Circulating mesenchymal stem cells with abnormal osteogenic differentiation in patients with osteoporosis




While the role of osteoclasts in bone loss has been well investigated, the involvement of osteoblast-lineage cells has not been completely elucidated. Several genes contribute to normal osteoblastic differentiation from mesenchymal stem cells (MSCs), but an understanding of their role in the pathogenesis of osteoporosis is still lacking. The present study was undertaken to evaluate a possible alteration of osteogenic gene expression as a mechanism contributing to bone loss.


We studied the osteogenic differentiation process in MSCs obtained from the peripheral blood of 31 patients with osteoporosis and 20 normal donors. The cells were evaluated by colony-forming unit–fibroblastic assay and cultured in osteogenic medium to analyze the transcription factors runt-related transcription factor 2 (RUNX-2) and Sp7 and the bone-related genes COL1A1, SPARC, and SPP1 after 3, 8, and 15 days of differentiation. In addition, to determine possible differences between the 2 groups in terms of osteoclastic and osteoblastic activation, we quantified the osteoprotegerin (OPG) and RANKL levels in the supernatants of osteoblastic culture.


Circulating MSCs were increased in osteoporosis patients compared with normal donors. In contrast, gene expression analysis revealed down-regulation of RUNX2, Sp7, COL1A1, SPARC, and SPP1 in patients with osteoporosis, associated with a lower OPG:RANKL ratio.


These results suggest that an alteration of osteoblastic differentiation may contribute to the pathogenesis of osteoporosis. The noninvasive approach used in the present study could be proposed as a useful tool for studying mesenchymal involvement in bone diseases.

Osteoporosis is a common age-related skeletal disease characterized by bone loss, which increases skeletal fragility. Postmenopausal osteoporosis is caused primarily by estrogen deficiency after cessation of ovarian function, leading to bone loss due to increased bone turnover and negative bone-remodeling (1). Bone remodeling plays a major role in the maintenance of the skeleton's mechanical integrity and involves a well-coordinated balance between bone formation (by osteoblasts) and bone resorption (by osteoclasts) (2).

An imbalance in bone remodeling, in which bone formation is not able to compensate for bone resorption, is the main mechanism leading to osteoporosis and bone fragility. While the role of increased activity of osteoclasts has been extensively studied, the involvement of osteoblasts and osteocytes has not been well elucidated. Osteoblasts derive from mesenchymal stem cells (MSCs), which are multipotent progenitor cells with the capacity to differentiate into connective tissue cells (3). Although bone marrow is believed to be the main source of these precursor cells (4), MSCs can be isolated from peripheral blood (5, 6) as well as from nonhematopoietic tissue such as adipose tissue, trabecular bone, dermis, dental pulp, synovium, and lung (7). Osteoblast differentiation, though not completely understood, is known to be a well-coordinated process regulated by signal transduction networks and specific transcriptional factors such as runt-related transcription factor 2 (RUNX-2) and Osterix (Osx).

RUNX-2 (also called Cbfa1) is known to be essential for osteoblastic differentiation, because mice with its null mutation exhibit as complete lack of bone (8). RUNX2 gene expression is up-regulated in proliferative chondrocytes, and it is expressed early in osteoblastic differentiation. This gene can be stimulated by multiple signal transduction pathways (9), and it can directly stimulate the transcription of osteoblast-related genes such as those encoding osteocalcin, type I collagen, osteopontin, and collagenase 3 by binding to specific enhancer regions containing the core sequence, PuCCPuCA (10). RUNX-2, together with Osx (Sp7), a recently identified zinc finger–containing transcription factor expressed in the osteoblasts of both endochondral and membranous bones, are the main transcription factors involved in osteoblastogenesis (11).

In Osx-null mice, no bone formation occurs even though RUNX2 is expressed (11), suggesting that Sp7 acts downstream of RUNX-2 during bone development. Moreover, other differences between RUNX2-null and Osx-null mutant animals have been reported, such as inhibition of chondrocyte hypertrophy with deletion of RUNX2 but not with deletion of Osx (12), suggesting that while RUNX-2 promotes chondrocyte terminal maturation, Osx does not. Studies have indicated that Osx is regulated by various growth factors in osteoblasts and chondrocytes (13), and it has been shown that tumor necrosis factor regulates expression of Osx by inhibiting the transcriptional activity of its promoter (14).

