SEARCH

SEARCH BY CITATION

Keywords:

  • Iron;
  • Menopause;
  • Osteoblast Differentiation;
  • Proliferation;
  • Osteoporosis

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Postmenopausal osteoporosis is characterized by an imbalance of bone resorption exceeding bone formation, resulting in a net loss of bone mineral density (BMD). Estrogen deficiency is known to promote bone resorption. However, the causative factors that impair bone formation have not been identified. Women after menopause experience not only estrogen deficiency but also iron accumulation as a result of cessation of menstruation. In this study we investigated whether increased iron plays a role in osteoporosis. By growing primary mouse osteoclast and osteoblast progenitor cells as well as immortalized cell lines in the presence of iron, we found that increased iron had minimal effects on osteoclast cell differentiation. Interestingly, iron, particularly in its inorganic form, and to a lesser extent ferritin and transferrin all suppressed alkaline phosphatase (ALP) activities in osteoblasts. Moreover, iron downregulated mRNA levels of several other osteoblastogenic markers such as Runx2, osterix, osteopontin, and osteocalcin. To further show that this in vitro finding is relevant to the in vivo condition, we demonstrated that iron-accumulated mice with intact ovaries exhibited a significant decrease in BMD. Although iron inhibited preosteoblast cell differentiation, it did enhance preosteoblast cell proliferation, as evidenced by increased cell growth and expression of cell cycle regulator genes such as CDK4, CDK6, cyclin D1, and cyclin D3 and G2/M phase cell population. Taken together, our results suggest that increased iron could be a factor that slows down bone formation in postmenopausal women. © 2011 American Society for Bone and Mineral Research.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Osteoporosis occurs in about 50% of women over 50 years of age, and those who suffer from hip fractures become dependent on others for the activities of daily living.1, 2 The causes of osteoporotic fractures are multifactorial. Among all the known risk factors,3 estrogen deficiency is considered the most important because bone loss accelerates 2 to 3 years after menopause at a rate of 1% to 1.5% annually.4 Overall bone turnover is significantly increased in postmenopausal women, but in an imbalanced fashion, with bone resorption exceeding bone formation.5 High levels of estrogen in young premenopausal women maintain remodeling balance by inhibiting osteoclast function via effects on the receptor activator of NF-κB ligand (RANKL)/RANK/osteoprotegrin system.5, 6 Estrogen also reduces the production of a number of proresorptive cytokines and directly deters osteoclast activity and lifespan.5, 6 Therefore, withdrawal of estrogen or estrogen deficiency promotes bone resorption and increases bone turnover. However, bone resorption increases by 90% at menopause, as assessed by biochemical markers, whereas bone-formation markers increase by only 45%.7, 8 Therefore, an increase in bone turnover rate and a remodeling imbalance lead to accelerated bone loss. The exact mechanism as to why bone formation after menopause cannot keep up with bone resorption remains unknown.

Menopause is a natural aging process during which a woman passes from the reproductive to the nonreproductive years. Using data collected initially from the New York University (NYU) Women's Health Study, we observed that ferritin, an iron storage protein with the capacity to bind up to 4,500 atoms of iron, and transferrin, an iron-transport protein with two binding sites for iron, are significantly higher in post- than in premenopausal women.9 A subsequent literature search of large population studies demonstrated that concurrent but inverse changes occur in levels of ferritin and estrogen during the menopausal transition.10 The normal functions of ovaries and menstruation lead to systemic levels of high estrogen but low iron in young women. As they get older, ovaries cease functioning, and iron is no longer lost through menstruation. Thus systemic levels of high estrogen and low iron in young premenopausal women are reversed to the low-estrogen and high-iron levels seen in older postmenopausal women.10 In view of the profound differences in iron levels between before and after menopause, we have tested a hypothesis that in addition to estrogen deficiency, increased iron as a result of menopause contributes to the development of postmenopausal osteoporosis.10

