The authors state that they have no conflicts of interest.
Restoration of Bone Mass and Strength in Glucocorticoid-Treated Mice by Systemic Transplantation of CXCR4 and Cbfa-1 Co-Expressing Mesenchymal Stem Cells†
Article first published online: 29 DEC 2008
Copyright © 2009 ASBMR
Journal of Bone and Mineral Research
Volume 24, Issue 5, pages 837–848, May 2009
How to Cite
Lien, C.-Y., Chih-Yuan Ho, K., Lee, O. K., Blunn, G. W. and Su, Y. (2009), Restoration of Bone Mass and Strength in Glucocorticoid-Treated Mice by Systemic Transplantation of CXCR4 and Cbfa-1 Co-Expressing Mesenchymal Stem Cells. J Bone Miner Res, 24: 837–848. doi: 10.1359/jbmr.081257
- Issue published online: 4 DEC 2009
- Article first published online: 29 DEC 2008
- Manuscript Accepted: 23 DEC 2008
- Manuscript Revised: 12 JUN 2008
- Manuscript Received: 10 MAR 2008
- mesenchymal stem cells;
- bone marrow homing;
Transplantation of gene-modified mesenchymal stem cells (MSCs) in animals for bone regeneration therapy has been evaluated extensively in recent years. However, increased endosteal bone formation by intravenous injection of MSCs ectopically expressing a foreign gene has not yet been shown. Aside from the clearance by lung and other tissues, the surface compositions of MSCs may not favor their bone marrow (BM) migration and engraftment. To overcome these hurdles, a gene encoding the chemokine receptor largely responsible for stromal-derived factor-1 (SDF-1)-mediated BM homing and engraftment of hematopoietic stem cells (HSCs), CXCR4, was transduced into mouse C3H10T1/2 cells by adenovirus infection. A dose-dependent increase of CXCR4 surface expression with a parallel enhanced chemotaxis toward SDF-1 in these cells after virus infection was clearly observed. Higher BM retention and homing of CXCR4-expressing MSCs were also found after they were transplanted by intramedullary and tail vein injections, respectively, into immunocompetent C3H/HeN mice. Interestingly, a full recovery of bone mass and a partial restoration of bone formation in glucocorticoid-induced osteoporotic mice were observed 4 wk after a single intravenous infusion of one million CXCR4-expressing C3H10T1/2 cells. In the meantime, complete recovery of bone stiffness and strength in these animals was consistently detected only after a systemic transplantation of CXCR4 and Cbfa-1 co-transduced MSCs. To our knowledge, this is the first report to show unequivocally the feasibility of ameliorating glucocorticoid-induced osteoporosis by systemic transplantation of genetically manipulated MSCs.
Osteoporosis is a worldwide health problem and a tremendous economic burden. For instance, in the United States alone, >2 million fracture incidences at a cost of $17 billion were estimated for 2005.(1) Even though the fracture rates in osteoporotic patients could be reduced by bisphosphonate therapy,(2–4) hormone replacement therapy (HRT),(5,6) selective estrogen receptor modulators (SERMs),(6) and calcitonin,(7) along with calcium and vitamin D,(8–10) restoration of their bone mass and strength by these antiresorptive regimens has as yet been achieved. On the other hand, improvements in both the quality and quantity of bone were reported in patients receiving “anabolic therapy.” By the use of PTH, sodium fluoride, growth hormone, or statins, this regimen stimulates new bone formation by acting primarily on osteoblasts that not only increases the production of bone matrix but also reverses the microarchitectural deterioration(11) and therefore may provide significant benefits to patients in whom antiresorptive therapy has proven insufficient.
