Accepted Manuscript

Orthosilicic acid, Si(OH)4, stimulates osteoblast differentiation in vitro by up-regulating miR-146a to antagonize NF-κB activation

Xianfeng Zhou, Fouad M. Moussa, Steven Mankoci, Putu Ustriyana, NianliZhang, Samir Abdelmagid, Jim Molenda, William L. Murphy, Fayez F. Safadi,Nita Sahai

PII: S1742-7061(16)30216-1DOI: http://dx.doi.org/10.1016/j.actbio.2016.05.007Reference: ACTBIO 4239

To appear in: Acta Biomaterialia

Received Date: 29 December 2015Revised Date: 28 April 2016Accepted Date: 5 May 2016

Please cite this article as: Zhou, X., Moussa, F.M., Mankoci, S., Ustriyana, P., Zhang, N., Abdelmagid, S., Molenda,J., Murphy, W.L., Safadi, F.F., Sahai, N., Orthosilicic acid, Si(OH)4, stimulates osteoblast differentiation in vitroby upregulating miR-146a to antagonize NF-κB activation, Acta Biomaterialia (2016), doi: http://dx.doi.org/10.1016/j.actbio.2016.05.007

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customerswe are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, andreview of the resulting proof before it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Orthosilicic acid, Si(OH)4, stimulates osteoblast differentiation in vitro by

upregulating miR-146a to antagonize NF-κB activation

Xianfeng Zhou1,*, Fouad M. Moussa2, Steven Mankoci1, Putu Ustriyana1,Nianli Zhang3, Samir

Abdelmagid2, Jim Molenda4, William L. Murphy4, Fayez F. Safadi2, and Nita Sahai1,5,6,*

1Department of Polymer Science, University of Akron, Akron, OH 44325, USA

2Department of Anatomy and Neurobiology, Northeast Ohio Medical University, Rootstown,

Biomedical Science Graduate Program, Kent State University, Kent, OH, USA

3Department of Biologic and Materials Sciences, University of Michigan, Ann Arbor,

MI 48109, USA

4Department of Biomedical Engineering, Department of Orthopedics and Rehabilitation,

University of Wisconsin, Madison, WI 53705, USA

5Integrated Bioscience Program, University of Akron, OH 44325, USA

6Department of Geology, University of Akron, OH 44325, USA

Correspondence should be addressed to*X.Z. and *N.S.

E-mail:xianfeng@uakron.edu; sahai@uakron.edu

Fax: +1-330-972-7001

Phone: +1-330-972-5795

2

ABSTRACT

Accumulating evidence over the last 40 years suggests that silicate from dietary as well as

silicate-containing biomaterials is beneficial to bone formation. However, the exact biological

role(s) of silicate on bone cells are still unclear and controversial. Here, we report that

orthosilicic acid (Si(OH)4) stimulated human mesenchymal stem cells (hMSCs) osteoblastic

differentiation in vitro. To elucidate the possible molecular mechanisms, differential microRNA

microarray analysis was used to show that Si(OH)4 significantly up-regulated microRNA-146a

(miR-146a) expression during hMSC osteogenic differentiation. Si(OH)4 induced miR-146a

expression profiling was further validated by quantitative RT-PCR (qRT-PCR), which indicated

miR-146a was up-regulated during the late stages of hMSC osteogenic differentiation. Inhibition

of miR-146a function by anti-miR-146a suppressed osteogenic differentiation of MC3T3 pre-

osteoblasts, whereas Si(OH)4 treatment promoted osteoblast-specific genes transcription, alkaline

phosphatase(ALP) production, and mineralization. Furthermore, luciferase reporter assay,

Western blotting, enzyme-linked immunosorbent assay (ELISA), and immunofluorescence

showed that Si(OH)4 decreased TNFα-induced activation of NF-κB, a signal transduction

pathway that inhibits osteoblastic bone formation, through the known miR-146a negative

feedback loop. Our studies established a mechanism for Si(OH)4 to promote osteogenesis by

antagonizing NF-κB activation via miR-146a, which might be interesting to guide the design of

osteo-inductive biomaterials for treatments of bone defects in humans.