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Cell culture

Mouse preosteoclast Raw 264.7 cells, preosteoblast C2C12 cells, and MC 3T3-E1 cells were purchased from the American Type Culture Collection (Manassas, VA, USA) and cultured in 5% CO2 at 37°C in Dulbecco's modified Eagle medium (DMEM) and α-minimal essential medium (α-MEM), respectively (Invitrogen, Carlsbad, CA, USA). These media were supplemented with 10% heat-inactivated fetal bovine serum (FBS), 100 IU/mL of penicillin, and 100 IU/mL of streptomycin. Primary mouse nonadherent bone marrow cells were isolated from 8-week-old C57BL/6 female mice (Jackson Laboratories, Bar Harbor, ME, USA) as described previously.11 Primary mouse mesenchymal stem cells (MSCs) with C57BL/6J background were a kind gift of Dr DJ Prockop (Texas A&M Health Science Center, Dallas, TX, USA).12

Cell treatment and osteoclast and osteoblast differentiation

Raw 264.7, C2C12, and MC 3T3-E1 cells were seeded initially at a density of 1 × 104/cm2 on 24-well plates in 2% FBS-containing medium. After 24 hours, these cells were adapted to two cell culture models that mimicked pre- and postmenopausal conditions for 3 to 7 days in the presence of RANKL (100 ng/mL, Raw 264.7 cells), bone morphogenetic protein 2 (BMP-2, 100 ng/mL, C2C12 cells), or induction medium without cytokines and growth factors (50 µg/mL of ascorbic acid and 10 mM β-glycerolphosphate for MC 3T3-E1 cells and 1 nM dexamethasone, 0.5 µM ascorbate 2-phosphate, and 10 mM β-glycerol-phosphate for primary mouse MSCs). The premenopausal model had high 17β-estradiol (E2) and low iron (Fe), and the postmenopausal model had low E2 and high Fe. The concentration of E2 in the premenopausal model was set at 500 pg/mL, equivalent to the premenopausal levels.13 Levels of ferritin in the postmenopausal model were equal to those in medium containing 10% serum (or 20 ng/mL of ferritin), given that the physiologic upper limit of serum ferritin is 200 ng/mL. Transferrin (Tf) at 5 µg/mL was added in its wholly unsaturated form (apo-Tf) or its fully 100% iron saturated form (holo-Tf) to the pre- and postmenopausal models, respectively. This simple system allowed us to initiate the proposed research.

To further confirm the roles of E2 and Fe in altering cell differentiation, inorganic iron (ferric sulfate, 5 to 50 µM), apo-transferrin (apo-Tf, 5 to 20 µg/mL), holo-transferrin (holo-Tf, 5 to 20 µg/mL), ferritin (Ftn, 20 ng/mL), E2 (0.5 to 1 µg/mL), and deferoxamine (DFO, 50 µM) (Sigma-Aldrich, Inc., St Louis, MO, USA) were tested separately. In addition to Raw 264.7 cells, primary mouse bone marrow cells were induced to differentiate to osteoclasts in the presence of macrophage colony-stimulating factors (M-CSF, 50 ng/mL) and RANKL (30 ng/mL) with or without iron (5 and 20 µM). Differentiation was measured by counting the number of tartrate-resistant acid phosphatase–positive (TRAP+) multinucleated cells (MNCs) with two or more nuclei, a marker of preosteoclast differentiation, in three different microscopic view fields. C2C12, MC 3T3-E1, and primary mouse MSC differentiations were determined by histochemical staining of alkaline phosphatase (ALP) with the Leukocyte Alkaline Phosphatase Kit (Sigma-Aldrich). Mineralization in MSCs was determined by staining 10% formalin–fixed cells with alizarin red (Sigma-Aldrich).

Measurements of ALP activity

In parallel, ALP activity was measured colorimetrically using p-nitrophenyl phosphate liquid substrate (MP Biomedicals, Santa Ana, CA, USA). Briefly, cells were lysed in RIPA buffer (Santa Cruz Biotechnologies, Santa Cruz, CA, USA), sonicated on ice, and centrifuged at 15,000g for 3 minutes at 4°C to remove cellular debris. Each lysate (50 µL) was mixed with 150 µL of phosphate substrate in a 96-well plate. After 30 minutes of incubation at 37°C, the reaction was terminated by adding 50 µL of 1 M NaOH. ALP activity was measured as absorbance at 405 nm using a UV-Vis microplate reader (Molecular Devices, Sunnyvale, CA, USA). All samples were quantified in triplicate and further normalized by the concentrations of total cellular proteins (BioRad, Hercules, CA, USA).