Mesenchymal stem cells (MSCs) are adult stem cells present in a wide variety of tissues that are capable of differentiating into various mesenchymal and non-mesenchymal lineages.(12–17) In addition, these cells have been shown to play an important role in hematopoiesis,(18) bone physiology,(19) and in part participate in the pathophysiology of bone diseases.(20) Because of the osteogenic differentiation potential and the relative ease of isolation and expansion, MSCs became promising materials for treating various bone degenerative disorders including osteoporosis. This idea was further supported by the preclinical findings that the autologous MSCs transplanted intravenously could engraft in bone marrow (BM).(21–25) Not surprisingly, systemic transplantation of BM-derived MSCs has already been applied to treat children with severe osteogenesis imperfecta (OI). Even though short-term clinical improvements were seen in some of these patients, long-term benefits were not observed thus far.(26–28) In other words, unless the problems of poor BM homing and engraftment of MSCs after their vascular administration could be solved, effective treatment of osteoporosis by systemic transplantation of these cells will be difficult to achieve.(29)
BM homing for hematopoietic stem cells (HSCs) is a coordinated, multistep process requiring the interplay between chemokines, growth factors, proteolytic enzymes, and adhesion molecules.(30) Among them, the interaction between stroma-derived factor-1 (SDF-1, also known as CXCL12) and its receptor, CXCR-4, has been identified to be most pivotal.(31,32) Constitutively expressed in human and murine BM endothelium,(33) this chemokine and receptor pair is essential for BM seeding of the stem cells during fetal development and for definitive repopulation in adult mice receiving BM transplants. Accordingly, SDF-1, applied either in vitro(34) or in vivo,(35) has been shown to enhance the engraftment of human HSCs in mice. Interestingly, a CXCR4-dependent BM migration has also been observed in a small subset of MSCs that express this chemokine receptor on their surface.(36) Besides BM, the SDF-1–CXCR4 interaction seems to be critical for the migration of MSCs to the impaired brain(37) and other injured tissues.(38)
In light of the importance of the SDF-1/CXCR4 axis in BM homing and engraftment of stem cells and the potential of MSCs in bone regeneration therapy, we decided to introduce CXCR4 into mouse C3H10T1/2 MSC-like cells (referred to as MSCs even though they are not primary MSCs) through adenovirus infection and examined its effects on homing and engraftment of these cells and the feasibility of using systemic transplantation of CXCR4-expressing MSCs to treat steroid-induced osteoporosis in mice. In the meantime, because bone formation requires osteogenic differentiation of MSCs that is controlled mainly by a master transcription factor, core binding factor α-1 (Cbfa-1),(39) we further analyzed the therapeutic effect of intravenous injection of CXCR4 and Cbfa-1 co-transduced MSCs on osteoporosis.
MATERIALS AND METHODS
Murine C3H10T1/2 cells were grown in DMEM (Sigma-Aldrich) supplemented with 10% FBS (Biological Industries), 100 U/ml penicillin, 100 μg/ml streptomycin, and 0.25 mg/ml amphotericin B (PSA; Biological Industries).
Preparation of recombinant adenoviruses
The human CXCR4 gene digested from plasmid pCMV-SPORT6-CXCR4 (obtained from the Genome Research Center of National Yang Ming University) by EcoRI and XhoI was inserted into the pShuttle-CMV (AdEasy XL Adenoviral vector system; Stratagene) to obtain pShuttle-CMV-CXCR4. Preparation of the recombinant adenovirus carrying human CXCR4 (AdCXCR4) was conducted as described previously. A similar procedure was applied to produce viruses carrying the genes encoding human Cbfa-1 (AdCbfa-1), enhanced green fluorescent protein (AdEGFP), and firefly luciferase (AdLuc-F, kindly provided by Dr. Ming-Hong Tai of the Department of Medical Education and Research, Kaohsiung Veterans General Hospital, Taiwan, R.O.C.), respectively.
Virus titer determination and cell infection
Virus titer was determined by infecting Ad293 cells with serial logarithmic dilutions of viruses, and the endpoint titer (multiplicity of infection [MOI]) = 1) was defined as the highest dilution that caused an apparent cytopathic effect (CPE) 48 h after infection. The optimal virus titer for transducing C3H10T1/2 cells was determined primarily as described.(60) Briefly, cells were incubated with different amounts of adenoviruses carrying a lacZ gene (AdLacZ) in serum-free DMEM medium for 1 h before viruses were removed and cells were washed twice with PBS. After being cultured under regular conditions for 48 h, cells were stained with 5-bromo-4-chloro-3-indolyl-β-d-galactoside (X-gal; Sigma-Aldrich). The MOI resulting in >90% of the cells being stained positively by X-gal (blue) was chosen as the optimal titer for cell infection.