Key words: Silicic acid, osteoblastic differentiation, osteoclastic differentiation, microRNA, NF-

κB.

3

1. INTRODUCTION

The earliest work involving silicate and bone dates back to the early 1970s, when it was

suggested that silicate may play an important role in skeletal development and repair [1-3].

Subsequently, the positive effect of silicate on bone has raised the interest of multiple research

groups working on bone graft substitutes [4]. Silicate-based bioactive glasses [4-9], silicate-

substituted ceramics [10-15], and natural silicate bioceramics [16-20] have been developed and,

in most of cases, have shown better biological properties in orthopaedic applications when

compared to their silicate-free counterparts. Systematic studies by Hench et al. [21-23] and De

Aza [16, 17, 24] have identified silicate in bioactive glasses or glass-ceramics could be released

as Si(OH)4 by hydrolyzing Si-O-Si bridges and re-precipitate to form surface silanol groups. The

concentration of silanol groups could be critical because they act as a nucleation agent for apatite

formation [25, 26]. A “passive” role for silicate has been proposed in which the presence of

silicate in the materials results in changes of grain size, surface topography, etc., leading

ultimately to a change of biological response [4, 8, 9, 27]. There is evidence in the literature

which indicates that inorganic dissolution products (Si(OH)4, Ca2+, PO43-, etc.) from inorganic

silicate materials can influence and control the cell cycle of osteogenic precursor cells and

ultimately control the differentiated cell population [28, 29]. An “active” role, as claimed by

researchers and companies, is that silicate can affect bone cell metabolism [30]. Thus, the silicate

materials are not simply bone graft substitutes but, effectively, can be cell- and gene-activating

biomaterials [31]. However, none of the studies focused on the active roles have demonstrated

yet that the positive biological effects can be attributed solely to soluble silicate, since Ca2+ and

PO43-, which are released from silicate-based materials, are known to have osteo-inductive

effects on bone cells [32, 33]. A number of in vitro studies using osteoblast-like cells have added

4

to the evidence that silicate may be beneficial for bone formation. For example, silicic acid

addition to MG63 osteosarcoma cells has been shown to increase collagen type I (Col I) levels

1.75-fold [34]. The authors proposed a structural role (stabilization of collagen) or a metabolic

role (a co-factor for prolyl hydroxylase) for silicate [35]. In osteoblasts and mesenchymal stem

cells, silicic acid also increased the gene expression of osteocalcin (OCN) and alkaline

phosphatase (ALP) [36-38], which are excellent indicators of osteogenic differentiation and

mineralization. Recently, silica (SiO2) nanoparticles and nanoplatelets were shown to mediate

potent stimulatory effects on osteoblast differentiation [39, 40]. The spherical silica nanoparticles

of 50 nm were shown to enter the cells through caveolae-mediated endocytosis, which triggered

the stimulation of the mitogen activated protein kinase (MAPK) ERK1/2 (p44/p42) pathway.

These nanoparticles were likely released from the endosome to stimulate autophagy, which

suggested a cellular mechanism for the stimulatory effects of silica nanoparticles on osteoblast

differentiation and mineralization [41]. It should be noted, as the authors themselves point out,

that subtle changes in surface charge, surface area, shape, size, and/or surface chemistry of the

particles may play direct roles in determining specific cellular responses [39]. For example,

hexagonal silica nanoparticles with size about 100 nm did not show any effect on differentiation

of MSCs in vitro [42]. It is possible that not silicate chemistry but the nano-size particles could

be responsible for the positive outcome in these studies. Therefore, the exact biological roles of

soluble silicate on bone cells remain unclear and highly controversial [31].

