In Vitro Characteristics and In Vivo Immunosuppressive Activity of Compact Bone-Derived Murine Mesenchymal Progenitor Cells



In contrast to the considerable amount of data that documents the biological properties of mesenchymal progenitor cells from human and other species, there is still paucity of information about mouse counterparts, as their purification and culture expansion procedures remain rudimentary. In the present study, murine mesenchymal progenitor cell (muMPC) culture was developed by explant culture of collagenase-digested bone fragments after removal of the released cells. During cultivation, fibroblastoid cells sprouted and migrated from the fragments, followed by adherent monolayer development. The cells exhibited homogenous surface antigen profile and presented in vitro multipotential differentiation along osteocyte, chondrocyte, and adipocyte lineages, as evaluated by matched cell or matrix staining and reverse transcription polymerase chain reaction techniques. Also, the surface antigenic epitope changed and potential of proliferation and multidifferentiation decreased with successive subculturing. Functional investigations demonstrated that these cells supported in vitro hematopoiesis and suppressed lymphocyte cell proliferation triggered by ConA or allogeneic splenocytes. Furthermore, muMPCs prolonged the mean survival time of skin grafts across the major histocompatibility barrier (H2b → H2d), suggestive of the immunosuppressive effects in vivo. The findings demonstrate that muMPCs obtained with this simple protocol are similar in property to their marrow counterparts, and thus, the protocol described here could be used for further investigations in mouse physiological and pathological models.


Mesenchymal progenitor cells (MPCs) are generally defined as a cell population that has the potential to differentiate along multiple lineages such as chondrocytes, osteocytes, adipocytes and hematopoiesis-supporting stromal cells both in vitro and in vivo [13]. This potential has rendered them as an intriguing source of cells for cellular replacement therapy and tissue engineering. Recent data also suggest that MPCs are characterized by their immunosuppressive activity in vitro and support for ex vivo expansion of hematopoietic progenitor cells by means of direct cell–cell contact and cytokine secretion [48], enhancing their therapeutic appeal in the setting of allogeneic bone marrow transplantation. However, MPCs exist predominantly within bone marrow at an exceedingly low frequency (approximately 1 per 106 mononuclear cells) and thus have prompted research interest in developing various approaches for MPC isolation and culture expansion.

MPC culture was first introduced by Friedenstein et al. and modified by others, using the physical propensity of MPCs to adhere to plastic flasks [911]. This methodology has been widely employed for bone marrow MPC propagation from human beings [1], rats [12], pigs [13], rabbits [14], cats [15], and dogs [16]. However, the standard method of plastic adherence has failed to yield relatively homogenous MPCs from murine bone marrow, mainly because of hematopoietic cell contamination [17].

Several groups have reported their independent work on murine MPC (muMPC) purification from bone marrow by magnetic selection [18], retroviral infection [19], or unique culture systems [17, 2022]. In our own previous work, muMPC culture was developed by addition of bone fragments with mouse bone marrow cells in the presence of basic fibroblast growth factor (basic FGF) and bone fragment-conditioned medium [23]. Furthermore, mouse colony-forming unit-fibroblast (CFU-F), a strong in vitro correlative of MPCs, has been successfully enriched by Simmons et al., who clearly identify the femoral bone itself as a richer source of muMPCs than the marrow plug within it [24, 25]. Based on these findings and those from other investigators demonstrating that MPCs from human compact bones are similar in property to their counterpart in bone marrow [26] and that human bone-derived MPC culture can be developed by adherent culture of either enzyme-treated bone fragments or the released cells [2731], we postulate that murine counterparts are also able to be purified with these simple methods. To test this hypothesis, mouse bones were digested with collagenase and the released cells or the remaining bone fragments were propagated as previously reported [23, 26]. Interestingly, muMPC culture could be developed only by inoculation of digested bone chips other than the suspending cells. These putative muMPCs were homogenous in surface epitopes, possessed the classic tridifferentiation potential, inhibited in vitro splenocyte proliferation, and maintained hematopoietic cells in long-term culture as effectively as mouse primary stromal cells. Furthermore, muMPC delivery could prolong the survival time of skin grafts transplanted across the MHC barrier (H2b → H2d). These findings add more proof that compact bones are an alternative source of muMPCs, and the protocol described in the present study provides novel cues for further investigations in mouse pathological models.