Western blot analyses

Then 40 µg of protein was separated on 10% or 12% SDS gel, transferred onto polyvinylidene difluoride membranes, and incubated overnight at 4°C with antibodies against CDK4, CDK6, cyclin D1, cyclin D3, β-tubulin, phosphor-ERK, JNK, p38 MPAK, and their nonphosphorylated counterparts (Cell Signaling Technology, Inc., Danvers, MA, USA), respectively. After probing with secondary antibodies at room temperature for 30 minutes, membranes were visualized by an enhanced chemiluminescence system (Amersham Biosciences, Piscataway, NJ, USA) according to the manufacturer's instructions.

mRNA expression of osteoblast marker genes by quantitative real-time polymerase chain reaction (qRT-PCR)

Total RNA was prepared using the SV total RNA Isolation Kit (Promega, Madison, WI, USA). One microgram of total RNA was used to synthesize first-strand cDNA by Superscript II cDNA Synthesis Kit (Invitrogen). qRT-PCR analyses of gene expression were performed in triplicate with the ABI 7900 SDS real-time PCR system using specific primers (Supplemental Table S1) and SYBR Green PCR Master Mix (BioRad). Relative expression levels of the genes of interest were normalized to the geometric average of three internal control genes (GAPDH, G6PD, and HPRT1). Data were expressed as fold change over the control at the corresponding time points.

Cell proliferation assays

Growth studies were performed using the 3-(4,5-dimethylthiazol-2-yl)-2.5-diphenyl-2H-tetrazolium bromide (MTT) colorimetric assay (Sigma). In some experiments, PD98059 (a specific inhibitor of ERK, 10 µM) or U0126 (a MEK1/2 inhibitor, 20 µM) was added to the medium 1 hour before iron treatment. Briefly, 10 µL of a stock MTT solution (5 mg/mL in PBS) was added to 100 µL of culture medium, and incubations were continued for an additional 4 hours. The medium was aspirated, and 100 µL of DMSO was added. The absorbances at 570 nm were determined using a UV-Vis microplate reader. Data were expressed as a percent of control.

Cell cycle analyses by flow cytometry

Cells were treated with Fe or E2 for 24 hours in α-MEM containing 2% FBS. The cells were fixed in ice-cold 80% ethanol at −20°C overnight. The fixed cells were permeabilized in a buffer containing 100 mM sodium citrate/0.1% Triton X-100 at room temperature for 15 minutes. After treatment with 0.2 mg/mL of RNase A for 10 min, the cells were stained with propidium iodide (50 µg/mL) at 4°C for 1 hour. Samples were detected using fluorescence-activated cell sorting (FACS; Becton Dickinson, Franklin Lakes, NJ, USA) and analyzed with FlowJo software (Tree Star, Ashland, OR, USA). A total of 20,000 cells were collected for each sample.

Animals

Hemochromatosis Fe (HFE) gene knockout mice with 129/SvEv background were a kind gift of Dr Nancy A Andrews (Duke University School of Medicine, Chapel Hill, NC, USA). These mice were shown to have an iron-overload condition.14 After crossing with wild-type (WT) 129/SvEv mice, HFE heterozygote (HFE+/−) mice with intact ovaries were used because these mice had mild iron accumulation (see “Results”), and the intact ovaries excluded estrogen as a confounding factor in iron's role. All mice were cared for according to the Institutional Animal Care and Use Committee-approved protocol.

X-ray and micro–computed tomographic (µCT) scanning

Bone mineral density (BMD) and microarchitecture of the femoral and tibial trabecular bones from 16-month-old WT and HFE+/− mice, with 5 mice per group, were measured using dual-energy X-ray absorptiometry (DXA) and µCT by the Kureha Special Laboratory (Iwaki, Japan; www.kureha-special-labo.com/). Iron and estrogen status in serum was analyzed as described previously.15, 16

Statistical analysis

Statistical evaluations of the data were conducted by Student's t test for paired comparison or by one-way analysis of variance for multiple comparisons, followed by a post hoc Newmann-Keuls test. The results were presented as means ± SD except for the BMD data. p Values < .05 and <.01 were considered to be significantly different.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Inhibitory effects of low E2 and high Fe on preosteoblast differentiation