In addition, the expression of bone-related marker genes, i.e., osteopontin (SPP1), osteonectin (SPARC), and (COL1A1), is activated during osteogenic differentiation (15, 16). In a previous study, we found that, during osteogenic differentiation of circulating MSCs, expression of the transcription factor RUNX2 and of COL1A1, SPARC, and SPP1 was increased whereas, in contrast, Sp7 expression declined after differentiation (6), consistent with its role in osteoblast maturation.

Although a strong genetic component has been identified, the molecular processes involved in osteoporosis and fragility fracture and the osteoblasticic potential expressed in individuals with osteoporosis are not well understood. Therefore, to investigate the osteoblastic potential in osteoporosis, we harvested MSCs from peripheral blood and, using gene expression analysis, assessed their capacity to differentiate in osteogenic cells. The osteoprotegerin (OPG):RANKL ratio, relevant in bone remodeling, was also evaluated, in order to explore the potential ability to prevent osteoclastic activation (17).


The study patient group comprised 31 postmenopausal women with osteoporosis who had been referred to the metabolic bone disease center at our institution between January 2006 and November 2008. As controls, we studied 20 healthy postmenopausal women who were matched with the patients for age and anthropometric parameters. Informed consent was obtained from all study participants, and the study was approved by the local ethics committee.

All women included in the study underwent densitometry of the lumbar spine and hip (QDR Discovery Acclaim; Hologic, Waltham, MA). Osteoporosis was diagnosed according to World Health Organization parameters, i.e., a lumbar spine or femoral T score lower than –2.5 SD (18).

To determine the presence of vertebral fractures, a radiologic examination of the dorsolumbar spine was performed on all patients with osteoporosis. A fracture was defined as a >20% reduction of the ratio between anterior and posterior height or between middle and posterior height, and vertebral deformities were classified into 3 types (wedge, biconcave, or crush) and 3 degrees (mild [grade 1; ratio <20%], moderate [grade 2; ratio 25–40%], or severe [grade 3; ratio >40%]). The severity of fractures was expressed as a spine deformity index (SDI), calculated as the total of all fracture grades. Patients with a history of clinical vertebral fracture occurring in the preceding 12 months were excluded from the study. Serum levels of calcium, phosphate, parathyroid hormone, carboxy-terminal telopeptide of type I collagen, alkaline phosphatase, and 25-hydroxyvitamin D, and the urinary calcium excretion rate, were measured in all osteoporosis patients to rule out secondary causes of osteoporosis.

In order to be included as normal donors, women had to have a T score above −2 SD and no history of bone fractures. Biochemical and radiologic analyses were not performed in normal donors for ethical reasons, but a standardized clinical evaluation was carried out to exclude possible comorbidities. Exclusion criteria included premature menopause (<45 years old), and presence of a disease or use of a drug that could affect bone or metabolism of calcium or phosphate.

Cell preparation and culture.

Cell culture experiments and gene expression analyses were performed on blood samples obtained from each study subject. MSC-like cells were isolated using 50 ml of blood, with a method that included 2 Ficoll procedures to deplete hematopoietic cells by application of a cocktail of antibodies, as previously described (6). The selected cells obtained were washed in phosphate buffered saline (PBS) and then cultured in 24-well plates with Mesencult Basal Medium (StemCell Technologies, London, UK) and 10% stimulatory supplements for human MSCs (catalog no. 05402; StemCell Technologies).

Analysis of cell phenotype.

We analyzed the cell phenotype for CD3, CD14, CD19, CD45, and CD34 markers at the RNA level, as previously described (6). This method allows the phenotypic analysis of small numbers of cells obtained with stringent stem cell purification techniques (19). RNA extraction and real-time reverse transcriptase–polymerase chain reaction (RT-PCR) were performed on human MSC samples from osteoporosis patients and normal donors as described below.

Quantification of human mesenchymal progenitors (colony-forming unit–fibroblastic [CFU-F]).

To perform CFU-F assays, selected cells were seeded at 3 different concentrations (4 × 105, 2 × 105, and 1 × 105) in 24-well plates and incubated for 14 days in medium containing stimulatory supplements for human MSCs, under a humidified atmosphere of 5% CO2 at 37°C. After 14 days, the cell suspension was removed and adherent cells were fixed with methanol, dried, and stained for 5 minutes with Giemsa staining solution. After washing with distilled water, colonies consisting of >50 cells were scored using an inverted microscope. The number of CFU-Fs from each donor was expressed as the average obtained under the 3 different culture concentrations.