Materials and Methods


Normal inbred C57BL/6 (H2b) and BALB/c (H2d) female mice were purchased from the Laboratory Animal Center, Chinese Academy of Military Medical Science (Beijing; and were housed in conventional cages. All experiments in this study were performed in accordance with the Chinese Academy of Military Medical Sciences Guide for Laboratory Animals.

Culture and Expansion of muMPCs

The femurs and tibiae were collected from 2- to 3-week-old C57BL/6 female mice. The epiphyses were removed, bone marrow was flushed out, and the bone cavities were washed thoroughly by drawing and expelling with a syringe. The compact bones were excised into chips of about 1 mm3 into plastic culture dishes, suspended in α-modified minimal essential medium (α-MEM) containing 10% fetal bovine serum (FBS) from selected lots in the presence of 1 mg/ml of collagenase II (Sigma-Aldrich, St. Louis,, and incubated for 2 hours at 37°C with shaking at a speed of 200 rpm. The released cells were aspirated, and the bone fragments were washed three times with α-MEM, followed by incubation in α-MEM containing 10% selected FBS at 37°C in 5% CO2. Medium was changed every 3–4 days. The adherent cells were harvested by trypsin digestion and passaged at an initial density of 1,500 cells per cm2 thereafter. The cells at the third passage were used for the following experiments except otherwise described. Also, the released cells after digestion were propagated separately in the same culture medium in some experiments.

Flow Cytometry

Adherent cells at the indicated passages were retrieved by trypsin digestion and aliquots of 1 × 106 cells were labeled with fluorescein isothiocyanate- or phycoerythrin-conjugated monoclonal antibodies against mouse CD11b, CD29, CD31, CD34, CD44, CD45, CD105, and stem cell antigen-1 (Sca-1) (all products from eBio-Science, San Diego, for 30 minutes at room temperature in the dark. After washing twice, events were acquired by FACSCalibur (Becton, Dickinson and Company, Franklin Lakes, NJ,, and data analysis was conducted with WinMDI 2.9 software (Joseph Trotter, The Scripps Institute, La Jolla, CA) after gating for the designated population.

Multipotent Differentiation Assay

Osteogenic differentiation was induced by osteoinductive medium as previously described in our laboratory [23]. A histochemical kit (Sigma) was used to assess alkaline phosphatase (ALP) activity according to the manufacturer's protocol.

An aggregate culture system was used according to the protocol described for chondrogenic differentiation of marrow muMPCs [22]. Briefly, aliquots of 2 × 105 cells were pelleted and cultured in serum-free inductive medium containing 10ng/ml recombinant human transforming growth factor β-3 (R&D Systems Inc., Minneapolis, and 500 ng/ml recombinant human bone morphogenetic protein-6 (BMP-6) (R&D Systems). Pellet cultures were maintained for 21 days, with medium changes every 3–4 days. The pellets were fixed in paraformaldehyde, embedded in paraffin, and sectioned for toluidine blue staining.

To assess adipogenic capacity, cells were propagated to confluence and maintained in α-MEM containing 10% FBS, 10−7 M dexamethasone, and 10 μg/ml insulin (Sigma-Aldrich). Intracellular lipid droplets were generally evident after 5 days, and lipid accumulation was further identified by in situ oil red-O staining as routinely described [23, 32].

Maintenance of CFU-F Formation and Multipotential Differentiation

Aliquots (100 cells per well) of muMPCs at the indicated passages were seeded in 12 replicates into a 24-well tissue culture plate and were maintained in culture for 12 days. All visible colonies larger than 5 mm in diameter were counted after Giemsa staining. For evaluating the multiple differentiation capacities, trilineage-associated transcripts in cells after dexamethasone induction were detected by reverse transcription-polymerase chain reaction (RT-PCR), with that of hypoxanthine phosphoribosyl-transferase (HPRT) as a control for mRNA quality.