To investigate whether increased iron levels as a result of menopause affect the balance between bone formation and bone resorption, mouse preosteoclasts Raw 264.7 and preosteoblast C2C12 cells were cultured in the two distinct media. Figure 1 shows that RANKL, M-CSF, and BMP-2 induced osteoclast and osteoblast differentiation, respectively. However, pre- and postmenopausal conditions seemed to have minimal effects on RANKL-mediated preosteoclast Raw 264.7 cell differentiation, as measured by the number of TRAP+ MNCs (Fig. 1A). To confirm that this observation with Raw 264.7 cell line is physiologically relevant, Fig. 1B shows that iron also had little effect on the differentiation of mouse primary bone marrow cells. Interestingly, preosteoblast C2C12 cells grown under postmenopausal conditions of low E2 and high Fe demonstrate that in the presence of BMP-2, significant inhibition was observed on the ALP activity compared with the same cells grown under premenopausal conditions or in the presence of BMP-2 alone (Fig. 1C). No meaningful difference in ALP activity was observed between premenopausal conditions and in the presence of BMP-2 alone.

thumbnail image

Figure 1. Effects of low E2 and high Fe on osteoclast and osteoblast differentiation. Mouse preosteoclast Raw 264.7 cells, primary bone marrow cells, and preosteoblast C2C12 cells were grown under the defined control, pre-, and postmenopausal conditions. (A) Raw 264.7 cells in the presence of RANKL at 100 ng/mL for 5 days. (B) Primary mouse bone marrow cells in the presence of M-CSF (50 ng/mL) and RANKL (30 ng/mL) for 7 days. (C) C2C12 cells in the presence of BMP-2 at 100 ng/mL for 5 days. *Significantly different from the BMP-2-treated cells (p < .05).

Download figure to PowerPoint

Inhibitory effects of iron on ALP activity and mineralization

To illustrate which compound is responsible for the inhibition, separate testing of the constituents in the pre- and postmenopausal conditions was carried out. Iron and ferritin had no significant effects on osteoclast differentiation (Supplemental Fig. S1). Figure 2A shows that ALP activity in C2C12 cells was strongly inhibited by the addition of inorganic iron (ferric sulfate, 20 µM) and to a lesser extent by all iron proteins, such as apo-Tf (5 µg/mL), Ftn (20 ng/mL), and holo-Tf (5 µg/mL). Among them, inorganic Fe had the most noteworthy suppression, totally blocking the BMP-2-induced ALP activity to an extent that was even lower than the control cells in the absence of BMP-2. In contrast, E2 slightly stimulated ALP. Figure 2B shows that the inhibitory effect of iron on BMP-2-induced ALP activity was dose-dependent. Histochemical staining of ALP shows that BMP-2 alone or BMP-2 with the premenopausal condition of high E2 and low Fe or with the iron chelator DFO promoted ALP formation (Supplemental Fig. S2). On the contrary, the postmenopausal condition of low E2 and high Fe or iron sulfate inhibited BMP-2-induced ALP. To further show that the inhibitory effect of iron is true for other types of preosteoblasts in the absence of BMP-2, differentiations of MC 3T3-E1 in a BMP-2-free medium showed similar results, with the most striking inhibitory effects of inorganic Fe on ALP formation. Figure 2C, D demonstrates that iron inhibited not only ALP but also mineralization of primary mouse MSCs.

thumbnail image

Figure 2. Effects of individual components in the model systems on ALP activity and mineralization. (A) C2C12 cells were pretreated with each component for 3 days, followed by exposure to BMP-2 (100 ng/mL) for an additional 3 days. ALP activities were measured and normalized to protein concentrations. E2 (1 µg/mL), apo-Tf (5 µg/mL), Fe3+ (20 µM), Ftn (20 ng/mL), and holo-Tf (5 µg/mL), as well as pre- and postmenopausal conditions. (B) Ferric sulfate at 5, 20, and 50 µM and E2 at 1 µg/mL. (C) ALP staining in mouse MSCs after 21 days. (D) Mineralization in mouse MSCs. *Significantly different from the BMP-2-treated cells (*p < .05; **p < .01).

