Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA
Skeletal Health and Osteoporosis Center and Bone Density Unit, Division of Endocrinology, Diabetes and Hypertension, Professor of Medicine, Harvard Medical School, Brigham and Women's Hospital, 221 Longwood Avenue, Boston, MA 02115, USA.
Clinical risk factors, low bone density, and an imbalance in bone remodeling contribute to increased fracture risk.1–3 Skeletal homeostasis is achieved through tissue remodeling by bone-resorbing osteoclasts and bone-forming osteoblasts. Many factors contribute to an imbalance of bone resorption and formation resulting in bone loss, including aging, low body mass index (BMI), and medical disorders such as hyperparathyroidism or renal insufficiency. Both aging and vitamin D deficiency are associated with reduced bone density, osteoporosis, and increased risk of fragility fractures. We reported extreme vitamin D deficiency in community-dwelling U.S. women presenting with hip fracture.4 Whereas chronic vitamin D-deficiency causes impaired mineralization and osteomalacia, it may also contribute to reductions in bone formation and strength. Elucidation of the relative importance of clinical risk factors on bone formation in humans may lead to improved preventive and therapeutic strategies for osteoporosis.5
Human marrow-derived stromal cells (hMSCs), also known as mesenchymal stem cells, are precursors of several mesenchymal cellular lineages, including osteoblasts.6–8 We9–12 and others13, 14 showed an age-dependent decline in osteoblast potential of hMSCs that may contribute to human skeletal aging. We found that other in vitro properties of hMSCs vary with the age of the subjects from whom the cells were obtained, including proliferation potential,10 production of cytokines,15, 16 expression of WNT genes,17 expression of the parathyroid hormone (PTH) receptor, and PTH signaling and osteoanabolic effects.11
It is known that 1,25-dihydroxyvitamin D (1,25(OH)2D) stimulates the in vitro differentiation of hMSCs to osteoblasts.18 Finding that osteoblast differentiation was also stimulated by 25-hydroxyvitamin D3 (25OHD3) led to the discoveries that hMSCs have the capacity to enzymatically activate 25OHD3 to 1,25(OH)2D3 with CYP27B1/1α-hydroxylase,19 and that CYP27B1 is necessary for 25OHD3's antiproliferative and prodifferentiation actions in hMSCs.20 The constitutive level of expression of CYP27B1 in hMSCs in vitro was related to the vitamin D status19 and age12 of the subjects from whom these cells were obtained. Less is known, however, about the effect of age, BMI, adiposity, renal function, or other clinical characteristics on in vitro differentiation of osteoblasts. Given the importance of these clinical risk factors, and recent debates about the level of 25OHD optimal for bone health, translational studies that bridge in vivo clinical attributes with in vitro regulation of osteoblast formation provide a unique approach to identify factors that contribute to reduced bone mass in humans. In this study, we investigated the effects of age, serum 25OHD, 1,25(OH)2D, PTH, estimated glomerular filtration rate (eGFR), BMI, and new standardized indices of fat and lean mass (fat mass index [FMI; defined as fat mass/height2]; lean mass index [LMI] (defined as lean mass/height2]) on in vitro hMSCs responsiveness to 1,25(OH)2D3.
Materials and Methods
Subjects and clinical characteristics
We obtained bone marrow samples from discarded femoral tissue collected during primary arthroplasty for osteoarthritis as described,19 through an institutional review board (IRB)-approved study. Subjects were excluded if they were taking medications or had comorbid conditions that could affect skeletal metabolism, including rheumatoid arthritis. A total of 53 subjects (aged 41–83 years, 21 men and 32 women) scheduled for hip arthroplasty were enrolled in this study; some data were not available for different subjects. Bone mineral density (BMD) of the spine (L1–L4) and proximal femur, and body composition were measured by dual X-ray absorptiometry (DXA) (Discovery H; Hologic Inc., Bedford, MA, USA) in the Skeletal Health and Osteoporosis Center.19 Body composition values were analyzed with APEX Software Version 3.3 (Lawrenceville, GA, USA) that allows calculation of fat and lean mass indices, FMI and LMI.21 FMI values were characterized according to new gender- and age-specific thresholds from the National Health and Nutrition Examination Survey (NHANES) database. Thresholds for individuals categorized as overweight (BMI > 25 kg/m2) are set at FMI >6 kg/m2 for males and >9 kg/m2 for females, and thresholds for obesity (BMI > 30 kg/m2) are >9 kg/m2 for males and >13 kg/m2 for females.22 Percent coefficient of variation (CV%) for fat and lean tissue measures in the Bone Density Unit were 1.09% ± 0.15% and 0.89% ± 0.28%.23 Blood chemistry tests, including measurements of serum 25OHD, 1,25(OH)2D, and PTH, and complete blood counts, were performed in hospital clinical laboratories or the Harvard Catalyst Core Laboratory as recently described.19 eGFR was estimated according to the Modification of Diet in Renal Disease (MDRD) Study equation [GFR (mL/min/1.73 m2) = 175 × (Scr)−1.154 × (Age)−0.203 × (0.742 if female) × (1.212 if African American) (conventional units)]. An additional set of bone marrow samples that were used for osteoblast differentiation experiments was obtained as discarded tissue from 13 deidentified individuals with IRB approval and the same preoperative exclusion criteria were used.