Bone marrow-derived human mesenchymal stem cells (hMSCs) are a population of self-

renewing multipotent cells and are excellent candidates for cell-based therapeutic strategies to

regenerate injured or damaged tissue. They can differentiate into multiple lineages, including

osteogenic lineage, in response to appropriate stimuli [43]. It is now evident that osteogenic

5

differentiation of MSCs is tightly controlled by several regulators including microRNAs

(miRNAs) [44, 45]. MiRNAs are small endogenous RNA molecules that govern gene expression

by targeting mRNA at post-transcriptional level. miRNAs have emerged as key regulators of

diverse cellular processes, including growth, apoptosis, development, metabolism, stress

adaption, differentiation, etc. [46-48]. For example, miR-138 was shown to inhibit osteoblast

differentiation of hMSCs by targeting FAK (focal adhesion kinase) translation. Consequently,

reduced phosphorylation of FAK and ERK1/2 (Extracellular signal-related kinase) resulted in a

decrease of Runt-related transcription factor 2 (Runx2) phosphorylation [49]. miR-29b has been

reported to promote osteogenesis by directly down-regulating known inhibitors of osteoblast

differentiation, HDAC4, TGFβ3, ACVR2A, CTNNBIP1, and DUSP2 proteins, through binding

to target 3’-UTR sequences in their mRNAs [50]. Various miRNAs that positively or negatively

regulate osteoblast differentiation were summarized by Vimalraj and Selvamurugan [51].

However, no miRNAs have been identified to bridge the biological functions of silicate with

bone formation.

In the physiological pH and at concentration below 2 mM (56 ppm Si), silicate exists

predominantly as orthosilicic acid (Si(OH)4). This tetrahedral Si(OH)4 is uncharged at neutral pH,

but has the tendency to polymerize to polysilicate at silicate concentration above 2~3 mM

(56~84 ppm Si) [52]. To clarify the direct effects of silicate on bone formation, we examined the

effect of (Si(OH)4) on bone cell proliferation, differentiation, and intracellular signaling events.

We demonstrated that soluble Si(OH)4 stimulated hMSCs to proliferate in growth medium and

enhance osteogenic differentiation in osteogenic medium. We also sought to ascertain at least one

mechanism by which Si(OH)4 accomplishes these bioactivities by regulating microRNAs

expression.

6

2. MATERIALS AND METHODS

2.1. Si(OH)4 preparation

Final concentrations of Si(OH)4 solution with 0-50 ppm Si were prepared by adding sodium

metasilicate (Na2SiO3.9H2O) in αMEM and then adjusting pH to 7.2 ~ 7.4 with 1M HCl before

adding FBS or any other supplements (Equation below). Inductively coupled plasma optical

emission spectrometry (ICP-OES) analysis was performed to establish the ion concentration for

Si in the medium using an Agilent Technologies 710 ICP-emission spectrometer. Standard

solutions (SQCSTD27-100B) were purchased from RICCA Chemical Company for the

calibration of the instrument. All the reagents, unless stated specially, were obtained from Sigma-

Aldrich and were of the highest available purity.

Si

O

O O

3H2O Si

OH

OHHO

OH

2OH-++

2.2. Cell Culture

Bone marrow-derived hMSCs (Lonza®) at passage 4 were grown in regular growth medium

(αMEM, 10% FBS, and 1% penicillin-streptomycin). The cells were cultured until 70~75%

confluence. Then, cells were trypsinized and seeded onto 24-well culture plates at the density of

3 × 103 cells/cm2. After 24 hours, the medium was replaced with Si-containing (0, 10, 50 ppm)

medium.

MC3T3-E1calvarial cells and mouse monocytic cell line RAW264.7 were purchased from

American Type Culture Collection (ATCC, Manassas, Virginia) and cultured in DMEM + 10%

FBS. All cultures were grown at 37 °C and 5% CO2.

2.3. The effects of Si(OH)4 on hMSC cytotoxicity and proliferation

Cytotoxicity of Si(OH)4 was determined after 24 hours incubation using Live/Dead staining

7

(Invitrogen®) according to the manufacturer’s protocol.