Download figure to PowerPoint

Inhibitory effects of iron on mRNA expression of osteoblast markers

To assess whether, in addition to ALP, other osteoblast marker genes are altered by iron, mRNA expression levels of Runx2, osteocalcin, osterix, and osteopontin in C2C12 cells were subjected to qRT-PCR analyses. Figure 3A shows that levels of ALP mRNA were increased 38.8- and 26.5-fold after 3 and 5 days of BMP-2 treatment compared with the untreated control. When the cells were pretreated with iron at 5, 20, and 50 µM for 3 days, followed by BMP-2 for an additional 3 days, the ALP mRNA expression was significantly decreased to 12.2-, 5.2-, and 1.8-fold, respectively (Fig. 3A). On day 5, while mRNA levels of ALP remained high in the presence of BMP-2, iron attenuated BMP-2-mediated mRNA levels of ALP (1.2-, 0.8-, and 0.9-fold of the untreated control, respectively). Runx2 also was inhibited by iron in a dose-dependent manner (Fig. 3B). Similarly, osteocalcin, osterix, and osteopontin, which were induced later in the differentiation process, also were inhibited by iron treatments (Fig. 3C–E). It is noteworthy that E2 had relatively small effects on these genes compared with BMP-2 alone.

thumbnail image

Figure 3. Inhibitory effects of iron on mRNA levels of osteoblastogenic genes. C2C12 cells were pretreated with ferric sulfate at 5, 20, and 50 µM or E2 at 1.0 µg/mL for 3 days, followed by exposure to BMP-2 at 100 ng/mL for 3 and 5 days, as indicated. Total RNA was extracted and qRT-PCR was performed as described in “Materials and Methods.” Graphs displayed the fold changes over the untreated controls. *Significantly different from the BMP-2 treatment (p < .05).

Download figure to PowerPoint

Decreased BMD in iron-loaded HFE+/− mice

To further provide in vivo evidence that increased iron leads to osteoporosis development, HFE+/− and age- and genetic background–matched WT control mice at 16 months of age were used. Figure 4A shows that HFE+/− mice had significantly higher levels of serum iron and transferrin saturation than WT mice. Transferrin saturation rates between the two types of mice were similar to those between pre- and postmenopausal women.16, 17 Serum E2 levels were comparable and remained low, simulating menopause. Figure 4B demonstrates that BMD was significantly lower in iron-accumulated HFE+/− mice than in WT mice. Images of the soft X-ray of the femur and µCT illustrate a lower BMD in the iron-loaded HFE+/− mice (Supplemental Fig. S3).

thumbnail image

Figure 4. Decreased BMD in iron-accumulated HFE+/− mice compared with WT mice. (A) Levels of serum iron, transferrin saturation, and E2 in WT and HFE+/− mice (n = 5/group). (B) Average levels of femur BMD ± SEM. *Significantly different from the WT mice (p < .01).

Download figure to PowerPoint

Enhancing effects of iron on C2C12 cell proliferation

In contrast to the inhibitory effects of iron on C2C12 cell differentiation, we found that iron significantly increased C2C12 cell proliferation in a dose- and time-dependent manner (Fig. 5A). As a parallel control, E2 had minimal effects, probably owing to a lack of estrogen receptors in these cells. C2C12 cells treated with holo-Tf also displayed significant cell proliferations compared with cells treated with apo-Tf, further confirming that iron is the component promoting cell proliferation (data not shown). Cell cycle analyses by FACS showed that G2/M phase was significantly increased by iron treatment and, to a lesser extent, by E2 treatment (Fig. 5B). As shown in Fig. 5C, iron increased the protein levels of cyclin-dependent kinase 4 (CDK4), CDK6, cyclin D1, and cyclin D3.

thumbnail image

Figure 5. Enhancing effects of iron on C2C12 proliferation. C2C12 cells were treated with ferric sulfate at 5, 20, and 50 µM and E2 at 1.0 µg/mL for various time periods. Cell proliferation rates were normalized to the cells before treatment, and data on days 1 and 2 are shown (A). Cell cycle analyses were carried out by FACS (B) and Western blot (C) as described in “Materials and Methods.” *Significantly different from the control cells (p < .05).