Preparation of hMSCs
We isolated low-density marrow mononuclear cells by centrifugation on Ficoll/Histopaque 1077 (Sigma, St. Louis, MO, USA).24 This procedure removes differentiated cells and enriches for undifferentiated, low-density marrow mononuclear cells that include a fraction of nonadherent hematopoietic cells and a fraction capable of adherence and differentiation into musculoskeletal cells. Adherent hMSCs were expanded at 37°C with 5% CO2 in monolayer culture with phenol red-free α modified essential medium (α-MEM medium, 10% fetal bovine serum–heat inactivated [FBS-HI], 100 U/mL penicillin, and 100 µg/mL streptomycin; Invitrogen, Carlsbad, CA, USA) as described.10 In some cases, MSCs were expanded in Dulbecco's Modified Eagle Medium (DMEM), 10% FBS (Atlanta Biologicals, Norcross, GA, USA), 100 U/mL penicillin, 100 µg/mL streptomycin, and 292 µg/mL L-glutamine (Irvine Scientific).
Alkaline Phosphatase Enzyme Assay
We cultured cells in triplicate in 12-well-plates in α-MEM with 10% FBS-HI until confluence; this required different times depending upon rates of proliferation. Upon confluence, medium was changed to osteogenic medium (α-MEM with 1% FBS-HI, 100 U/mL penicillin, 100 µg/mL streptomycin plus 10 nM dexamethasone (Dex), 5 mM β-glycerophosphate, 50 µg/mL ascorbate-2-phosphate) for 7 or 14 days. Some were supplemented with 10 nM 1,25(OH)2D3 (Sigma). Alkaline phosphatase (ALP) enzyme activity was measured after 7 or 14 days of treatment; other cells were stained at day 14 with an ALP kit (86R; Sigma)25 or assayed at day 7 for ALP enzyme activity; shown as µmol/min/g protein, as described.24 Effects of 1,25(OH)2D3 are presented as relative to control.
We also assessed another set of MSCs from 13 men for osteoblast differentiation. The medium was changed to an osteoblastogenic media containing DMEM, 10% FBS (Atlanta Biologicals), 5 mM β-glycerophosphate (Sigma), and 170 µM ascorbic phosphate (WAKO, Osaka, Japan). Some were supplemented with 10 nM Dex and/or 10 nM 1,25(OH)2D3 (Sigma). After 6 days, ALP activity was measured colorimetrically.
Matrix mineralization assay
For each sample assayed for mineralization, we seeded hMSCs in triplicate in 12-well plates or in quadruplicate in 24-well plates in α-MEM with 10% FBS-HI. Upon confluence, the medium was replaced with osteogenic medium with or without 10 nM of 1,25(OH)2D3. Mineralization was assayed after 21 days with Alizarin red S staining; quantitative data are presented as nanomoles of Alizarin Red S per well.11
RNA isolation and RT-PCR
We isolated total RNA from hMSCs with Trizol reagent (Invitrogen). For RT-PCR, 2 µg of total RNA was reverse-transcribed into cDNA with MMLV (Promega, WI, USA), following the manufacturer's instructions. One-twentieth of the cDNA was used in each 50-µL PCR reaction (30–40 cycles at 94°C for 1 minute, 55–60°C for 1 minute, and 72°C for 2 minutes) as described.24 The gene-specific primers for human osteocalcin,26IGF-I,27 and IGF-BP328 were used for amplification. Concentration of cDNA and amplification conditions were optimized to reflect the exponential phase of amplification. Data were expressed by normalization of densitometric units to that for GAPDH (internal control), as described.10, 24
All experiments were performed at least in triplicate. Group data are presented as mean values ± SD. Unless otherwise indicated, quantitative data were analyzed with nonparametric tools, either the Mann-Whitney test for group comparisons or the Spearman correlation test. If data allowed, parametric tools were used, either t test for two-group or one-way ANOVA for multiple group comparisons or Pearson correlation test. A value of p < 0.05 was considered significant.