Cells at different passages were detached and plated into plastic flasks at a density of 1,000 cells per cm2. Culture was maintained for 2 weeks in α-MEM containing 10% FBS in the presence of 10−8 M dexamethasone for inducing nonspecific differentiation as previously described [33]. After 2 weeks, total cellular RNA was extracted using TRIZOL reagent (Gibco, Grand Island, NY, and mRNA was reverse-transcribed with 15-mer poly-(d)-T (Promega; Madison, WI, for cDNA amplification. Primers used were as those reported previously [20]. PCR products were separated on a 2% agarose gel containing 1 μg/ml ethidium bromide and visualized under UV light.

Long-Term Culture of Hematopoietic Cells on muMPC or Stromal Feeders

Stromal feeder layers for supporting maintenance of long-term culture-initiating cell (LTC-IC) were established as previously described with some modifications [34]. Briefly, mononucleated cells from adult female C57/BL bone marrow were separated by Ficoll density gradient centrifugation and were seeded in LTC medium (Myelocult M5300; Stem Cell Technologies, Vancouver, British Columbia, Canada, containing 10−6 M hydrocortisone. Nonadherent cells were removed by thorough washing after 24 hours, and culture was maintained until an adherent layer formed. The adherent cells were retrieved by trypsin/EDTA digestion and seeded into a gelatin-precoated 96-well culture plate at 2 × 104 cells per well. Cells were irradiated (20 Gy) to eliminate hematopoietic activity without affecting the ability of the stroma to support hematopoiesis. muMPCs were also collected, plated, and irradiated as described above. Lineage-negative (Lin) cells from adult C57BL/6 bone marrow were prepared by labeling mononuclear cells with a cocktail of antibodies against CD3, CD4, CD8, Ter-119, Mac-1, and Gr-1 (eBioScience) and subsequently reacted with immunomagnetic beads (DynalBiotech; Brown Deer, WI, and separated by a magnetic particle concentrator (Dynal MPC-S) per the manufacturer's instructions. In a typical experiment, about 10% of marrow mononuclear cells were harvested as Lin cells. Graded number of enriched Lin cells (6,000, 3,000, 1,500, 750, and 375 cells per well) were overlaid in 12 replicates per dilution onto the adherent cell layers and maintained in LTC medium supplemented with 10−6 M hydrocortisone in humidified 5% CO2 atmosphere at 33°C. Culture was continued for 5 weeks with weekly changes of half of the medium, followed by cell collecting by trypsinization. All the cells from an individual well were plated in MethoCult 3434 (Stem Cell Technologies) into one dish for determination of colony-forming cells (CFCs). Seven days later, the dishes were scored as positive or negative for the presence or absence of CFCs. Frequency of LTC-IC was calculated, and the data were further statistically analyzed with the L-Calc software (Stem Cell Technologies).

In Vitro Splenocyte Proliferation Assays

Graded numbers of muMPCs (104, 5 × 103, 103, and 0 per well) were seeded in a gelatin-precoated 96-well culture plate and maintained at 37°C for 6 hours before 20 Gy γ-irradiation. Mononuclear spleen cells from C57BL/6 mice were prepared by Ficoll density gradient centrifugation, and aliquots of 2 × 105 cells were added into each well and cultivated in the presence of 20 μg/ml Con A or 2 × 104 irradiated splenocytes from BALB/c. The total volume was 200 μl per well. In mitogen stimulation experiments, cells were maintained at 37°C for 72 hours and pulsed with 5 μCi of 3H-thymidine deoxyribonucleoside/ml for an additional 8-hour culture. In mixed lymphocyte reaction experiments, 3H-TdR (1 μCi/well) was pulsed for 18 hours after a 5-day culture. Cells were harvested onto glass fiber filters, and radioactivity was measured on a Wallac Microbeta Trilux 1450-02P (PerkinElmer, Wellesley, MA, The results are expressed as mean cpm for at least triplicate cultures.