Osteogenic differentiation.

For osteogenic differentiation studies, the MSCs were plated onto 48-well plates at a density of 5 ×104 cells/well. Osteogenic differentiation was achieved using osteogenic medium containing 15% osteogenic stimulatory supplements (StemCell Technologies), 10−8M dexamethasone, 3.5 mM β-glycerophosphate, and 50 μg/ml ascorbic acid (StemCell Technologies). After 3, 8, or 15 days of osteogenesis induction, supernatant and cells were harvested and stored at –80°C until use.

Alkaline phosphatase staining.

Alkaline phosphatase staining was performed with Alkaline Phosphatase kit no. 85 according to the instructions of the manufacturer (Sigma-Aldrich, St. Louis, MO). The osteogenic medium was removed after 3 days or 15 days of incubation, cultures were washed twice with PBS, and staining solution was applied. After incubation, the staining solution was removed, and the cultures were washed 3 times with distilled water to remove excess color.

Extraction of total RNA.

Total RNA was extracted from each cell pellet using the RNeasy Mini kit (Qiagen, Chatsworth, CA) with DNase I treatment. The amount and purity of the RNA was checked by measuring absorbance at 260 nm and 280 nm; a ratio of 1.8–2.0 was taken to indicate purity. The absence of degradation of the RNA was confirmed by RNA electrophoresis on a 1.5% agarose gel containing ethidium bromide.

Reverse transcription.

First-strand complementary DNA (cDNA) was generated using the First Strand cDNA Synthesis Kit with random hexamers, according to the protocol recommended by the manufacturer (GE Healthcare, Piscataway, NJ). RT product was aliquoted in equal volumes and stored at –80°C.

Real-time RT-PCR.

Messenger RNA (mRNA) was quantified by the relative standard curve method (Chemistry guide; Applied Biosystems, Foster City, CA). PCR was performed with a total volume of 50 μl containing 1× TaqMan Universal PCR Master Mix, No AmpErase UNG, and 5 μl of cDNA from each sample. Predesigned gene-specific primers and probe sets for each gene (Assays-on-Demand gene expression products [CD3, Hs00174158_m1; CD14, Hs02621496-s1; CD19, Hs00174333_m1; CD45, Hs00174541_m1; CD34, HS00156373_m1; Sp7, Hs00541729_m1; RUNX2, Hs00231692_m1; COL1A1, Hs00164004_m1; SPARC, Hs00234160_m1; SPP1, Hs00167093_m1; B2M, Hs999999_m1]) were obtained from Applied Biosystems. The real-time amplification conditions were as follows: 10 minutes at 95°C (AmpliTaq Gold activation), followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. Thermocycling and signal detection were performed with an ABI Prism 7300 Sequence Detector (Applied Biosystems). Signals were detected according to the manufacturer's instructions. The technique allows identification of the cycling point where PCR product is detectable, by means of fluorescence emission (Ct value). As previously reported (20), the Ct value correlates with the starting quantity of target mRNA (20); we selected the ΔRn in the exponential phase of amplification plots to determine the Ct values and to obtain the linearity of calibration curves.

PCR efficiencies were calculated with a relative standard curve derived from a 4-cDNA dilution series (from peripheral blood cells for cluster differentiation expression and from osteoblastic-like cells [MG-63] for osteogenic differentiation expression) tested in triplicate, yielding regression coefficients of >0.98 and efficiencies of >96%. To normalize gene expression, we amplified the housekeeping gene B2M.

The amount of cDNA in each sample, calculated as the average from triplicate experiments, was determined from the standard curves. The results are expressed as percentages for expression of cluster differentiation marker genes and as pg equivalent to osteogenic RNA obtained from MG-63 cells.

Measurement of OPG and RANKL.

Supernatants were collected from osteoblast cell cultures after 20 days of osteogenic differentiation and stored at −80°C until they were analyzed for quantitative detection of OPG and RANKL. The amounts of OPG and RANKL in conditioned media were determined using 2 different enzyme-linked immunosorbent assays (ELISAs) performed on 96-well immunoplates, according to the procedures recommended by the manufacturers (for OPG, Human Osteoprotegerin Instant ELISA [BMS2021INST]; Bender MedSystems, Milan, Italy and for RANKL, Human sRANKL ELISA Kit [K0331187]; Koma Biotech, Seoul, South Korea). Briefly, 50 μl of each sample was added in duplicate to precoated wells and incubated for 3 hours. After 3 washes, the plates were incubated with tetramethylbenzidine substrate solution for 20–30 minutes. The reaction was stopped and absorbance was determined at 450 nm, with correction at 540 nm, using a microplate autoreader. The sensitivity range of the ELISAs for both OPG and RANKL was 31.2–2,000 pg/ml.