Clinical attributes of the study subjects
There were 53 consenting subjects from whom MSCs were isolated from bone marrow discarded during orthopedic surgery (Table 1). The mean age was 65 ± 10 years, ranging from 41 to 83 years. There were 21 men (65.0 ± 10.6 years) and 32 women (65.7 ± 10.3 years). There were wide ranges of values for serum 25OHD, 1,25(OH)2D, PTH, eGFR, BMI, FMI, LMI, and BMD of whole body (WB), femoral neck (FN), or femoral total (F-Tot), and BMD T-scores (Table 1). We found that 15% of the subjects had low 25OHD levels and were classified as vitamin D–deficient (≤20 ng/mL serum 25OHD), that 45% were “insufficient,” with levels from 21 to 29 ng/mL, and that 40% were vitamin D–sufficient (≥30 ng/mL).29
Table 1. Characteristics of the Study Subjects
Mean ± SD
PTH = parathyroid hormone; eGFR = estimated glomerular filtration rate; BMD = bone mineral density; FN = femoral neck; WB = whole body; BMI = body mass index; FMI = fat mass index; LMI = lean mass index.
65 ± 10
25OHD (ng/mL) (normal, 20–57)
30.1 ± 11.0
1,25(OH)2D (pg/mL) (normal, 18–62)
42.3 ± 13.0
PTH (pg/mL) (normal, 10–65)
36.4 ± 15.0
eGFR (mL/min/1.73 m2) (normal, >60)
83.7 ± 18.0
BMD (FN) (g/cm2)
0.82 ± 0.17
BMD (FN) T score
−0.64 ± 0.17
−3.1 to 3.8
BMD (WB) (g/cm2)
1.2 ± 0.2
BMD (WB) T score
1.0 ± 1.7
−1.4 to 7.5
28.8 ± 7.2
11.4 ± 5.4
15.6 ± 2.3
Relationships for age and vitamin D status with clinical attributes of the study subjects
There was an inverse correlation between T-score for WB BMD and age (r = −0.32, p = 0.024, n = 49; Fig. 1A), but there was only a trend for an effect of age on WB BMD (gm/cm2; r = −0.25, p = 0.077, n = 49; Table 2). There were inverse relationships with age for BMD (r = −0.32, p = 0.037, n = 42; Table 2) and T-score at the FN (r = −0.44, p = 0.0038, n = 42; Fig. 1B) and a trend for an inverse correlation between total hip BMD and age (r = −0.28, p = 0.084, n = 39). There was an inverse correlation between eGFR—an index of renal function—and age (r = −0.36, p = 0.032, n = 36; Fig. 1C). There were no effects of age (Table 2) or gender on serum 25OHD. There were significant inverse correlations with age for BMI (Pearson r = −0.37, p = 0.0065, n = 53; Fig. 1D), FMI (r = −0.43, p = 0.0018, n = 51; Fig. 1E), and LMI (r = −0.35, p = 0.011, n = 51; Fig. 1F).
Table 2. Correlations for Age, Serum 25OHD, and ALP With Other Clinical Parameters
BMD (FN) T-score
BMD (WB) T-score
ALP = alkaline phosphatase activity of hMSCs in vitro; hMSC = human marrow stromal cells; PTH = parathyroid hormone; eGFR = estimated glomerular filtration rate; BMD = bone mineral density; FN = femoral neck; WB = whole body; BMI = body mass index; FMI = fat mass index; LMI = lean mass index; r = Pearson correlation coefficient; p = two-tailed p value; n = number of subjects.
As expected, serum 25OHD was inversely correlated with serum PTH (r = −0.29, p = 0.05, n = 47); there were no correlations for 25OHD with BMD or serum 1,25(OH)2D for this cohort (Table 2). In addition, there were significant inverse correlations with serum 25OHD for BMI (r = −0.41, p = 0.0029, n = 50), FMI (r = −0.30, p = 0.033, n = 50), and LMI (r = −0.32, p = 0.023, n = 50) (Table 2, Fig. 2).