Cell proliferation was measured with the CyQuant Cell Proliferation Assay kit (Invitrogen®). At

the end of incubation time (Day 1, 7, 14, and 21), the cells were washed with PBS to remove

non-adherent cells and frozen at -80 °C until being assayed. After thawing, 100 µL of lysis buffer

was added to the cell pellet at room temperature for 1 h and then 100 µL of CyQuant GR dye

was mixed with lysis buffer in 96-well plate and incubated for 5 min. Fluorescent signals were

detected with excitation at 480 nm and emission at 520 nm using a fluorometricplate reader

(Synergy H1, Biotek®). Values were determined against a standard curve and displayed as ng/mL

DNA. Five replicates were performed.

2.4. The effects of Si(OH)4 on osteogenic differentiation of hMSCs

Osteoblastic differentiation was induced using osteogenic medium (growth medium

supplemented with 0.1 µM dexamethasone, 50 µM ascorbic acid 2-phosphate, and 10 mM β-

glycerol phosphate) with or without Si(OH)4. The medium was changed twice weekly. The

osteoblast phenotype was evaluated by determining ALP activity and corrected by the total DNA

content [19]. ALP activity was measured using a fluorometric endpoint assay (ab83371), which

quantified ALP enzyme by cleaving the phosphate group of non-fluorescent 4-

methylumbellifferyl phosphate disodium salt (MUP) substrate resulting in an intense fluorescent

signal. Alizarin Red S staining (ARS) was performed to detect matrix mineralization. On day 28,

cells were washed three times with PBS and then fixed with 10% formalin (10 minutes). The

fixed cell were further washed with distilled water in order to remove any salt residues and then

a solution of ARS (40 mM) with a pH adjusted to 4.2 was added to the wells. After an incubation

of 20 minutes at room temperature, the excess of ARS was washed with distilled water.

Mineralization was quantified by measuring the absorbance of orange-red ARS. After staining,

8

10% (v/v) acetic acid was incubated with the cells overnight at -20 °C. Then, the cells with the

acetic acid were transferred to tubes and centrifuged for 15 minutes at 20,000 g. The supernatant

was removed to other tubes and neutralized with 10% (v/v) ammonium hydroxide. 100 µL of

each sample was added to 96-well plates and the OD405 was read using microplate reader.

2.5. The effects of Si(OH)4 on osteoclast differentiation of mHSCs

Mouse hematopoietic stem cells (mHSCs) were isolated and cultured as described: 4 - 8 weeks

old male C57/Blk-6 mice were dissected and both tibiae and femur marrow cavities were flushed

with 10 ml media (αMEM 10% FBS, 1% penicillin-streptomycin, and 0.1% Amphotericin) into a

50 ml centrifuge tube using a sterile 20-gauge needle. All flushed bone marrow was collected

and centrifuged at 1200 rpm for 12 minutes at 4 °C. The supernatant was discarded and the pellet

was re-suspended with fresh media (αMEM, 10% FBS, 1% penicillin-streptomycin, and 0.1%

amphotericin). The re-suspended mix was plated on 10 cm dish to incubate for 24 hours and non-

adherent cells were cultured at a density of 1 × 105 cells/cm2 for 2 days more in the presence of

20 ng/ml of macrophage colony-stimulating factor (MCSF). The resultant preosteoclasts were

further cultured in medium supplemented with 20 ng/ml of MCSF and 20 ng/ml of receptor

activator of nuclear factor-kappa B ligand (RANKL). The osteoclast phenotype was evaluated by

determining tartrate resistant acid phosphatase (TRAP) activity using a commercially available

kit. The number of mature osteoclasts (TRAP positive with ≥ 3 nuclei) was counted under light

microscopy. Pit formation assays were conducted using BD BioCoatOsteologic bone cell culture

system (BD Biosciences®). Resorption areas were examined under normal light microscope and

pictures were analyzed using Image J and Photoshop.