C3H10T1/2 cells, after being infected with different titers of AdCXCR4 (MOI 200, 400, and 800), were fixed with 2% paraformaldehyde at 4°C for 10 min and stained with a monoclonal anti-human CXCR4 antibody (10 μg/ml; Sigma) at 4°C for 30 min, followed by incubating with a rhodamine-conjugated anti-mouse IgG at 4°C for 30 min. Flowcytometry was subsequently performed to analyze surface expression of CXCR4.
Transwell migration assay
Cell migration was analyzed using Boyden chambers with 8-μm pore membranes (Corning). Four days postinfection, C3H10T1/2 cells were seeded onto the upper chamber at a density of 1 × 104 per well, and 0, 10, and 50 ng/ml of recombinant SDF-1 (R&D) was used as chemoattractant. After 6-h incubation, cells migrated to the undersurface of the filters were stained by Giemsa and counted microscopically.
BM retention assay using luciferase-tagged MSCs
Four-month-old female C3H/HeN mice (National Laboratory Animal Center, Taipei, Taiwan) were anesthetized by 2% isoflurane/air mixture by the Open Circuit System (Stoelting) before 1 × 106 C3H10T1/2 cells co-infected with AdLuc-F (MOI 400) plus AdCXCR4 (MOI 600) or AdLuc-F (MOI 400) plus AdLacZ (MOI 600) were injected, respectively, into the right and left femoral cavities of each animal. Briefly, the skin around the mouse knee was incised, and the leg was folded to expose the knee joint. Cells (resuspended in 10 μl PBS) were injected through the joint into femur cavity using 29-G insulin needles. Immediately and 3 and 7 days after transplantation, mice were intraperitoneally injected with 150 mg/kg d-luciferin (Xenogen), and in situ luciferase activities (bioluminescence) were measured by an in vivo imaging system (IVIS; Xenogen 50).
BM homing assay using quantum dots-labeled MSCs
C3H10T1/2 cells infected, respectively, with AdLacZ and AdCXCR4 (MOI 800) were labeled with quantum dots using the Qtracker Cell Labeling Kits (Invitrogen). After being washed with PBS twice, cells (1 × 106) were injected into the tail vein of 4-mo-old female C3H/HeN mice. Three hours and 7 days after transplantation, animals were killed, and single cell suspensions were prepared from tissues including BM, brain, heart, lung, liver, and spleen using a cell strainer (BD Falcon). Flowcytometry was applied to analyze the homing efficiencies of the transplanted cells to these tissues by measuring the fluorescence emitted from the quantum dots taken by them.
BM engraftment assay using β-galactosidase-expressing MSCs
One million C3H10T1/2 cells co-infected with AdLacZ (MOI 400) plus AdEGFP (MOI 800) or AdLacZ (MOI 400) plus AdCXCR4 (MOI 800) were injected, respectively, into the tail veins of 4-mo-old female C3H/HeN mice. One week after transplantation, animals were killed, and engraftment of these cells to bone and lung was examined by X-gal staining.
BM homing efficiency of EGFP-expressing MSCs
One million C3H10T1/2 cells co-infected with AdEGFP (MOI 400) plus AdCXCR4 (MOI 800) or AdLacZ (MOI 800) were injected into the tail vein of 4-mo-old female C3H/HeN mice. One week after transplantation, animals were killed, and marrow cells from both femurs were flushed out and plated. One day after plating, the attached BM stromal cells were stained with anti-CXCR4 antibodies as described above. Flowcytometry was carried out to examine the expression levels of EGFP and CXCR4.
BM homing efficiency of firefly luciferase-expressing MSCs
Virus infection of C3H10T1/2 cells and their transplantation and harvest after tail vein injection were performed essentially as above described except that MOI 100 was used for AdLuc-F co-infection. One day after harvesting, luciferase activity in the attached BM stromal cells was examined by an in vivo imaging system (IVIS; Xenogen 50) immediately after the addition of 75 mg/kg d-luciferin (Xenogen).