Tail Skin Grafting

muMPCs (1 × 106) from female C57/BL were injected intravenously into BALB/c mice, and sex-matched skin grafting was conducted as described elsewhere [35]. Briefly, full-thickness tail skin (around 0.5 × 0.5 cm2) from C57/BL (allografting) or BALB/c (syngeneic grafting) was transferred to the sides of the recipient's tail, from which an equivalent amount of skin had been removed. Each mouse received two allografts and one syngeneic graft. The grafts were protected with a glass tube and monitored daily by visual inspection. A graft was considered rejected when more than 90% was necrotic, and grafts transferred to the tail on which the syngeneic skin had not survived were screened out for further analysis. On day 10, some of the skin grafts were removed for histological examination after H&E staining.

Statistical Analysis

Data were presented as the mean ± standard deviation and analyzed with Student's t test. Data from skin grafting were analyzed with the log-rank test. A p value less than .05 was considered statistically significant.


Establishment of muMPC Culture from Digested Bone Fragments

The concept is generally accepted that it is difficult to isolate mesenchymal stem/progenitor cells from murine marrow. Interestingly, previous data have definitely shown that human MPCs derived from compact bone exhibit properties similar to those of their counterparts from bone marrow [26] and that mouse compact bone presents a richer source than bone marrow for muMPC purification [24, 25]. To verify whether muMPC culture could be developed as reported with human MPCs [26], the suspended cells from digested mouse compact bones were seeded into plastic dishes in our initial experiments. However, several problems hindered its practical application. First, it was difficult to disperse the cells into single-cell suspension by trypsin digestion because of obvious extracellular matrix accumulation and, therefore, cell aggregates usually formed, and subculture could not be readily developed. Second, CD45-positive cells contaminated the cultures (Fig. 1), although CD45 negative cells gradually overgrew with passaging. Third, most of the adherent cells presented ALP activity, as evaluated by histochemical staining (data not shown). These results were consistent in several independent experiments.

Therefore, digested bone fragments, instead of the enzyme-released cells, were inoculated as previously described in humans [31]. Fibroblastic cells appeared around the fragments during 48-hour cultivation (Fig. 2A, 2B). Also, some small colonies were evident apart from the fragments after 72 hours (Fig. 2C), at which time, culture medium was changed after removal of nonadherent cells and tissue debris. Typically, cells reached −80% of confluence within 5 days (Fig. 2D). The cells proliferated and grew rapidly and could be serially passaged twice weekly at a split ratio of 1:4 or 1:3 before the 10th passage; however, some of the cells became flattened and polygonal with further subculturing, and cell senescence was usually obvious in culture at passage 15.

Immunophenotyping Characteristics

The cell surface antigen profile of putative muMPCs was analyzed by flow cytometric technique. As shown in Figure 1, more than 90% of the cells expressed Sca-1 (a murine hematopoietic and mesenchymal stem/progenitor cell marker), CD29 (β1-integrin), and CD44 (receptor for hyaluronate and osteopontin). Meanwhile, they were homogenously negative for CD34 (non-specific marker for progenitors of mouse hematopoietic, endothelial, and mesenchymal lineages), CD45 (pan-hematopoietic marker), CD11b (monocyte marker), and CD31 (endothelial cell marker), demonstrating that the culture was devoid of hematopoietic and endothelial cell lineages. However, cells seemed heterogeneous in CD105 (endoglin) expression in that the positive proportion of the passage 3 cells was around 70%.

To address whether antigenic profile of these putative muMPCs varied with increasing population doubling, cells at different passages were collected for flow cytometric analysis. As shown in Figure 1, the proportion of CD105+ cells decreased progressively, and no CD105 expression was evident from the 11th passage. It is also of interest that reduction in CD29 and increase in CD34 expression was observed on cells at late passages.

