviral cDNA libraries, we infected Col1a1GFP ES cells with either one under the condition that a single copy of the retroviral genome would be integrated into a host chromosome and screened for a single gene that induced osteogenic differentiation in serum-free medium using GFP fluorescence as an indicator. Although we observed the cells for as long as 10 days, no GFP-positive cells appeared (data not shown).
We next chose five signaling pathways (BMPs, Hedgehogs, Runx2, Wnts, and insulin-like growth factor 1, or IGF-1) based on the in vivo phenotypes of their mutant animals (12
13
14
15
16)
and attempted to screen their random combinations. After preparation of adenoviral vectors encoding genes activating or inhibiting these signaling pathways (Table 1
), specific expression of each protein was confirmed by immunoblot analysis (Fig. 1A
). We infected Col1a1GFP ES cells with each combination of genes, including the neutral one encoding LacZ(35=243 combinations), then screened for the combination that induced osteogenic differentiation within a week and in serum-free medium. Although the stimulation of all signaling pathways (ALL) strongly induced GFP fluorescence within a week, stimulation of each signaling pathway failed to do so, as expected. Among all combinations, caALK6+Runx2 was found to be the minimal unit that induced GFP fluorescence as potently as ALL (Fig. 1B
). Neither the number of GFP-positive cells nor the strength of their GFP fluorescence was significantly enhanced by adding the other genes (data not shown). These data suggest that caALK6+Runx2 may be the combination meeting our criteria.
|
Osteogenic induction in stem cells and terminally differentiated nonosteogenic cells by caALK6+Runx2
To confirm that caALK6+Runx2 actually induced the osteoblast phenotype in mouse ES cells (mESs), we analyzed the expression of osteoblast marker genes and matrix calcification. caALK6+Runx2 strongly induced mRNA expression of the osteoblast marker genes within a week in serum-free medium (Fig. 2
A). As shown by von Kossa staining, caALK6+Runx2 induced matrix calcification within 10 days (Fig. 2B
), but other individual factors did not (data not shown). The level of osteocalcin expression induced by caALK6+Runx2 was comparable to that induced by ALL and that of mouse primary osteoblasts, whereas other individual factors did not induce its expression (Fig. 2C
).
|
In serum-free medium, caALK6+Runx2 exerted similar effects on human hMSCs (supplemental Fig. 2), hDFs (Fig. 3
), and terminally differentiated nonosteogenic cell lines, including NIH3T3 cells (32)
(Fig. 4
A, B) and HeLa cells (33)
(Fig. 4C
). It is worth noting that vascular endothelial growth factor A (VEGF-A) and matrix metalloproteinase-13 (MMP-13), which played important roles in angiogenesis and tissue remodeling, respectively, were strongly induced by caALK6+Runx2 in hDFs (Fig. 3A
). The results from nonosteogenic cell lines (Fig. 4)
enabled us to exclude the possibility that a small number of stem cells mingled in primary DFs might selectively expand and differentiate into osteoblasts. In addition, caALK6+Runx2 failed to induce expression of the type X collagen, a differentiation marker of hypertrophic chondrocytes (data not shown), ruling out the possibility that mineralization was induced by hypertrophic chondrocytes rather than osteoblasts.
|
|
Taken together, caALK6+Runx2 is the minimal unit that meets the optimization criteria (Fig. 3A
) and in stem cells and terminally differentiated nonosteogenic cells (Fig. 3B
), suggesting that caALK6+Runx2 is the potent osteogenic unit that meets all three criteria.
The molecular mechanisms underlying the cooperative action of caALK6+Runx2
The data so far strongly suggest the presence of a synergistic action between caALK6, which transduces BMP signaling, and Runx2 for the induction of osteogenic differentiation. Because Cbfb forms a complex with Runx2 and enhances its DNA binding and transcriptional activation (34
, 35)
, we hypothesized that Cbfb might be involved in the synergistic action. To clarify the role of Cbfb, the effects of the loss or overexpression of Cbfb on caALK6+Runx2-induced differentiation were investigated. Cbfb-KO (–/–) cells failed to induce mRNA expression of osteocalcin in response to caALK6+Runx2, which was restored by adenoviral reintroduction of Cbfb to the level of that of WT cells (Fig. 5
A). This regulation was also observed in mRNA expression of ALP and osteopontin (data not shown). In NIH3T3 stimulated by caALK6+Runx2, additional treatment with Cbfb accelerated the speed of the osteocalcin expression without enhancing its maximum expression level (Fig. 5B
). These data suggested that Cbfb was necessary for the synergistic action of BMP signaling and Runx2 and might be regulated by the two signaling pathways, leading us to focus on Cbfb in investigating the molecular mechanisms underlying the synergistic action.