Animal model of osteoporosis
Blood samples were collected from 4-mo-old female C3H/HeN mice before and after they were injected intraperitoneally with 10, 20, or 50 mg/kg/d dexamethasone (Sigma-Aldrich) for 21 consecutive days. Bone mass loss in these animals was examined by measuring their serum levels of the N-telopeptide of type I collagen (NTX) using an ELISA kit (Osteomark). Daily injection of 10 mg/kg dexamethasone for 3 wk was sufficient to induce 50% of bone loss in these mice, which was not worsened by higher dosages of dexamethasone (data not shown). A similar protocol was subsequently applied in two separate sets of experiments using 7-mo-old and 5-wk-old female mice.
Analysis of bone formation
Systemic transplantation of C3H10T1/2 cells in osteoporotic mice was carried out as described above. Serum osteocalcin levels in young animals were measured before and every 7 days after transplantation, whereas those in adult mice were analyzed before and every 2 wk after cell injection using ELISA kits (Invitrogen).
Four-point bending test
Right femurs collected from mice receiving various treatments were loaded to failure in four-point bending using ProLine table-top machines Z100 (Zwick Roell) at a displacement rate of 0.05 mm/s and constant force of 10 N. Bones were positioned in the test apparatus with the dorsal side in compression and ventral side in tension. Mechanical properties measured included strength (ultimate force that the specimen sustained) and stiffness (the slope of the initial, linear portion of the load-deformation curve).
Femurs were collected and stored at −20°C until analyzed by pQCT (STRATEC XCT-2000). The femur was arbitrarily divided into proximal, middle, and distal regions. Analysis was carried out in each 4-mm distance of femurs, and BMD was expressed as the average of the values from cortical and trabecular portions.
Each data point in the figures and table represents mean ± SD for at least three independent determinations. Statistically significant differences (p < 0.05) between each group were determined by Student's t-test and one-way ANOVA plus Tukey's test.
Ectopic expression of human CXCR4 in mouse C3H10T1/2 cells enhances their migration toward SDF-1
To facilitate BM homing and engraftment of MSCs, a human CXCR4 gene was introduced in mouse C3H10T1/2 MSC-like cells through adenovirus infection. Surface expression of CXCR4 on these cells was analyzed by flow cytometry using a human-specific anti-CXCR4 antibody. In comparison with those infected by a control virus (AdLacZ), surface CXCR4 levels in cells transduced with the corresponding human gene were indeed proportional to the amounts of virus infected (Fig. 1A), despite that some endogenous CXCR4 molecules were also found on their surface (data not shown). Transwell migration assay was subsequently performed to examine whether CXCR4-overexpressing C3H10T1/2 cells responded more effectively to its natural ligand, SDF-1. In comparison with their parental cells, mouse MSCs transduced with human CXCR4 showed a dramatic increase of migration toward SDF-1 (Fig. 1B), which might be accounted for by a high sequence similarity (∼89%) of this chemokine receptor between two species.
CXCR4-expressing MSCs exhibit higher BM retention efficiency
Because SDF-1/CXCR4 signaling has been postulated to be a principal axis regulating retention, migration, and repopulation of HSCs during homeostasis and injury, we examined whether an ectopic expression of CXCR4 in MSCs also enhanced their BM retention. Firefly luciferase-tagged C3H10T1/2 cells infected with AdLacZ and AdCXCR4 were injected directly into the left and right femurs of syngeneic C3H/HeN mice, respectively. As can be seen, higher BM retention of CXCR4-transduced MSCs was clearly found in these animals. In fact, ∼8-fold more CXCR4-expressing MSCs were retained in bone 7 days after their intraosseous transplantation (Fig. 2B).