Statistical analysis.

Results were expressed as the mean ± SD. The significance of differences was assessed by Student's t-test for independent samples. In all analyses, P values less than 0.05 were considered significant. Statistical analyses were performed using SPSS for Windows, version 16.0 (SPSS, Chicago, IL).


Densitometric, radiologic, and biochemical evaluation.

Densitometric results are shown in Table 1. Bone densitometry values in osteoporosis patients were lower than those in normal donors at all skeletal sites evaluated. Morphometric analysis of the dorsolumbar spine of osteoporosis patients revealed 1 mild vertebral fracture (SDI 1) in 5 patients, 2 mild vertebral fractures (SDI 2) in 1 patient, and 3 vertebral fractures (2 mild and 1 medium [SDI 4]) in 1 patient. None of the patients included in the study reported any nonvertebral fractures. Levels of all biochemical parameters tested in the osteoporosis patients were within the normal range.

Table 1. Clinical and densitometric data on the osteoporosis patients and normal donors*
 Osteoporosis patients (n = 31)Normal donors (n = 20)
  • *

    Except where indicated otherwise, values are the mean ± SD. BMI = body mass index; BMD = bone mineral density.

  • P < 0.0001 versus normal donors.

Age, years58.93 ± 5.5958.29 ± 8.25
Height, cm157.24 ± 6.07158.72 ± 6.83
Weight, kg60.89 ± 8.7462.75 ± 8.08
BMI, kg/m224.62 ± 3.3325.02 ± 3.96
Lumbar spine  
 BMD, gm/cm20.72 ± 0.050.96 ± 0.14
 T score−2.97 ± 0.46−1.00 ± 1.08
 Z score−1.62 ± 0.540.21 ± 1.27
Femoral neck  
 BMD, gm/cm20.63 ± 0.070.74 ± 0.09
 T score−1.90 ± 0.68−1.07 ± 0.63
 Z score0.68 ± 0.730.02 ± 0.71
Total hip  
 BMD, gm/cm20.74 ± 0.080.85 ± 0.07
 T score−1.63 ± 0.69−0.81 ± 0.56
 Z score−0.75 ± 0.720.07 ± 0.65
No. with vertebral fractures70
No. with nonvertebral fractures00

Cell phenotype and CFU-F assay findings.

The isolated MSCs from osteoporosis patients and normal donors expressed cluster differentiation markers in similar percentages. In particular, the selected cells were CD3−,CD19−,CD34low; the mean ± SD percentage of CD14+ cells in osteoporosis patients and normal donors was 0.51 ± 0.08% and 0.48 ± 0.06%, respectively, and that of CD45+ cells was 2.26 ± 0.32% and 2.30 ± 0.47%, respectively (P not significant for both).

The selected cells from both normal donors and osteoporosis patients formed fibroblastic colonies after 14 days of cultivation (Figure 1). The colonies grew essentially in isolation and represented the clonal expansion of a single CFU-F. When the frequency of CFU-F was analyzed, we observed a significantly higher number of fibroblastic colonies formed by selected cells in osteoporosis patients (mean ± SD 7.02 ± 4.37) compared with normal donors (2.52 ± 1.69) (P < 0.0001) (Figure 2).

Figure 1.

A and B, Selected cells from normal donors (A) and osteoporosis patients (B) after 14 days of cultivation. The colonies grew essentially in isolation and represented the clonal expansion of a single colony-forming unit–fibroblastic. C and D, Alkaline phosphatase staining in cultured cells from normal donors (C) and osteoporosis patients (D) in the early phase (3 days) and late phase (15 days) of osteogenic differentiation (original magnification × 20).

Figure 2.

Lumbar spine bone mineral density (BMD) and the number of colony-forming units–fibroblastic (CFU-F) in osteoporosis patients (solid bars) and normal donors (hatched bars). As expected, BMD was significantly lower in the osteoporosis patients compared with normal donors; in contrast, the number of mesenchymal cells as evaluated by CFU-F assay was higher in osteoporosis patients. Values are the mean and SEM. ∗ = P < 0.0001 versus normal donors.