Multilineage Differentiation

The commonly used standard for identifying MPCs is their inherent capacity to differente into osteoblasts, chondroblasts, and adipoblasts under the appropriate conditions [1]. To test whether the cells met these criteria, the passage cells were maintained in matched inductive medium, and cultures in muMPC medium served as controls. Although the cells revealed specific differentiation into osteocytes and adipocytes after induction as assessed by ALP and oil red-O staining, none of the controls displayed characteristics of differentiative phenotypes (Fig. 3A–3D). In complement to the results of histochemical analysis, the cells after induction exhibited mRNA expression of osteocalcin (OCN) and peroxisome proliferator-activated receptor γ2 (PPARγ2) (Fig. 4), which are molecular hallmarks of differentiated osteoblasts and adipocyte, respectively. On the other hand, little evidence of proteoglycan secretion was observed in pellet culture for chondrogenic induction (data not shown), which is consistent with the previous findings in marrow MPCs from BALB/c and C57/BL mice [22]. Moreover, transcripts for collagen II were detectable in dexamethasone-induced cells, as assessed by sensitive RT-PCR, and similar constitutive expression of HPRT mRNA was observed in treated and untreated cells (Fig. 4). Therefore, the cells prepared with this protocol are designated muMPCs herein.

To investigate whether the multipotentials could be maintained during in vitro expansion, muMPCs at passage 3, 5, 7, 9, 11, and 13 were coaxed for nonspecific differentiation, as previously reported [33], and RT-PCR was performed to analyze the expression levels of the trilineage-specific genes. As shown in Fig. 4, the expressions of OCN, PPARγ2, and COL II were evident in differentiated cells at passages 3, 5, 7, and 9, but not passages 11 and 13, at which the positivity was exclusively observed for OCN. It is noteworthy that OCN mRNA was also detectable in control cells at passage 11 and 13, indicating of spontaneous differentiation along the osteocytic lineage.

Maintenance of CFU-F Formation

To observe whether the proliferation features could be sustained with serial passaging, CFU-F frequencies of the cultured cells at passage 1, 3, 5, 7, 9, 11, and 13 were evaluated. As indicated in Fig. 5, the CFU-F frequency remained relatively constant in culture before passage 11, around 6% (p = .3515, .8582, .7897, .4975, and .1653 for P3, P5, P7, P9, and P11, respectively, compared with that of P1). However, it decreased significantly at passage 13 (around 2%) (p < .0001 compared with that of P1), and CFU-F could not be detected from cells at passage 15.

Support of Hematopoiesis In Vitro

A number of reports have documented that human marrow MPCs can maintain LTC-IC of ex vivo-expanded hematopoietic cells [7, 8, 36, 37]. To observe whether muMPCs derived from long bones have activity similar to that of human marrow counterparts, in vitro maintenance of LTC-IC in marrow Lin cell population was evaluated with muMPCs or primary marrow stromal cells as the feeders. The results from two separate experiments showed that after a 5-week culture, LTC-IC frequency was 1/2,935 (range: 1/3,397 to 1/2,535; test for internal inconsistency: p = .2390) in muMPC cultures, which was slightly higher than that in stromal cell cultures (mean: 1/3,491, range: 1/4,061 to 1/3,002; test for internal inconsistency: p = .074). Statistical analysis showed that the difference is not significant (p = .4092), revealing the functional similarity of muMPCs and marrow stromal cells to support in vitro hematopoiesis.

muMPCs Inhibited In Vitro Lymphocyte Proliferation Triggered by Con A or Allogeneic Cellular Stimuli

It has been well documented that human or murine MPCs suppress in vitro T cell activation stimulated by nonspecific mitogens or allogeneic lymphocytes [46, 23]. To observe whether muMPCs defined in this study displayed similar features, cells of graded numbers were cocultured with splenocytes from the same strain (C57BL/6) in the presence of Con A. Consistent with previous reports, muMPCs inhibited Con A-stimulated splenocyte proliferation in an MPC dose-dependent manner (Fig. 6A). The proliferation activity was completely restrained if the MPC to splenocyte ratio exceeded 1:40, although the majority of the splenocytes in culture were viable, as evaluated by trypan blue exclusion test (data not shown). Further, muMPCs suppressed lymphocyte activation by allogeneic cellular stimuli in one-way mixed lymphocyte reaction, and likewise, the suppressive effect exhibited itself in a dose-dependent way (Fig. 6B).