BM homing of MSCs is facilitated by CXCR4 expression
Four different approaches were next applied to assess the effect of enforced expression of CXCR4 in MSCs on their in vivo homing after systemic transplantation (tail vein injection). First, 1 × 106 quantum dot-labeled C3H10T1/2 cells transduced without or with CXCR4 were intravenously injected into C3H/HeN mice. Three hours and 7 days later, homing efficiency of the transplanted MSCs in various tissues was analyzed by flow cytometry. As can be seen, whereas homing of the intravenously infused C3H10T1/2 cells to brain, heart, spleen, and liver was not affected by the presence of CXCR4, upregulation of this chemokine receptor significantly enhanced BM homing and dramatically decreased lung trapping of mouse MSCs (Figs. 3A and 3B). A favorable BM homing with a reduced lung trapping in MSCs resulted from CXCR4 upregulation were also observed in LacZ-expressing cells (Fig. 3C). Furthermore, much higher bioluminescence was detected in BM stromal cells harvested from femurs of mice receiving C3H10T1/2 cells co-transduced with CXCR4 and luciferase (Fig. 3D), suggesting that BM migration and/or survival of mouse MSCs might be increased by CXCR4 overexpression. Finally, flow cytometric analysis showed a higher BM engraftment of AdCXCR4-infected EGFP-expressing MSCs (Figs. 3E and 3F).
Bone resorption in osteoporotic mice is dramatically reduced by a single systemic transplantation of CXCR4-expressing MSCs
To evaluate whether MSCs with higher BM homing ability had better therapeutic effects on systemic bone loss, intravenous transplantation of these cells was performed in steroid-induced osteoporotic animals.(40) Serum levels of the NTX, a good indicator for bone matrix degradation, were used to assess bone resorption in C3H/HeN mice before and after they received intraperitoneal injection of 10, 20, or 50 mg/kg of dexamethasone for 21 consecutive days. To our surprise, ∼50% bone mass loss was found in all animals regardless of the dosage of steroid (data not shown). Because the age of animals might affect therapeutic outcome, MSC transplantation was performed in two-young (e.g., 5 wk old) and adult (7 mo old)—groups of mice separately. Although significant decrease in bone resorption could be detected in both young and adult mice as early as 1 wk after a single intravenous injection of 106LacZ- or Cbfa-1–expressing MSCs, the best result accomplished by this treatment in both groups was ∼70% of their normal levels (Fig. 5). In contrast, bone mass recovery in young mice receiving CXCR4-transduced or CXCR4 plus Cbfa-1 co-transduced MSCs reached a plateau (∼90% of normal) 3 wk after transplantation (Fig. 4A), whereas a continuous increase that eventually returned to its normal level 4 wk after MSCs infusion was observed in adult animals (Fig. 4B). Intriguingly, bone resorption reversed by transplantation of CXCR4-expressing MSCs in young mice was completely abolished by incubating these cells with a CXCR4-neutralizing antibody for 2 h before their infusion (Fig. 4A).
Bone matrix formation in osteoporotic mice is markedly increased after a single systemic transplantation of CXCR4-expressing MSCs
In the meantime, to assess the effects of MSCs transplantation on bone matrix formation (osteoblast activity) in these animals, serum osteocalcin levels were also measured by ELISA. As can be seen, intravenous injection of non-CXCR4- expressing MSCs only marginally enhanced matrix formation in both groups (Fig. 5). In contrast, significant increase of matrix formation in young mice receiving either CXCR4-transduced or CXCR4 and Cbfa-1 co-expressing MSCs could be detected as early as 14 days after transplantation (Fig. 5A). Moreover, a better matrix formation recovery (85% versus 70% of normal levels) resulting from the above-mentioned treatment was also found in young animals 4 wk after MSCs infusion.
Bone stiffness and strength in osteoporotic mice are greatly improved by a single systemic transplantation of CXCR4-expressing MSCs
Finally, to assess the recovery of bone mechanical properties, femurs were collected from adult mice 1 mo after cell transplantation, and their stiffness and strength were determined by the four-point bending test.(41) Even though no significant improvement in bone stiffness was found in young mice receiving AdLacZ- or AdCbfa-1–infected MSCs, transplantation of these cells into adult animals markedly enhanced their bone stiffness (Table 1). Accordingly, better therapeutic efficacies from intravenous infusion of CXCR4-transduced MSCs were also observed in adult mice. On the other hand, complete restorations of both bone stiffness and strength in two age groups were consistently detected only when CXCR4 and Cbfa-1 co-expressing MSCs were transplanted (Table 1). In agreement with the above findings, an overall density of proximal, middle, and distal parts of femurs in adult osteoporotic mice were significantly increased by a single intravenous injection of AdCXCR4- or AdCXCR4 plus AdCbfa-1–infected MSCs. Interestingly, a stronger density increase in the proximal region of femurs was detected in animals transplanted with the latter (Fig. 6).