|
When Runx2 was overexpressed in NIH3T3, its mRNA and protein expressions were up-regulated, with the protein being accumulated in the nucleus, which was not altered by coinfection with caALK6 (Fig. 6
A, B). At the basal level, NIH3T3 cells expressed a moderate amount of endogenous Cbfb mRNA, which was not altered by infection with caALK6, Runx2, or both (Fig. 6A
). On the other hand, the basal expression of Cbfb protein was weak, which was markedly increased in the nucleus by Runx2 or caALK6+Runx2 (Fig. 6B
). Although Runx2 and Cbfb were colocalized in the nucleus upon infection with Runx2 or caALK6+Runx2, coimmunoprecipitation analysis revealed that these two proteins bound to each other only upon infection with the latter (Fig. 6B
). Upon treatment with the ubiquitin-proteasome inhibitors such as lactacystin and MG132, the Cbfb protein level was increased without being affected by infection with Runx2 or caALK6+Runx2 (Fig. 6C
). In addition, chromatin immunoprecipitation analysis revealed that caALK6+Runx2, but not Runx2 alone, recruited Cbfb to the osteocalcin promoter whereas Runx2 was constantly recruited (Fig. 6D
).
|
Taken together, these data suggest the following molecular mechanisms. 1) Overexpressed Runx2 protects Cbfb from degradation by the ubiquitin proteasome, and both proteins accumulate in the nucleus. 2) Upon activation of BMP signaling by caALK6, Runx2 associates with Cbfb, and the complex of Runx2 and Cbfb binds to the promoters of the osteoblast-specific genes to activate their expression. This is a sequential and progressive process, a cascade of molecular interactions via regulation of Cbfb, which leads to transcription of the target gene. Thus, caALK6+Runx2 synergistically elicits a potent osteogenic signaling that is distinct from its component.
In vivo osteogenic effects of dermal fibroblast cell sheets stimulated by caALK6+Runx2
To investigate the in vivo relevance of caALK6+Runx2, we tested whether DFs stimulated by caALK6+Runx2 would have osteogenic effects on bone defects. For transplantation of DFs, we modified the cell sheet technology (26)
to create sheets of cells by using collagen films to support the cell sheets. Autologous mouse DFs (mDFs) were infected with adenoviruses expressing LacZ, caALK6, Runx2, or caALK6+Runx2, cultured on collagen membranes, then transplanted as monolayer cell sheets onto bone defects created in the mouse calvarias. To estimate the contribution of donor cells, transgenic mice expressing GFP ubiquitously (36)
were used as donors. Soft X-ray analysis revealed that while no bone formation occurred 2 wk after transplantation in either group, bone formation was strongly induced at 4 wk in the caALK6+Runx2-infected group (Fig. 7
A), which was confirmed by quantitative analysis of the regenerated bone area (Fig. 7B
). Histological analysis revealed that woven bones with marrow cavities, thicker than the original calvaria bones, appeared 4 wk after transplantation (Fig. 7C
). Immunohistochemistry for GFP revealed that both GFP-positive donor cells and GFP-negative host cells were observed in regenerated bone tissues (Fig. 7D
), suggesting that the transplants induced osteogenic differentiation of recipient cells. In contrast, cell sheets of mDFs infected with either caALK6 or Runx2 did not mineralize the matrix in culture or induce bone formation when implanted (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The screening of cDNA libraries using this system suggested that a single gene could not induce osteogenic differentiation under our experimental conditions. The integrity of the libraries seems high because we succeeded in isolating several chondrogenic genes in other experiments using them (data not shown). This result prompted us to screen combinations of genes rather than individual ones. Based on an educated guess, we chose five major osteogenic signaling pathways for the combination because random combinations of millions of genes or signaling pathways amount to an astronomical number, which is impossible to screen using this system or any technology available at this time. By setting the three criteria, we successfully identified caALK6+Runx2, which cooperatively induced osteogenic differentiation much more rapidly and potently than the conventional osteogenic genes. Two recent reports strongly support our results. Phimphilai et al. reported that the transcriptional activity of Runx2 required BMP signaling and that the sensitivity of cells to BMPs was enhanced by Runx2 (37)
. In addition, Runx2, especially its C terminus containing the nuclear matrix targeting signal and Smad interacting domain, was shown to be indispensable for execution of the BMP2 osteogenic signal (38)
. These reports indicate that BMP signaling and Runx2 require each other for their full osteogenic effects, suggesting they may act in synergism. We provided the potential molecular mechanism involving Cbfb for this synergism.