2.6. RNA extraction, cDNA synthesis and quantitative real-time RT-PCR.

9

The cell culture medium was removed and, after washing cells with RNase-free PBS, total RNA

was extracted using Sure PrepTM RNA purification kit (Fisher Bioreagents®). The purity and

quantity of RNA was determined spectrophotometrically using a NanoDrop 1000 (Thermo

Fisher Scientific®). The ratio of sample absorbance at 260 and 280 nm was used to assess the

purity of the RNA. A ratio of ~ 2.0 is acceptable as pure RNA. The RNA samples were treated

with DNAse and reverse-transcribed using the SuperScript first-strand synthesis system for real-

time PCR (Invitrogen®). SYBR (N’, N’-dimethyl-N-[4-[(E)-(3-methyl-1,3-benzothiazol-2-

ylidene) methyl]-1-phenylquinolin-1-ium-2-yl]-N-propylpropane-1,3-diamine) green mRNA

qRT-PCR was performed using ABI PRISM 7700 Real PCR system for analyzing expression of

Runx2, OCN, OPN, Col 1, NFATc1, CALCR, DC-STAMP, and TRAP genes. GAPDH served as

the house-keeping gene and the expression for the gene of interest was normalized to the

GAPDH expression. The miRNA qRT-PCR primer specific for miR-146a was purchased from

Applied Biosystem®. Each sample was run in triplicate. The relative target mRNA expression

was calculated by using the ∆∆Ct method.

2.7. Microarray.

Total RNA with small RNA (< 200 nucleotides) from cultured cells was isolated, immediately

frozen in dry ice, and submitted to the Case Comprehensive Cancer Center in Case Western

Reserve University/University Hospital of Cleveland. 800 ng of RNA was used for RNA labeling

with FlashTag® HSR RNA Labeling Kit (Genisphere®). The biotin-labeled RNAs were

hybridized with miRNA 2.0 arrays (containing 633 human miRs) at 48 °C for 16 hours, stained

using theGeneChip Fluidics Station 450, and scanned with GeneChip Scanner 3000.

2.8. Transfection of oligonucleotides.

Synthetic mmu-miR-146a mimic (Qiagen®) is a small, double-stranded RNA molecule designed

10

to mimic endogenous mature miR-146a molecule when introduced into cells. The mature

sequence is 5’UGAGAACUGAAUUCCAUGGGUU3’. Synthetic mmu-miR146a inhibitor

(MISSION®), also a double-stranded RNA molecule, was designed to inhibit the mature miR-

146a. Transfection of 5-20 nMmiR mimic (or inhibitor) was performed according to the

manufacturer’s instructions using siLentFectTM Lipid reagent (Bio-Rad®). In brief, MC3T3-E1

preosteoblasts were plated onto 24 wellplates for 24 hours prior to miRNA transfection.

Transfection mix was prepared in serum-free medium and transfection was done in antibiotic-

free medium. Transfection efficiency was monitored by using 5’ carboxyfluorescein-labeled

oligonucleotides. After 4 hour’s incubation, the medium was changed to normal medium with or

without Si(OH)4. Cells were harvested for RNA isolation 72 hours post transfection. miR-146a

levels were confirmed by qPCR. Transcription levels of the target proteins in transfected cells

were analyzed by qPCR using the corresponding primers. The osteoblast phenotype was

evaluated by ALP staining at day 10 and ARS staining at day 21.

2.9. NF-κB Reporter Transfection and Luciferase Assays.

Cells (MC3T3-E1 or RAW264.7) were seeded in 12-well culture plates until ~ 70% confluency

and transfected with 1 µg of commercial NF-κB responsive reporter pNF-κB-Luc plasmid (a

kind gift from Kent State University, Ohio, USA), using Lipofectamine 2000 reagent

(Invitrogen®) as per the manufacturer’s protocol. For miRNA inhibition experiments, cells were

co-transfected with reporter clone of NF-κB plasmid and anti-miR-146a and harvested 72 hours

post transfection. NF-κB activation was induced in MC3T3 cells by addition of TNFα (10 ng/ml).