Osteoporosis is a devastating systemic bone degenerative disease resulted primarily from an imbalance between resorption and formation on endosteal and trabecular bone surfaces. For patients affected by this disorder, fractures at the wrist, spine, and hip occur most frequently, which are associated with significant morbidity and, in case of the latter two, excessive mortality.(42) Although fracture risks could be reduced and bone mass might be increased respectively by antiresorptive and anabolic therapies in these patients, alternative regimens are still being sought because even a combination of both may not guarantee a better outcome.(43) In this aspect, the potential of systemic and/or local transplantations of MSCs, the most critical osteoblastic precursors, for osteoporosis therapy, have been evaluated extensively in preclinical settings(29,44–46) because these cells play important roles not only in normal skeletal homeostasis but also in fracture repair. Intriguingly, even though systemic transplantation of MSCs in mice did not seem to increase bone formation,(44,45) transient clinical benefits have nevertheless been found in children with severe OI who had received intravenous infusion of allogeneic BM stromal cells.(26–28)
The major hurdle for treating osteoporosis with systemic transplantation of MSCs is their poor BM homing and engraftment efficiencies. During our search for novel solutions for these problems, we were inspired by the finding that SDF-1/CXCR4 signaling is crucial for BM homing, engraftment, and repopulation of HSCs.(31,32) Interestingly, surface expression of this chemokine receptor has also been detected in a small subset of MSCs that migrate specifically to BM.(36) The CXCR4/SDF-1 signaling, moreover, seems to be critical for the migration of MSCs to certain damaged tissues.(37,38) Taking into account of the importance of CXCR4 in tissue-specific migration of MSCs, we decided to introduce the gene encoding this chemokine receptor into mouse MSCs and examined the in vivo behavior of these cells before assessing their usefulness in osteoporosis therapy in a mouse model.
In contrast to an earlier observation that surface CXCR4 expression rarely occurs in human MSCs,(36) the majority of the ectopically expressed human CXCR4 molecules were found in the plasma membrane of mouse MSCs-C3H10T1/2 cells (Fig. 1A), which drastically facilitated the chemotaxis of these cells toward SDF-1 (Fig. 1B). In accordance, a much higher BM retention was found in CXCR4-transduced C3H10T1/2 cells when they were injected directly into the bone cavity (Fig. 2). In fact, ectopic CXCR4 expression increased the retention rate of these cells (3 days after transplantation) from ∼30%(29) to > 70% of day 0. As expected, BM homing efficiency of C3H10T1/2 cells was also markedly enhanced by CXCR4 overexpression, which might be accounted at least in part by a reduction of their lung trapping (Fig. 3A and B), a postulated major barrier for efficient BM engraftment of the intravenously infused stem cells.(47) Interestingly, in contrast to CXCR4-overexpressing human CD34+ cells transplanted in mice,(48) no preferential spleen homing of CXCR4-transduced MSCs was detected, suggesting that the SDF-1/CXCR4 signaling is not the sole determinant for specific migration of stem cells to this organ. Clearances of cells in brain, heart, lung, and liver 7 days after transplantation were observed, which indicated that microenvironment cues may be important for transplanted MSC survival (Fig. 3B). Another notable observation was that, even though surface expression of human CXCR4 was detected in ∼4% of the BM stromal cells recovered 7 days after transplantation, only 1/10th of them expressed EGFP (Fig. 3E), suggesting that either the expression of this reporter might interfere with BM homing/engraftment of MSCs or the bone marrow microenvironment might silence the expression of nonessential genes such as EGFP in these cells.