There were significant correlations for WB BMD with BMI (Pearson, r = 0.34, p = 0.020, n = 46) and LMI (r = 0.60, p < 0.0001, n = 45), but not with FMI (r = 0.15, p = 0.33, n = 45) (Fig. 3); this suggests that the association between BMI and BMD is mainly due to LMI. Although FMI was not correlated with BMD, applying the new gender-based thresholds for FMI22 showed that the mean BMD of obese subjects (n = 23) (FMI > 9 for men, >13 for women) was significantly higher than for overweight subjects (n = 13) (FMI >6 to 9 for men, >9 to 13 for women) (p < 0.05, ANOVA), and for normal weight subjects (n = 11) (FMI from 3 to 6 for men, from 5 to 9 for women) (p < 0.05, ANOVA; Fig. 3D).
In vitro effects of 1,25(OH)2D3 on osteoblastogenesis of hMSCs
After the hMSCs reached confluence, the medium was changed to α-MEM with 1% FBS-HI plus osteogenic supplements (10 nM Dex, 5 mM β-glycerophosphate, 50 µg/mL ascorbate-2-phosphate) for assessment of osteoblast differentiation. Representative studies showed that 1,25(OH)2D3 (10 nM) significantly stimulated ALP histochemical staining (Fig. 4A), and that there was a dose-dependent stimulation of ALP enzymatic activity (Fig. 4B). There was an average 160% increase in ALP activity with 10 nM 1,25(OH)2D3 in hMSCs from 53 subjects (Table 3). Further, after 21 days of culture, 1,25(OH)2D3 (10 nM) significantly enhanced matrix accumulation of mineral (Fig. 4C, D). There was an average 260% increase in mineral accumulation with 10 nM 1,25(OH)2D3 in hMSCs from 22 subjects (Table 3).
Table 3. The Effects of 1,25(OH)2D3 on Osteoblastogenesis and Mineralization of hMSCs
hMSC = human marrow stromal cell; IGF = insulin-like growth factor; BP = binding protein.
Values from paired t test.
Alkaline phosphatase (enzyme assay)
158.9 ± 68.9
p < 0.0001, n = 53
Mineralization (Alizarin red)
257.3 ± 433.5
p = 0.011, n = 22
1742.2 ± 1682.9
p < 0.0001, n = 28
137.6 ± 32.0
p < 0.0001, n = 21
175.8 ± 66.2
p < 0.0001, n = 44
Effects of age on 1,25(OH)2D3 stimulation of osteoblastogenesis in hMSCs
A pilot study with a set of samples from 13 men (age range 27–79 years) showed that constitutive ALP activity in basal culture medium was low (Fig. 5). Osteoblastogenic supplements were added separately for assessment of osteoblast differentiation. A representative study showed that Dex (10 nM) ± 1,25(OH)2D3 (10 nM) significantly stimulated ALP activity in hMSCs (Fig. 5A). With this series of samples, Dex stimulated ALP (mean 266%; range, 111% to 615%) and 1,25(OH)2D3 further stimulated ALP activity (mean 457%; range, 85% to 1135%). There was a striking inverse correlation with age for the magnitude of stimulation of ALP enzymatic activity by 1,25(OH)2D3 plus Dex (r = −0.803, p = 0.0009; Fig. 5).
With a larger set of samples from well-characterized, enrolled subjects, there was an inverse correlation between stimulation of ALP enzymatic activity by 1,25(OH)2D3 and subject age with hMSCs obtained from subjects >55 years old (r = −0.46, p = 0.032, n = 46; Pearson correlation) (Fig. 6A). The effects of 1,25(OH)2D3 on osteoblast differentiation were also monitored by changes in expression of osteoblast signature genes. Assessment of effects on expression of early osteoblast signature genes (day 2) showed that 1,25(OH)2D3 significantly upregulated IGF-I (138% ± 32%, p < 0.0001, n = 21) and IGFBP-3 (176% ± 66%, p < 0.0001, n = 44; Table 3). After 14 days in osteogenic media, 1,25(OH)2D3 stimulated osteocalcin (OC) gene expression (1742% ± 1683%, n = 28; Fig. 6B). There was an inverse correlation between magnitude of upregulation of OC by 1,25(OH)2D3 and subject age (r = −0.56, p = 0.0019, n = 28; Fig. 6C). There was 33% greater expression of OC in hMSCs obtained from younger subjects (<65 years old) subjects (25.9 ± 18.4, n = 13) than in hMSCs from older (≥65 years) subjects (7.8 ± 5.3, n = 15, p = 0.0045, t test).