Download figure to PowerPoint

Possible involvement of ERKs in iron-induced cell proliferation but not differentiation

Our previous studies have shown that iron greatly activates ERK and p38 but not the JNK pathway of the mitogen-activated protein kinase (MAPK) family.18, 19 Figure 6A shows a significant ERK1/2 activation after 1 hour of iron treatment, although there were no significant changes in the activation of the p38 and JNK pathways (data not shown). To check whether ERK signaling plays a role in iron-induced proliferation or differentiation, pretreatments of cells with PD98059, an ERK inhibitor, and U0126, an upstream ERK inhibitor of MEK/1/2, demonstrates that blocking the ERK pathway resulted in significant suppressions of C2C12 cell proliferation (Fig. 6B). However, PD98059 and U0126 did not restore iron-induced ALP inhibition (Fig. 6C).

thumbnail image

Figure 6. Involvement of iron-mediated ERK activation in C2C12 cell proliferation but not differentiation. (A) ERKs activation by iron. C2C12 cells were pretreated with PD98059 for 1 hour before the addition of ferric sulfate and E2 for an additional 3 days. Cells then were collected for MTT and ALP assays as described in the “Materials and Methods.” (B) Attenuation of the ERK inhibitor PD98059 on iron-mediated cell proliferation. (C) PD98059 and U0126 had no effect on iron-mediated ALP inhibition. *Significantly different from the iron-treated cells.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Postmenopausal osteoporosis is characterized by a remodeling imbalance of bone resorption outpacing bone formation. Although it is well known how bone resorption is stimulated by estrogen deficiency,20 factors that impair bone formation have not yet been identified. We have shown previously that in addition to estrogen deficiency, women during menopausal transition experience a two- to threefold iron increase as a result of cessation of menstruation.10 Hence it was logical for us to investigate the role of increased iron as a result of menopause in the bone-remodeling imbalance.

In this study, we began to culture mouse preosteoblast C2C12 and preosteoclast Raw 264.7 cells in the well-defined low-E2 and high-Fe postmenopausal condition or the high-E2 and low-Fe premenopausal condition. This simple but innovative in vitro system allowed us to demonstrate that the postmenopausal condition of low E2 and high Fe inhibited osteoblast progenitor C2C12 cell differentiation but seemed to have minimal effects on preosteoclast Raw 264.7 and primary mouse bone marrow cell differentiations. Separate testing of each individual compound further revealed that inorganic iron and, to a lesser extent, iron proteins such as transferrin and ferritin all were capable of inhibiting ALP activity. Our results support recent reports showing that iron and ferritin suppress osteoblast-specific genes such as ALP and osteocalcin in a dose-dependent manner.21, 22 Although the ferroxidase activity of the ferritin was shown to inhibit calcification,21 our data indicate that iron itself possesses the most pronounced inhibitory effects on ALP as well as mineralization. ALP activity is critical to the initiation of mineralization,23, 24 and inhibitors of ALP were shown to suppress vascular calcification.25 Our results suggest that iron could inhibit bone formation.

BMP-2 is one of the most potent bone-inducing agents.26 BMP-2 treatment increases ALP mRNA expression and ALP activity during osteoblast differentiation.26, 27 To exclude the possibility that iron inactivates BMP-2 and does not have a direct effect on ALP activity in C2C12 cells, MC 3T3-E1 and primary mouse MSCs were used subsequently. Differentiation and mineralization of the MC 3T3-E1 cells and primary mouse MSCs took place in BMP-2-free induction medium. Interestingly, iron alone or a postmenopausal condition of high Fe and low E2 inhibited ALP formation in MC 3T3-E1 cells as well as ALP activity and mineralization in primary mouse MSCs.

To ascertain that the inhibition is not limited to ALP, mRNA levels of other osteoblast specific genes, such as Runx2, osteocalcin, osterix, and osteopontin, were measured. Among many mediators that regulate the BMP-2-induced ALP expression pathway, Runx2 is of particular importance. Runx2 is a key regulatory transcription factor essential to osteoblast differentiation by inducing osteoblastic gene expression.28, 29 BMP-2 stimulates Runx2 mRNA expression in C2C12 cells, and indeed, overexpression of Runx2 in C3H10T1/2 cells stimulates ALP promoter activity.30 Osterix is another prerequisite transcription factor for preosteoblast differentiation,31 and its expression is also induced by BMP treatment in immature mesenchymal cell lines.32 Our results demonstrated that iron pretreatment significantly blocked BMP-2-induced expression of Runx2 and osterix. Osteocalcin is secreted solely by osteoblasts and is implicated in bone mineralization.33 Osteopontin is a well-recognized osteoblast differentiation marker. During osteogenesis, osteopontin is found in bone-forming cells almost concomitantly with the appearance of ALP. Both osteocalcin and osteopontin derived from osteoblasts play a role in osteoclastic differentiation and activation.34 We also found that iron suppressed BMP-2-induced osteopontin and osteocalcin mRNA expression, which is in line with changes observed in mRNA expressions of Runx2 and osterix but not as dramatic as the suppression of ALP mRNA expression.