Having shown a drastic increase in BM homing of the intravenously injected mouse MSCs by an ectopic CXCR4 expression, we next asked whether systemic transplantation of CXCR4-transduced MSCs offered a better therapeutic efficacy in osteoporosis. A glucocorticoid-induced osteoporotic mouse model(40) was used because bone mass loss in these animals might result from a combination of an apoptosis of osteoblasts,(49,50) a decrease in osteoprotegerin (OPG), and an increase in RANKL.(49,51,52) Additionally, because animals' age might affect the therapeutic outcome, both young (5 wk old) and adult (7 mo old) C3H mice were used. Contrary to the ineffective MSC transplantations applied earlier by others,(29) a partial but significant bone resorption decrease was detected in recipients of both ages after a single intravenous injection of 1 × 106LacZ- or Cbfa-1-transduced C3H10T1/2 cells (Fig. 4), which might be accounted for by a higher basal expression of surface CXCR4 in these cells (data not shown). Fittingly, a complete reversion of bone resorption was found in mice receiving a similar number of mouse MSCs transduced with CXCR4 alone or CXCR4 plus Cbfa-1 (Fig. 4). These results suggested that a single intravenous injection of MSCs expressing CXCR4 or CXCR4 plus Cbfa-1 could alleviate bone resorption in osteoporotic animals. It will be of interest to examine in the future whether osteoclastic cell activity in BM is reduced (or suppressed) by the transplanted MSCs. Intriguingly, no apparent bone formation recovery in the these mice could be detected after LacZ- or Cbfa-1-expressing MSCs were transplanted (Fig. 5), indicating that an efficient homing/engraftment is necessary for these cells to either undergo osteoblastic conversion directly or to induce osteogenic differentiation of the preexisting stem/progenitor cells. The critical contribution of CXCR4 ectopic expression in MSCs to their therapeutic effects in osteoporosis was further shown by a complete abrogation of both bone mass and formation recoveries in young animals receiving MSCs that had been treated briefly with a CXCR4-neutralizing antibody before their transplantation (Figs. 4A and 5A). Fittingly, a drastic downregulation of surface CXCR4 triggered by this antibody was observed (Supplemental Fig. 1). In the meantime, co-expression of Cbfa-1 in MSCs failed to improve their efficacies in restoring bone formation that might be explained by the following reasons. First, because SDF-1/CXCR4 signaling has recently been shown to play a crucial role in bone morphogenetic factor (BMP)-2–induced osteogenic differentiation of some MSC-like cells,(53)CXCR4-transduced C3H10T1/2 cells might automatically differentiate into osteoblasts, therefore bypassing the need of Cbfa-1. Second, either the recipients' BM microenvironment per se was sufficient for inducing osteoblastic differentiation of the transplanted MSCs or the homed naïve MSCs were capable of guiding the osteogenesis of stem/progenitor cells residing in the BM of these animals. However, the last two possibilities might be negated by the atrophy of MSCs and the incompetence of these cells in osteogenic differentiation caused by BM microenvironment alterations found in osteoporotic individuals.(54,55) On the other hand, much better bone formation recovery induced by MSC transplantation was found in young osteoporotic mice, suggesting less severe deterioration of the BM microenvironment in these animals.
In agreement with the above observations, transplantation of LacZ- or Cbfa-1–transduced MSCs only induced a partial improvement of bone stiffness in adult mice (Table 1). In contrast, both stiffness and strength of a long bone in adult animals were fully restored by an intravenous injection of CXCR4-expressing MSCs. However, this treatment improved bone stiffness in young mice only modestly. Complete bone stiffness and strength recoveries in both groups were consistently detected when CXCR4 and Cbfa-1 co-expressing MSCs were transplanted. If the mechanical parameters were better indicators of bone functions, Cbfa-1 co-expression in MSCs is necessary for maximizing their bone-regenerating effects. Finally, our treatment might be helpful in preventing and/or reducing hip fractures resulting frequently from a loss of cortical bone in the proximal femurs,(56) because a greater BMD recovery of this region was found in adult mice transplanted with CXCR4 and Cbfa-1 co-expressing MSCs (Fig. 6). Even though histomorphometry was not used to further analyze the bones of these animals, which would be a minor deficiency of this study, surrogates including bone resorption and matrix formation (biochemical assays), as well as bone stiffness, strength, and mineral density (biomechanical tests), were nonetheless measured, which should be sufficient to compensate for the lack of histomorphometry.