Transcription factors and osteogenic gene expression.

The monolayer peripheral blood MSC-like cells were incubated with a mixture containing dexamethasone, β-glycerophosphate, and ascorbic acid to obtain osteoblastic cells. In order to analyze the differentiation at the molecular level, we quantified transcription factors and osteogenic gene expression after 3, 8, and 15 days of osteogenic induction.

During the course of osteogenic induction, levels of the bone-related genes COL1A1, SPP1, SPARC, and the transcription factor gene RUNX2 increased, whereas levels of Sp7 decreased, in osteoblasts derived both from osteoporosis patients and from normal donors (Figure 3). However, expression levels of all genes investigated were different in osteoporosis patient osteoblasts versus osteoblasts from normal donors. In particular, down-regulated expression of RUNX2 and Sp7 in osteoporosis patient osteoblasts compared with osteoblasts from normal donors was observed after only 3 days of osteogenic induction (in osteoblasts from osteoporosis patients versus normal donors, respectively, mean ± SD RUNX2 level 0.76 ± 0.028 pg versus 0.89 ± 0.013 pg, Sp7 level 0.0045 ± 0.002 pg versus 0.056 ± 0.001 pg [both P < 0.0001]) (Figure 3). Expression of the osteogenic genes was also reduced in osteoporosis patient osteoblasts compared with osteoblasts from normal donors after 3 days of differentiation (COL1A1 0.0926 ± 0.0035 pg versus 0.114 ± 0.004 pg, SPP1 2.247 ± 0.099 pg versus 2.780 ± 0.116 pg, SPARC 1.239 ± 0.084 pg versus 1.539 ± 0.122 pg [all P < 0.0001]) (Figure 3). Analysis of transcription factors and osteogenic gene expression after 8 days and 15 days of osteogenic induction showed that all genes investigated remained significantly down-regulated in osteoporosis patient osteoblasts compared with osteoblasts from normal donors (Figure 3).

Figure 3.

Expression of mRNA for RUNX2, Sp7, SPP1, SPARC, and COL1A1 after 3, 8, or 15 days of osteogenic induction of mesenchymal cells from osteoporosis patients (solid bars) and normal donors (hatched bars), as determined by real-time polymerase chain reaction with correction for B2M levels. Values are the mean and SEM. ∗ = P < 0.0001 versus normal donors.

The expression of osteogenic differentiation genes correlated significantly with bone density parameters. The relationship between lumbar spine bone mineral density and RUNX2 expression (r2 = 0.455, P < 0.0001) is depicted in Figure 4.

Figure 4.

Left, RUNX2 levels (pg) according to presence or absence of osteoporosis of the lumbar spine. Data are presented as box plots, where the boxes represent the 25th to 75th percentiles, the lines within the boxes represent the median, and the lines outside the boxes represent the 10th and 90th percentiles. Circles indicate outliers. The difference between levels of RUNX-2 in osteoblasts from patients with osteoporosis (OPOb) and osteoblasts from normal donors (NDOb) was statistically significant (∗ = P < 0.0001). Right, Direct correlation between lumbar spine bone mineral density (BMD; gm/cm2) and levels of RUNX2. The correlation was statistically significant (P < 0.0001).


To determine the difference between the 2 groups in terms of osteoclastic and osteoblastic activation, OPG and RANKL levels and the OPG:RANKL ratio were analyzed in supernatants of osteoporosis patient osteoblasts and osteoblasts from normal donors after 20 days of osteogenic differentiation, when expression of mRNA for osteogenic markers, such us SPP1, SPARC, and COL1A1, had increased. Levels of OPG were lower in osteoporosis patient osteoblasts compared with osteoblasts from normal donors (mean ± SD 44.05 ± 7.26 pg/ml versus 49.09 ± 5.92 pg/ml; P = 0.01), whereas RANKL levels were increased in osteoporosis patient osteoblasts (50.06 ± 13.26 pg/ml versus 40.05 ± 9.55 pg/ml; P < 0.005). Accordingly, the OPG:RANKL ratio was lower in osteoporosis patient osteoblasts compared with that in osteoblasts from normal donors (0.94 ± 0.33 versus 1.29 ± 0.33; P < 0.0001) (Figure 5). A similar result was also observed in the early phase of differentiation, but the difference did not reach statistical significance.

Figure 5.