NF-κB activation was stimulated in RAW264.7 cells by treatment with RANKL (50 ng/ml).

Luciferase activity was measured using the Luciferase Assay kit (Promega®) and normalized

with total protein content (Pierce BCA Protein assay). Data are expressed as Relative Light Units

11

(RLU) and represent the mean ± SD of four replicate wells for each data point.

2.10. Nuclear protein extraction and NF-κB Transcription Factor Assay.

MC3T3 cells were treated with TNFα (10 ng/ml) and miR-146a mimic in presence or absence of

Si(OH)4 to assess the effect of silicate exposure on NF-κB expression.RAW264.7 cells were

stimulated with RANKL (50 ng/ml) and miR-146a (10 nM) in presence or absence of Si(OH)4 to

evaluate the NF-κBactivation. The EpiSeeker Nuclear Extraction kit (abcam®) was used to

extract nuclear proteins from cells 72 hours post transfection. NF-κB level in the nuclei was

analyzed by western blot with the corresponding antibody. DNA-binding ELISA for NF-κB p65

Transcriptional Factor Assay kit (abcam®) was used to quantify the nuclear NF-κB levels,

according to the instructions of the manufacturer. Briefly, nuclear extractions were incubated in

96-well plates coated with immobilized oligonucleotide containing a binding site for p65 subunit

of NF-κB. NF-κB p65 was detected by incubation with the primary antibody specific for the

activated form of p65, visualized by anti-IgG horseradish peroxidase (HRP)-conjugate and

developing solution and quantified by spectrophotometry at 450 nm with a reference wavelength

of 655 nm. Nuclear NF-κB expression was normalized by total nuclear fraction protein and fold

changes were calculated with respect to the control.

2.11. Immunofluorescence of Runx2.

MC3T3 cells were treated with TNFα (10 ng/ml) in presence or absence of Si(OH)4 to assess the

effect of Si(OH)4 exposure on Runx2 expression. After 14 days of culture withSi(OH)4, cells

were fixed with 10% formalin for 10 min and rinsed with PBS. Then, cells were permeabilized

with 0.1% Triton X100 for 5 min and nonspecific binding was blocked with 10% (m/v) BSA.

Cells were incubated overnight at 4 °C with primary antibody (mouse monoclonal to Runx2,

abcam®, ab76956). After incubation, cells were rinsed 3 times with PBS for 5 min and incubated

12

with secondary antibody (Alexa Fluor goat-anti-mouse IgG). Actin was counterstained with

phalloidin-rhodamine and then rinsed 3 times. Immunolabeling was qualitatively analyzed using

an IX51 Epifluorescence Microscope (Olympus Co., Japan) equipped with fluorescent light

source and filters.

2.12. Statistical Analyses.

Data are represented as mean ± SD from ≥ 3 replicates for each condition/time point, and each

experiment was repeated at least three times independently. Comparisons were made by using

one-way ANOVA with post-hoc testing or Student’s t test. All the comparisons and statistical

analyses have been done in Microsoft excel. Observations were considered to be significantly

difference for p < 0.05.

3. RESULTS

3.1. Effects of Si(OH)4 on hMSC Osteoblastic Differentiation.

Using the CyQuant® cell proliferation assay, we found the steady increase of cell proliferation

with time (Fig.1a), which indicated that the cells were proliferating and the Si(OH)4

concentration in this range had no overtly toxic effect to the hMSCs. This result was confirmed

by Live/Dead cell staining, which showed that in the presence of Si(OH)4, the percentage of

fluorescent green cells (living cells) did not drop below 98% (Fig.1b). We also found that

Si(OH)4 significantly increased the proliferative activity of hMSCs between 3 and 7 days of

incubation in growth medium (Fig. 1a). However, no significant difference in cell proliferation

was shown for different doses of Si(OH)4 after 7 days culture. To study the impact of Si(OH)4 on

the osteogenic differentiation of hMSCs, the hMSCs were cultured in osteogenic medium in

different doses of Si(OH)4. A convenient fluorogenic assay was used to detect the activity of ALP,

13

an enzyme involved in bone mineralization produced by osteoprogenitors and osteoblasts (Fig.