muMPCs Prolonged Skin Survival in an MHC-Incompatible Skin Grafting Model

To further address whether immunosuppressive effects of muMPCs could occur in vivo, skin grafting was performed in a mouse model (C57BL/6→BALB/c) with or without muMPC pretransfusion. The results showed that muMPC delivery gave rise to a significant survival prolongation of allogeneic skin grafts (10.13 ± 1.51 vs. 11.95 ± 2.92 days, n = 26 and 20, respectively; p = .029) (Fig. 7, left). Histological examination on day 10 showed intact epidermis in grafts from muMPC-treated mice, although obvious leukocyte infiltration was evident in the subcutaneous layer, further proving the inhibitory activities that muMPCs exert on the ongoing immune response in vivo.


Several independent reports have described protocols in detail on the purification and culture expansion of murine marrow MPCs, all of which aim to circumvent the problem of hematopoietic cell contamination prominently observed in murine marrow culture by means of immunomagnetic selection of marrow cells to eliminate blood cells [21], addition of FGF-2 to maintain the stemness of MPCs in culture [23], or successive subculturing to promote MPC outgrowth [22]. However, some of the procedures seem not easily handled, or in some instances, hematopoietic cell contamination remains in the initial several passages, indicating the need to search for alternative sources for muMPC isolation in an operational way.

In fact, the presence of mesenchymal stem or stem-like cells in murine compact bone has been previously reported by Short et al., who enriched muMPCs successfully from enzyme-released cells by immunodepletion of Lin hematopoietic cells and further FACS selection of Sca-1+ cells [25]. Also, a relatively simple method of isolating human bone-derived MPCs by inoculation of enzyme-released cells for developing an adherent culture has been well described [26]. To test the postulate that MPCs residing in mouse compact bones could also be purified with this simple methodology, the suspended cells from murine long bone fragments after collagenase digestion were cultured in our initial experiments. Although most of the adherent cells were fibroblastoid in morphology, the presence of CD45 positive cells was evident in culture as evaluated by flow cytometry, denoting the presence of hematopoietic cell contamination. This finding is not surprising, as hematopoietic cells reside along the bone surface, and enzyme digestion could release them with other cell types from the densely structural arrangement. The proliferation and/or differentiation of the contaminated hematopoietic cells could be further supported by osteoblasts and/or muMPCs in culture [38], yielding a heterogeneous cell population in adherence. Hence, it seems impractical to isolate and culture expand pure muMPCs by cultivation of the released cells because they were not homogenous per se, and some of them, mainly osteoblasts in the cell suspension, possessed the propensity to adhere to plastic flasks. The high-level ALP expression supports this possibility and indicates that some of them may represent preferably as osteoblasts or mesenchymal progenitors at hierarchically late stage (data not shown).

In the present study, therefore, muMPC culture was initiated by inoculation of the remaining bone fragments after digestion as described for purification of human bone-derived MPCs [31]. In contrast, an adherent cell layer yielded within several days, which seemed much earlier than was observed in human MPC culture [31]. This cell population comprises morphologically homogenous fibroblast-like cells, and the population doublings could reach at least 30× with successive subculture. Flow cytometric analysis showed that these cells were positive for CD29, CD44, Sca-1, and CD105 but negative for hematopoietic and endothelial cell markers, including CD11b, CD45, and CD31, which is consistent with results reported in the literature for their counterparts from human and murine bone marrow and is thus expected of mesenchymal progenitors. Furthermore, the capacity of these cells to differentiate along osteogenic, chondrogenic, and adipogenic pathways in response to classic stimuli was identified. The cells underwent osteogenic differentiation as verified by cell ALP positivity and expression of osteocalcin specific for differentiated osteoblasts. In vitro chondrogenesis was revealed by the expression of collagen type II gene, although cells could secrete little proteoglycan in pellet culture system. The presence of intracellular lipid droplets and PPAR-γ expression proved the differentiation competence into adipocytic lineage. The results here are in agreement with the routinely accepted standard for MPC identification. Therefore, it would be reasonable to conclude that these cells arising from the propagation of digested bone fragments enable themselves to be defined as bona fide muMPCs.