Left, Osteoprotegerin (OPG):RANKL ratio in osteoporosis patients (solid bars) and normal donors (hatched bars). Values are the mean and SEM. ∗ = P < 0.0001 versus normal donors.Right, Direct correlation between lumbar spine bone mineral density (BMD) and OPG:RANKL ratio. The correlation was statistically significant (P < 0.0001).

Interestingly, we observed a significant correlation between bone density and the OPG:RANKL ratio (P = 0.0001). This finding confirmed the important role of OPG:RANKL in conditioning bone mass.


Osteoporosis is characterized by reduced bone mass associated with microarchitectural alterations leading to reduced bone strength and increased risk of fracture (21). As noted above, diagnosis of osteoporosis has been based on the evaluation of bone density, defined as a reduction of T score to below –2.5 (18). However, the relationship between bone density and risk of fracture is not linear, and the mechanisms leading to bone loss are multifactorial. Impairment of bone cell activity and disequilibrium between osteoblast and osteoclast performance are the main factors inducing bone loss. While the role of increased activity of osteoclasts has been well analyzed, the involvement of osteoblasts and osteocytes has not been completely elucidated. Several studies have highlighted the hypothesis that decreased activity of osteoblasts and osteocytes contributes to the pathogenesis of bone loss (22, 23).

In a previous study, we demonstrated the possibility of obtaining mesenchymal stem cells from peripheral blood by a 2-step method; the number of cells obtained with this approach is adequate to induce osteoblastic differentiation and for performing gene expression analysis starting from a simple blood sample (6). On the basis of that finding, we used this approach to examine, in a controlled in vitro setting, the hypothesis that MSCs and relative osteoblastic differentiation in patients with osteoporosis may differ from those in normal donors. In order to address this question, we analyzed the frequency of CFU-F, the expression of osteogenic genes, and the OPG:RANKL ratio from MSCs isolated from the peripheral blood of osteoporosis patients and normal donors.

An important finding of the present study was that levels of peripheral blood MSCs evaluated by CFU-F assay in vitro were significantly higher in osteoporosis patient osteoblasts than in osteoblasts from normal donors. These data may reflect a major mobilization of MSCs in osteoporosis, due to an increase of MSCs for the bone remodeling process. In addition, we observed a significant correlation between bone density and osteoblastic gene expression, suggesting a direct relationship between bone mass and osteoblast differentiation and confirming the involvement of osteoblast-lineage cells in the mechanism of bone loss. In spite of these findings, however, the levels of expression of genes involved in osteogenic differentiation were significantly lower in cultured osteoblasts from patients with osteoporosis. We also found that the increase of progenitor cells due to bone injury was inversely correlated with the level of transcription factor and osteogenic gene expression. These results support the hypothesis that in osteoporosis there is increased recruitment of osteoblastic precursors as a result of increased bone turnover, but these cells are not able to differentiate into mature osteoblasts because of an altered gene expression profile. This defect can contribute to the pathogenesis of osteoporosis, indicating that it could result both from increased osteoclast activity and from a concomitant reduction in differentiation of osteoblasts from precursors due to alteration of gene expression.

Eghbali-Fatourechi et al demonstrated that concentrations of osteocalcin-positive cells in peripheral blood increase with age in men (24). In addition, those authors showed that the percentage of circulating osteocalcin-positive cells is higher during the adolescent growth spurt, which may be associated with stimulation of bone formation, than in adulthood (5). Therefore, the osteoblast-lineage cells may represent a circulatory component of the bone formation process. These data are in accordance with the higher CFU-F numbers found in osteoporosis patients versus normal donors in the present study; nevertheless, we also observed altered osteoblastic differentiation in precursor cells from osteoporosis patients. In addition, we have recently demonstrated increased adipogenic differentiation of MSCs in osteoporosis patients (Valenti MT, et al: unpublished observations); this finding could suggest that impaired osteoblastic differentiation may be associated with adipogenic differentiation at the expense of osteoblasts. Moreover, the possibility of a shift from osteogenic to adipogenic differentiation has been confirmed by histomorphometric analyses showing adipose replacement of the bone marrow in osteoporosis (25), suggesting increased MSC adipogenic differentiation in osteoporosis. Rodriguez et al have also reported that increased bone marrow adipocyte production is counterbalanced by diminished production of osteogenic cells in osteoporosis (26).