1c). hMSCs cultured for 7-14 days in osteogenic medium typically differentiate into ALP-

producing osteoblasts. In the Si(OH)4-supplemented medium, the cells revealed higher levels of

ALP production compared to control cells cultured without Si(OH)4 at days 7, 14 and 21. An

extended culture time of 28 days allowed positive identification of mineralized nodules, detected

by ARS staining for calcium deposition. Mineralization was quantified by extracting the ARS

from the stained wells. In contrast to the results of ALP activity, increasing Si(OH)4 in medium

decreased the matrix mineralization , shown in Fig. 1d.

To further evaluate the effect of Si(OH)4 on the osteogenic differentiation of hMSCs, the

transcription levels of osteoblast-specific differentiation markers were assayed by qRT-PCR

under osteogenic condition. As shown in Fig. 2, Si(OH)4 treatment increased mRNA

transcription levels of Col 1, the major structural component of the organic matrix. This increase

was maximal on day 7 after Si(OH)4 treatment, consistent with an early role of Col 1 in the

osteoblast maturation sequence. Another notable finding is the up-regulation of mRNA levels of

Runx-2, the master transcription factor responsible for osteogenic differentiation [53].

Transcription of OCN and osteopontin (OPN), major osteoblast proteins downstream of Runx-2,

were also induced by addition of Si(OH)4 (Fig. 2). However, treatment with Si(OH)4 did not alter

the mRNA transcription of the house-keeping gene GAPDH (Fig. S1). Together, these findings

suggest that Si(OH)4may act as a positive regulator of osteogenic differentiation. Based on the

cytotoxicity, proliferation, and osteogenic differentiation studies, we selected 50 ppm of

Si(OH)4to gain a better understanding of the mechanism of action.

3.2. Identification of Differentially Expressed miRNA in Si(OH)4-treated hMSCs.

To identify differentially expressed miRNAs during osteogenic differentiation of hMSCs treated

14

with Si(OH)4, total RNA from the hMSCs were labelled using FlashTag® HSR RNA labelling kit

for miRNA microarray analyses. A total of 663 human miRNAs were analyzed (Fig. S2). Of the

663 miRNAs, 26 were consistently altered by Si(OH)4 treatment (50 ppm). Among them, eleven

miRNAs were down-regulated (blue color), and fifteen miRNAs were up-regulated (red color).

MiR-146a wasthe most robustly upregulated miRNA by Si(OH)4treatment during hMSC

osteogenic differentiation (Fig. 3a). To further validate that miR-146a is indeed upregulated by

Si(OH)4, qRT-PCR was performed and showed miR-146a was significantly up-regulated during

the later stages of hMSC osteogenic differentiation(day 14 and day 21, Fig. 3b).

Pre-osteoblasts are the late-stage cell lineage of MSCs during osteo-differentiation. To evaluate

whether the biological effects of miR-146a is of critical importance in later stages of osteogenic

differentiation, we inhibited miR-146a levels in MC3T3-E1 pre-osteoblasts using synthetic

antisense miRNA inhibitor (anti-miR-146a, Mission®). Also, in determining the role of miR-

146a, it is advantageous to use MC3T3-E1 cells because MSCs are more difficult to transfect

successfully with commercial transfection reagents. The degree of miRNA inhibition was

monitored by qRT-PCR after transfection of anti-miR-146a. Mature miR-146a levels were

decreased ~ 1.6 fold relative to control cells 72 h post-transfection (Fig. S3).