It is noteworthy, however, that muMPCs maintained in the defined medium at increasing passages progressively lost their multiple differentiation potentials, accompanied by a decrease in CFU-F contents. Cells at passages 11 and 13 could not differentiate along chondroblast and adipoblast lineages; at these points, however, OCN mRAN transcripts could be detected both in control and dexamethasone-treated cells. This phenomenon can also be observed in the culture expansion of human marrow MPCs. The differentiation potential for chondrogenesis, adipogenesis, and, subsequently, osteogenesis disappears with increasing cell doubling [2]. These findings indicate the need for further optimization of culture systems for the maintenance of multipotentials during extensive ex vivo expansion by means of, perhaps, addition of growth factors such as FGF-2, platelet-derived growth factor BB, and epidermal growth factor as observed in marrow muMPCs [17]. Yet further investigations might be needed to clarify the effects of growth factors on the proliferation and differentiation of MPCs. Evidence in a recent study has shown that a supplement of FGF-2 in human MPC culture favors their osteogenic [39, 40] or chondrogenic potential [41].

Furthermore, we have found phenotypic changes of muMPCs with successive subcultures, in accordance with the progressive loss of multiple potentials. The expression of CD105, a vascular-specific TGF-β coreceptor and a molecular indicator for MPCs and long-term hematopoiesis reconstructing cells, decreases dramatically, whereas that of CD34 increases progressively, and CD29 expression is evident at passage 13. Interestingly, Sca-1, a surface marker for mesenchymal or hematopoietic stem/progenitor cells, is highly expressed as marrow MPCs from the same mouse strain [22] and remains unchanged with passaging. Previous knockout studies have demonstrated that CD105 is actively involved in angiogenesis [42] and that Sca-1 plays a critical role in the self-renewal of stem/progenitor cells from both hematopoietic and mesenchymal lineages [43, 44]. Meanwhile, CD34 expression has been shown on human [45] and murine [22] marrow MPCs, which is observed on bone-derived muMPCs defined in this study exclusively at late passages. Interestingly, both marrow [22] and bone-derived muMPCs from the same strain (C57/BL) seem not display in vitro chondrogenic potential in aggregate cultures in spite of the discrepancy in CD34 expression. Hence, further detailed investigations might be needed to clarify the exact association between phenotypic characteristics and differentiation potential of muMPCs.

Accumulating data have proven that human marrow or placenta-derived MPCs support in vitro hematopoiesis by means of secretion of cytokines and direct contact with hematopoietic progenitors, in that MPCs enhance ex vivo expansion of CD34+ cells in the presence of cytokine cocktail and sustain LTC-IC for 5 weeks in the absence of additive cytokines [7, 8, 36, 37]. In this study, we have shown that feeder layers of muMPCs maintain week 5 LTC-IC as effectively as those of primary stromal cells. This finding denotes that muMPCs might be a crucial cellular component for hematopoietic niches within compact bone and play an active role in the control and regulation of proliferation and differentiation of hematopoietic stem/progenitor cells in this unique niche [38, 46].

Here, we also demonstrate that compact bone-derived muMPCs suppress in vitro lymphocyte proliferation elicited by mitogen or allogeneic splenocyte, which is consistent with the previous reports on murine marrow MPCs [23]. In an attempt to further expand our knowledge about whether bone muMPCs exert similar action in vivo, we employed a skin grafting model across MHC disparity to evaluate the regulatory effects of muMPC transfer on immune alloreactivity. The results showed that muMPCs significantly prolong the survival time of skin grafts between mice across strains (C57BL/6→BALB/c), which is consistent with the results in a baboon model [47] and suggestive of the suppressive effect of muMPCs on the ongoing immune process in vivo. These findings provide further support for the clinical use of MPCs in settings such as the prevention of graft versus host disease in allogeneic bone marrow transplantation.