It is well known that transcription genes are important in osteoblast differentiation and activity. Several studies have shown that these genes play an important role in osteoblastogenesis. RUNX-2, a member of the runt family of transcription factors that regulate osteoblast differentiation, has a crucial role in the early determination stage of osteoblast-lineage cells (27), whereas Osx acts downstream of RUNX-2/Cbfa1 (11). Osx has been shown to function more specifically than RUNX-2 in regulating osteoblast differentiation, since Osx-knockout mice lack bony skeletons while the cartilage anlagen are fully formed (11).

Type I collagen, a heterotrimeric molecule composed of two α1 chains and one α2 chain, is the major fibrillar component in most connective tissue including mineralized tissue (28), and the gene encoding the α1 chain of type I collagen (COL1A1) is an important functional candidate for the pathogenesis of osteoporosis. During differentiation, osteoblasts secrete osteopontin, a phosphorylated glycoprotein. Owing to its overall acidity, it binds to calcified matrices, and it has been proposed to link organic and inorganic phases to provide tissue adhesion (29). Also, SPARC is an important matricellular protein. The capacity of SPARC to bind to several resident proteins of the extracellular matrix, to modulate growth factor efficacy, to affect the expression of matrix metalloproteinases, and to alter cell shape as a counteradhesive factor supports the notion that SPARC acts to regulate cell interaction with the extracellular milieu during development and in response to injury (30). In our in vitro model, the finding that expression levels of these osteoblastic genes were significantly reduced in osteoblasts from osteoporosis patients can provide new insight into the altered molecular pathway in osteoporotic bone; furthermore, our approach may contribute to the understanding of cellular and molecular mechanisms involved in fragility fractures associated with osteoporosis.

The analysis of the OPG:RANKL ratio in osteoblastic cultures of peripheral blood MSCs enabled us to eliminate the potential effect of different growth factors and cytokines potentially influencing the system. We found a significant direct correlation between bone density and the OPG:RANKL ratio, with significantly lower ratios in osteoporosis patient osteoblasts than in osteoblasts from normal donors. This result confirms the direct relationship between the OPG/RANK/RANKL system and bone mass, suggesting that a decrease in the OPG:RANKL ratio may be one of the mechanisms leading to bone loss. It is known that OPG and RANKL are expressed in osteoblastic cells and that both are involved in the osteoclast regulatory pathway and, consequently, in bone remodeling.

RANKL activates its receptor, RANK, which is expressed on osteoclasts and their precursors, promotes osteoclast formation and activation, and, by suppressing apoptosis, prolongs osteoclast survival (31). Moreover, evidence that OPG acts as an inhibitor of osteoclastogenesis has emerged from experiments with transgenic mice, in which overexpression of OPG led to severe osteopetrosis and reduced the number of mature osteoclasts. In contrast, mice lacking the gene for OPG are osteoporotic (32). The biologic effects of OPG on bone cells include inhibition of the terminal stages of osteoclast differentiation and suppression of mature osteoclast activation (17).

Therefore, the lower OPG:RANKL ratio observed in osteoblasts from patients with osteoporosis may reflect a reduced ability of osteoblasts to prevent osteoclastic activation in osteoporosis, even without an influence of external factors. The OPG:RANKL ratio has been analyzed in animal cultures, tumor cells, and cell lines, but this is the first reported study comparing findings in in vitro–evaluated osteoblasts from osteoporosis patients and normal donors. Although larger studies are needed to confirm the present findings, our results support the hypothesis that impairment of osteoblast differentiation can be an important mechanism in the pathogenesis of bone loss.

From a clinical standpoint, these findings confirm the usefulness of this approach for diagnosis of bone pathology, as we have previously suggested (6). These intriguing results suggest a noninvasive method for studying the bone remodeling process. A noninvasive approach for evaluating osteoblastic differentiation from mesenchymal stem cells offers the possibility of comparing patients with suspected osteoporosis or other bone diseases and normal donors, significantly improving the reliability of the findings observed.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Dalle Carbonare had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Dalle Carbonare, Valenti.

Acquisition of data. Zanatta, Donatelli.

Analysis and interpretation of data. Dalle Carbonare, Valenti, Lo Cascio.


The grant from IFB Stroder was awarded as the Stroder-SIOMMMS first prize for the best scientific project presented during the VI Congress of the Società Italiana dell'Osteoporosi, del Metabolismo Minerale e delle Malattie dello Scheletro. IFB Stroder was not involved in the study design, data collection, or analysis and interpretation of the data, and had no influence on the publishing of the data.