Regarding the interactions of BMP signaling, Runx2, and Cbfb, several points remain to be clarified. The first is the mechanism by which Cbfb is stabilized by Runx2. Although stabilization of Runx1 occurred by its dimerization with Cbfb (39)
, the stabilization of Cbfb was not related to the dimerization with Runx2, suggesting that Runx2 may stabilize Cbfb in an indirect manner. The second is the mechanism by which Cbfb enters the nucleus. Our data suggest that dimerization takes place in the nucleus, but Cbfb has no known nuclear localization signal. Cbfb may enter the nucleus alone by an unknown mechanism (39)
; alternatively, there may be some other protein partner(s) helping Cbfb enter the nucleus. The third is the mechanism by which BMP signaling promotes dimerization of Runx2 and Cbfb. Because Runx2 is known to interact with Smad, Smads may promote dimerization; alternatively, other proteins interacting with Runx2 may mediate BMP signaling and dimerization (40
, 41)
. These points are now under investigation. Worth noting is that Runx2+Cbfb did not up-regulate the expression of endogenous osteocalcin in our experiments, although Cbfb was shown to enhance Runx2-mediated transactivation of the osteocalcin promoter in the luciferase assay (34
, 35)
. This difference likely reflects the difference in the chromatin structure of endogenous genes and exogenous plasmids (42)
, and likely accounts for the inability of either caALK6 or Runx2 alone to induce expression of the endogenous osteocalcin gene in our experiments.
Although applications of Runx2 or BMP signaling to bone regeneration have been reported, most depended on transplantation of stem cells, osteoblast lineage cells, or cell populations containing either one (17
, 43
44
45)
. A few groups used primary fibroblasts transduced with BMP genes, but their in vitro calcification was never shown (27
, 46)
. In addition, although the use of recombinant BMPs has been studied extensively as a clinically useful procedure in bone regeneration, a large amount of BMP is required and BMP-containing devices fail in a certain percentage of cases, raising concerns over costs and safety (47
48
49)
. As Franceschi et al. pointed out, the reasons may be related to a lack of controlled and sustained BMP delivery, its short biological half-life, and the inability of its presentation to mimic the biological condition (50)
. In the above studies, the possibility was not excluded that other signaling molecules, including their combinations with BMP, might exert a stronger effect on bone regeneration because neither BMP nor Runx2 was selected through comprehensive screening. Using the combination of caALK6+Runx2 identified through screening by the cell-based sensor, we succeeded in inducing both the expression of osteoblast marker genes and in vitro calcification in terminally differentiated fibroblasts and in inducing rapid bone regeneration by transplantation of a monolayer sheet of fibroblasts transduced with the combination.
In the present study, using the cell-based sensor system, we successfully identified a potent combination of genes for bone regeneration that helped broaden cell sources to terminally differentiated adult nonosteogenic cells. Our approach using cell-based sensors may provide tools to clarify mechanisms underlying regeneration of various tissues or, more generally, may help identify in vivo effective signaling factors, including their combinations for various biological processes including developmental and pathological ones.
|
ACKNOWLEDGMENTS |
|---|
Received for publication November 6, 2006. Accepted for publication January 11, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Ohba, F. Yano, and U.-i. Chung Tissue Engineering of Bone and Cartilage IBMS BoneKEy, November 1, 2009; 6(11): 405 - 419. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