Taken together, our data here demonstrate that muMPCs share similar phenotypic and functional properties with their marrow counterparts, including plastic adherence, fibroblast-like morphology, phenotypic characteristics, tridifferentiation potential in appropriate conditions, hematopoietic support, and immune suppression both in vitro and in vivo. This study provides new clues for further investigation on mesenchymal progenitors in murine physiological and pathological models.

Figure Figure 1..

Flow cytometric analysis on epitope characteristics of the adherent cells at passage 3 or passages indicated. The cultures were initiated with the released cells (P3-R) or digested bone chips (other panels). Open histogram shows background signal, and shaded histogram shows reactivity with the indicated antibodies.

Figure Figure 2..

Morphological features of bone-derived murine mesenchymal progenitor cells. Mouse femurs and tibiae were cut into fragments and enzyme-digested for 2 hours. After removal of the suspended cells, the digested chips were cultivated into plastic dishes. (A): Forty-eight hours later, fibroblastoid cells sprouted from the chips. (B): More cells migrated and distributed raying out from the fragment after additional 24 hours. (C): Subsequently, colony formation was evident. (D): An adherent layer of spindle-shaped cells developed 5 days after the initial culture.

Figure Figure 3..

Differentiation of muMPCs. Cells were subjected to classic inductive stimuli (B and D) or incubated in control medium (A and C) as described in Materials and Methods and stained for alkaline phosphatase (A and B) and oil red-O (C and D).

Figure Figure 4..

Reverse transcription-polymerase chain reaction analysis on the lineage-specific gene expression of muMPCs at different passages after dexamethasone treatment. Lane A: cells in muMPC medium; Lane B: cells treated with dexamethasone. HPRT is shown as a control for mRNA sample quality. The data are representative of two individual experiments. Abbreviations: muMPC, murine mesenchymal progenitor cell; HPRT, hypoxanthine phosphoribosyl-transferase; OCN, osteocalcin; PPAR, peroxisome proliferator-activated receptor γ2; COL-II, collagen type II.

Figure Figure 5..

Maintenance of CFU-F formation. Cells at different passages were seeded into 24-well plate and colonies larger than 5 mm in diameter were counted after 12 days. The x-axis shows different passages, and the y-axis represents the CFU-F frequency (%). Data are from the pooled results of three separate experiments. Abbreviations, CFU-F, colony-forming unit-fibroblast; p, passage.

Figure Figure 6..

muMPCs inhibit in vitro C57BL/6 lymphocyte proliferation upon Con A (A) (n = 22 for each group) or BALB/c splenocyte (B) (n = 6 for each group) stimuli. Shaded bars represent the cpm values in coculture with graded numbers of muMPCs, and open bars represent those in the absence of muMPCs. Statistical analysis showed that the cpm values in non-muMPC groups were significantly higher than those in any corresponding groups with muMPCs. The data are representative of two separate experiments. Abbreviation: muMPC, murine mesenchymal progenitor cell.

Figure Figure 7..

muMPCs prolong skin graft survival. Skin grafts from C57BL/6 were transplanted onto the tails of BALB/c mice and graft survival was monitored daily. Left: the x-axis shows the survival days and the y-axis represents the percentage of survived grafts. Statistical analysis demonstrates that muMPC transfusion significantly prolongs the graft survival (10.13 ± 1.51 vs. 11.95 ± 2.92 days, n = 26 and 20, respectively; p = .029). Right: Representative H&E sections of skin grafts on day 10 from muMPC-untreated (A) or muMPC-treated (B) recipients, showing the necrotic dermis (A) and leukocyte infiltration in the subcutaneous layers (A and B). Abbreviations: ctr, control; muMPC, murine mesenchymal progenitor cell.


We thank Dr. Bing Liu, Dr. Xiaosan Chen, and Prof. Delin Du for their helpful comments and critical reading of the manuscript. We are also grateful to Prof. Yinglin Lu for his assistance in histological examination and to Ying Wu and Chunmei Hou for their helpful assistance in flow cytometry and animal experiments. This work was supported by National 863 Grant of China (No. 2003AA205170). Z.G., H.L., and X.L. contributed equally to this work.


The authors indicate no potential conflicts of interest.