To our knowledge, this is the first study to unequivocally show the feasibility of ameliorating osteoporosis by a systemic transplantation of gene-modified MSCs. However, several questions need to be addressed before this approach is evaluated in clinical settings. First, its therapeutic efficacy in other osteoporotic models (i.e., resulted from ovariectomy and/or aging) where bone mass loss was more severe should be examined. Second, to mimic more closely a real clinical situation, C3H10T1/2 pluripotent fibroblasts should be replaced by primary or early passage syngeneic or allogenic mouse MSCs that express low (or negligible) CXCR4. Third, the human CXCR4 gene should be substituted by its mouse counterpart to minimize the risk of rejection even though neither an immediate immune response to human protein (i.e., a dramatic increase in serum total IgG levels and the presence of anti-CXCR4 antibodies; Supplemental Figs. 2 and 3) nor a deleterious tissue damage (i.e., pathological analyses of liver, lung, and spleen) were detected in recipient mice. Fourth, Cbfa-1 might be replaced by other osteogenic genes such as Dlx5(57) or Osterix (Osx)(58) because formation of the ectopic mineralized foci rather than mature bone trabeculas in animals receiving subcutaneous implantation of Cbfa-1–modified MSCs has already been reported.(59) Fifth, even though the use of replication-defective adenoviral vectors offers many advantages in gene therapy, they nonetheless trigger strong innate immune responses when administered directly into humans. In addition, these vectors may cause the destruction of transduced cells, resulting in local damage and inflammation.(60,61) Hence, if our approach was further developed for clinical application, the aforementioned drawbacks should be considered even when MSCs used for transplantation were transduced by a mere ex vivo approach. Finally, a combination of MSC transplantation with any of the currently available antiresorptive and/or anabolic therapies should be applied for testing their synergism in alleviating osteoporosis.
The authors thank Tom K. Kuo for technical assistance and Dr. S. Sundar for assistance with pQCT. Financial support was provided by the National Science Council (NSC95-2320-B-010-043-MY3 and NSC95-2320-B-010-066-MY2) of Taiwan to Y.S.
- 3Osteoporosis Methodology Group and The Osteoporosis Research Advisory Group 2002 Meta-analyses of therapies for postmenopausal osteoporosis. III Meta-analysis of risedronate for the treatment of postmenopausal osteoporosis. Endocr Rev 23: 517–523., , , , , , , , , ,
- 6Osteoporosis Methodology Group and The Osteoporosis Research Advisory Group 2002 Meta-analyses of therapies for postmenopausal osteoporosis. IV. Meta-analysis of raloxifene for the prevention and treatment of postmenopausal osteoporosis. Endocr Rev 23: 524–528., , , , , , , , ,
- 8Osteoporosis Methodology Group and The Osteoporosis Research Advisory Group 2002 Meta-analyses of therapies for postmenopausal osteoporosis. VIII: Meta-analysis of the efficacy of vitamin D treatment in preventing osteoporosis in postmenopausal women. Endocr Rev 23: 560–569., , , , , , , , , , , ,
- 9Osteoporosis Methodology Group; Osteoporosis Research Advisory Group 2004 Calcium supplementation on bone loss in postmenopausal women. Cochrane Database Syst Rev CD004526., , , , , , , , , , , ,
- 322003 Stem cell mobilization. Hematology (Am Soc Hematol Educ Program) XXX: 419–437., , , , , ,
- 382006 Migration of bone marrow and cord blood mesenchymal stem cells in vitro is regulated by stromal-derived factor-1- CXCR4 and hepatocyte growth factor-c-met axes and involves matrix metalloproteinases. Stem Cells 24: 1254–1264., , , , , , ,
- 522006 Glucocorticoid regulation of osteoclast differentiation and expression of receptor activator of nuclear factor-kappaB (NF-kappaB) ligand, osteoprotegerin, and receptor activator of NF-kappaB in mouse calvarial bones. Endocrinology 147: 3613–3622., , ,