In vivo attributes that influence in vitro stimulation of osteoblastogenesis by 1,25(OH)2D3
To test the hypothesis that clinical parameters of these subjects affected the osteoblastogenic response of hMSCs to 1,25(OH)2D3 in vitro, we performed correlation analyses for the stimulation of 1,25(OH)2D3 on ALP activity with clinical attributes of the individual from whom the hMSCs were obtained (Fig. 7, Table 2). There was significantly greater responsiveness to 1,25(OH)2D3 for hMSCs obtained from subjects who had lower serum 25(OH)D (r = −0.36, p = 0.03, n = 37; Fig. 7A). Osteoblastogenesis was stimulated to a greater degree by 1,25(OH)2D3 in hMSCs that were obtained from subjects with inadequate or deficient 25OHD levels (≤20 ng/mL, n = 5) than the subjects who were vitamin D “insufficient” (21–29 ng/mL, n = 17) or vitamin D sufficient (≥30 ng/mL, n = 15) (p < 0.05, ANOVA; Fig. 7B). The osteoblastogenic response to 1,25(OH)2D3 was greater in hMSCs obtained from subjects with higher serum PTH (r = 0.49, p = 0.0016, n = 38; Fig. 7C) and with higher eGFR (r = 0.37, p = 0.039, n = 31; Fig. 7D). The two other in vivo parameters that were significantly correlated with in vitro responsiveness to 1,25(OH)2D3 were higher WB BMD (r = 0.35, p = 0.038, n = 36) and WB BMD T-scores (r = 0.39, p = 0.018, n = 36; Table 2).
Research related to the differentiation of adult human mesenchymal stem cells is increasing, but often there is no information about the clinical characteristics of the subjects from whom the cells were obtained. We recently observed that primary cultures of hMSCs are influenced by the clinical environment from which the cells were obtained. For example, there were discrepancies in the literature regarding expression of WNT genes in studies using a single sample of hMSCs that led us to discover that age and gender accounted for significant differences in their expression.17 Marrow discarded from well-characterized subjects provides an opportunity to evaluate regulation of differentiation of hMSCs and to define the contributions of extrinsic factors and intrinsic mechanisms of skeletal aging. Extrinsic refers to systemic age-related changes that have an impact on the skeleton, such as menopause and hormone replacement therapy.15, 30, 31 Intrinsic refers to fundamental cellular properties that are altered with age, such as a decrease in in vitro proliferation of hMSCs, and increases in apoptosis and senescence associated–β-galactosidase,10 a marker for in vivo age or in vitro senescence. Using early-passage hMSCs avoids senescent effects of expansion in culture. Advanced subject age was also related to impaired in vitro responsiveness of hMSCs to PTH signaling11 and reduced biosynthesis of 1,25(OH)2D.12 Those inherent age-related properties of hMSCs may contribute to skeletal aging. In this study, we tested whether in vivo clinical attributes influenced the in vitro osteoblast potential of MSCs. We first evaluated the effect of age on serum parameters and body composition indices in a cohort of osteoarthritic subjects for whom surgical marrow was available. Consistent with data for non-osteoarthritic subjects,32, 33 FN BMD and renal function were inversely associated with age. As part of the aging process, a measurable decline in renal function occurs in most people,34 and appears to be a part of the normal physiologic process of cellular and organ senescence associated with structural changes in the kidneys.35 This study indicates that subjects with advanced osteoarthritis are similar to non-osteoarthritic subjects regarding the effects of age on BMD and on eGFR.
Clinical and epidemiological studies indicate the vitamin D is important for bone health, but there is little information about the effects of circulating 25OHD levels on in vitro properties of marrow cells. We reported that hMSCs express vitamin D–hydroxylating enzymes and that serum 25OHD levels were related to levels of expression and activity of CYP27B1/1α-hydroxylase.10 This study was designed to test whether vitamin D status of the subject influenced the in vitro effect of 1,25(OH)2D3 on osteoblast differentiation. The serum 25OHD levels for this cohort ranged from 7.6 to 64.9 ng/mL. As expected,36, 37 serum 25OHD was inversely correlated with serum PTH. Vitamin D deficiency in adults is associated with reduced bone density and an increased risk of fractures.38 There has been considerable debate about what constitutes vitamin D deficiency and sufficiency. Prior studies considered 25OHD levels below 50 nmol/L (20 ng/mL) as deficiency, and sufficiency as greater than 62.5 to 80 nmol/L (25 to 32 ng/mL) or as 75 to 80 nmol/L (30 to 32 ng/mL).32, 39, 40 A recent Institute of Medicine (IOM) report supported a 25OHD level of 50 nmol/L (20 ng/mL) as an adequate vitamin D level for the general population in North America.41, 42 Other organizations including the International Osteoporosis Foundation43 and the Endocrine Society proposed 25OHD levels above 75 nmol/L (30 ng/mL) as sufficient for bone health.29 The circulating 25OHD level may be very important to support non-renal production of 1,25-dihydroxyvitamin D.43
This cohort showed a range of values for BMD and body composition indices. High values for BMD, BMI, and FMI were not unexpected because these subjects presented to orthopedic surgery for advanced hip osteoarthritis. Finding a subset of osteoarthritic subjects with low values for BMD, BMI, and FMI is similar to a previous cohort of osteoarthritic subjects in which 25% had occult osteoporosis.44, 45 It is appreciated that both fat mass and lean mass contribute to the effect of body weight on bone.46 Our data indicate that BMI, FMI, and LMI were inversely associated with age and serum 25OHD, and were positively correlated with WB BMD; the strongest contribution to BMD was LMI, not FMI.