To further confirm our in vitro finding, HFE gene knockout mice were used. Hemochromatosis Fe patients, who have an iron overload disorder mainly owing to a C282Y mutation of the HFE gene, have been shown to be at high risk of developing osteoporosis.35, 36 In this study, HFE+/− mice had mild iron overload conditions compared with WT mice. Because these HFE mice did not undergo ovariectomies, estrogen did not play a role in the study because differences in estrogen levels were not statistically significant. Interestingly, lower BMD values were observed in HFE+/− mice with mild iron overloads. This finding is highly significant because the inhibitory effects of iron on osteoblasts notably differ from the effects of estrogen on osteoclasts and bone resorption.5, 6 Although iron has been shown to enhance mitochondrial biogenesis and promote osteoclast activity,37, 38 the inhibitory effects of iron on osteoblast differentiation appear to be more striking than those on osteoclasts in this study. It is noteworthy that our study measured only TRAP+ cell formation, which does not reflect the bone-resorbing activity of these cells. Therefore, a cautious interpretation is required for the effects of iron on osteoclast differentiation.

In the context of postmenopausal osteoporosis, it is well accepted that estrogen deficiency caused by menopause increases the lifespan of osteoclasts, bone resorption, and bone turnover rate. Our results suggest that bone formation slows down when iron arises during and after menopause. Since iron also increases in men,10 the inhibitory effects of iron on bone formation may be implicated in both genders, leading to an age-related bone loss. The exact molecular mechanisms as to how increased iron inhibits osteoblast differentiation and bone formation remain to be elucidated. It has been demonstrated that the inhibitor of κB kinase (IKK)-NF-κB pathway is involved in slowing down bone formation.39 Whether this pathway participates in the iron-mediated inhibition of bone formation awaits further investigation.

Iron activated CDK4, CDK6, cyclin D1, and cyclin D3, as well as ERK1/2, in C2C12 cells. These results indicate that iron did not cause cytotoxicity in C2C12 cells and that the observed inhibitory effects of iron on bone-formation markers were not due to cell death. It has been shown that the Ras/MAPK/AP-1 signaling pathway is involved in the regulation of BMP-2 induced osteoblast differentiation in C2C12 cells.40 ERK inhibitors partially reversed an enhancement of ALP activity by E2, suggesting that the MAPK pathway may contribute to E2- and BMP-2-comediated ALP formation. However, ERK inhibitor PD98059 blocked iron-mediated cell proliferation (Fig. 6B) but did not restore iron-mediated ALP inhibition, indicating that iron could keep the osteoblast progenitor cells in a proliferative but undifferentiated state.41

Now the argument is that increased iron as a result of menopause is considered within the normal physiologic range. Is this normal range healthy? Using estrogen as an analogue to iron, estrogen variations over the lifespan of a woman are considered to be within the norm. Yet a long duration of exposure to a normal range of estrogen, including early age at menarche, nulliparity, late first full-time pregnancy, and late menopause, is a well-established risk factor for breast cancer.42 Conversely, estrogen deficiency has been considered the major cause of menopausal symptoms and diseases such as osteoporosis.43 Iron deficiency is a health problem long recognized by the medical community.44 Our study suggests that iron accumulation and a prolonged exposure as a result of menopause could cause adverse health effects in older women. Thus the healthy ranges of estrogen and iron levels in women may need to be reconsidered and then established.

In conclusion, iron could slow down bone formation by inhibiting osteoblast progenitor cell differentiation. Based on the inhibitory effects of iron on bone formation shown in this study and the proven effects of estrogen deficiency on bone resorption, we suggest that it is the combined effects of estrogen deficiency and iron accumulation that cause bone resorption to outpace bone formation and accelerate bone loss in postmenopausal women.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

This research was supported in part by the Applied Research Support Fund from the NYU School of Medicine. We would like to thank Dr NA Andrews for generously providing the HFE gene knockout mice, Dr DJ Prockop for primary mouse mesenchymal stem cells, and Dr Mortimer Levitz for measurements of E2 in mouse sera.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
JBMR_337_sm_SuppFig.ppt583KSupplementary Figures
JBMR_337_sm_SuppTab1.docx14KSupplementary Table 1

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.