This research identified several clinical attributes that were associated with greater in vitro stimulation of osteoblastogenesis by 1,25(OH)2D3, including age less than 65 years, vitamin D insufficiency, elevated PTH, and normal eGFR. Animal studies indicated that effects of some clinical features are likely to influence MSCs. For example, Noh and coworkers47 demonstrated functional incompetence of murine MSCs from mice made uremic by partial kidney ablation. In the mouse ovariectomy (OVX) model, MSCs obtained from OVX mice had lower constitutive ALP activity, gene expression of Runt-related transcription factor 2 (RUNX2), transforming growth factor β1 (TGF-β1), and bone morphogenic protein (BMP)-2, and proliferation capacity, and more apoptotic cells than cells from sham mice.25 In a study with rats, Huff and coworkers48 reported that there was impaired expansion and multipotentiality of MSCs from rats after chronic alcohol abuse. Our finding that in vitro stimulation of osteoblast differentiation by 1,25(OH)2D3 was greatest in cells from vitamin D–deficient elders may mean that repletion of vitamin D–deficient subjects may lead to more vigorous bone formation. Consistent with this idea, von Hurst and coworkers49 reported that vitamin D suppressed the age-induced increase in bone turnover and reduced bone resorption only in vitamin D–deficient older women.
We found that there were better responses to 1,25(OH)2D3 in hMSCs from subjects with better renal function as indicated by higher eGFR, but there was not a significant correlation between serum 1,25(OH)2D and eGFR, nor between serum 25OHD and eGFR. A limitation of this study is that was not possible to estimate the relative contributions of age, vitamin D status, and renal function on in vitro responsiveness of MSCs to 1,25(OH)2D3, but finding these associations should stimulate further research on the impact of vitamin D status on bone formation. Classically, 1,25(OH)2D3 mediates its actions through activation of the Vitamin D Receptor (VDR), a ligand-dependent transcription factor. There may be additional effects of polymorphisms or epigenetic regulation of VDR in osteoblast progenitors.
These studies indicate greater stimulation of in vitro osteoblast differentiation by 1,25(OH)2D3 in hMSCs from younger subjects (<65 years old), subjects with inadequate circulating 25OHD levels, and those with better renal function. Our results suggest that it is clinically important to correct vitamin D deficiency to enhance bone formation at the cellular level. In vivo–in vitro studies with hMSCs from well-characterized subjects provide an innovative opportunity to evaluate effects of clinical attributes such as age, body composition, vitamin D status, and renal status on regulation of osteoblastogenesis of hMSCs.
All authors state that they have no conflicts of interest.
We greatly appreciate the help from S Anderson, K Johnson, N Glass, L Gao, N Setty, M Tuteja, and C Yu for aspects of these experiments. This study was supported by grants from the National Institutes of Health R01 AG 025015 and R01 AG 028114 (JG); Swiss National Science Foundation Fellowship award 81BE-53101 (SMM); China Scholarship Council (SG); American Federation for Aging Research grant A09052 (SZ); and BWH-BRI Fund (SZ). The research laboratory tests were supported by the General Clinical Research Center, National Institutes of Health grant RR-02635 (ML). The discarded marrow was obtained and studied with approval and annual review from the Partners Human Research Committee.
Author's roles: Study design, data collection, analysis, and interpretation, and manuscript preparation: SZ, JG, and ML; experimental work: SZ, SWK, JH, SG, SMM, LS, and IB; all authors approved the submitted version of this manuscript.