MC3T3-E1 pre-osteoblasts were then induced to differentiate into osteoblasts after transfection

with anti-miR-146a, in the presence or absence of Si(OH)4. Inhibition of miR-146a decreased

osteoblastic differentiation, which was indicated by lower transcription levels of the osteoblast-

specific genes Runx2, Col 1, ALP, and OCN (Fig. 3c), as well as by decreased ALP production

(visualized by ALP staining) and mineralization (Fig. 3d). However, this suppression of

osteoblast marker gene transcription, ALP production, and mineralization could be reversed by

Si(OH)4 treatment (Fig. 3c&3d). Therefore, our results suggest that Si(OH)4may be a positive

15

regulator of osteoblast differentiation via miR-146a.

3.3. Si(OH)4 targets NF-κB signaling via miR-146a.

Since miR-146a has been reported to be NF-κB dependent and directly down-regulate NF-

κBsignaling[54], the knockdown studies of miR-146awere performed to determine the direct

regulatory role of Si(OH)4on the NF-κB activation by regulating miR-146a expression. MC3T3

pre-osteoblasts were transfected with a luciferase reporter driven by tandem NF-κB consensus

motifs and were treated with anti-miR-146a (co-transfection) in the presence or absence of

Si(OH)4. TNFα-induced NF-κB promoter activity was increased upon anti-miR-146a

transfection, as shown by increase in luciferase activity (Fig. 4a), which indicated the inhibitory

role of miR-146a on NF-κB promoter activation. This result is in agreement with the previous

studies in which the inhibition of miR-146a increased the activity of NF-κB[55]. Treatment of

MC3T3-E1 cells with Si(OH)4 was found to significantly decrease TNFα-induced NF-κB

transactivation even in presence of anti-miR-146a (Fig. 4a), which means the up-regulation of

NF-κB can be effectively counter-balanced by increasing the cellular levels of miR-146a through

Si(OH)4 treatment.

We further validated the capacity of Si(OH)4 to suppress NF-κB signaling in MC3T3-E1 cells by

extracting cell nuclear proteins and subsequently performing western blot analysis.These data

showed that the nuclear NF-κB protein level in miR-146a-transfected cellswas less at day 3 when

compared with TNFα-treated control cells.Treatment with Si(OH)4to miR-146a-transfected cells

further markedly decreased NF-κB translocation to the nucleus (Fig. 4b).To quantify the Si(OH)4

regulating NF-κBactivation, we also exposedthe nuclear extract of MC3T3-E1 cells towells

coated with NF-κB binding DNA (abcam®) and performed anELISA(Fig. 4c). TNFα induced

significant binding to the specific double-stranded DNA containing the NF-κB consensus, shown

16

from the about200% increase of NF-κB p65 trapped in the wells. MiR-146a transfection to

MC3T3-E1 cellsdecreased NF-κB binding to its consensus binding site (~28% decrease after

transfection). Si(OH)4 treatment to the miR-146a transfected cells further potently diminished

NF-κB translocation (p < 0.05) to the nucleus and decreased DNA binding (~50% decrease with

Si(OH)4 treatment). These observations support the inhibitory role of Si(OH)4 on NF-κB activity

and indicate miR-146a as an important component to the effect of Si(OH)4.

3.4. Si(OH)4 increased Runx2 expression via NF-κB signaling

To further understand the roles of Si(OH)4 regulating NF-κB activity to influence osteoblastic

differentiation, immunofluorescence was used to detect expression of the bone-specific

transcription factor Runx2, which is known to be downstream of NF-κB and essential for

osteoblastic differentiation (Fig. 5). All MC3T3-E1 cells cultured for 14 days in osteogenic

medium exhibited good cell density. Cells treated with TNFα remained bipolar and

predominantly fibroblast-likein appearancebut showed negligible Runx2 production. However,

the cells cultured in Si(OH)4-containing medium were observed to raise the level of Runx2

expression. Furthermore,the TNFα-induced Runx2 inhibition could be rescued by treating the

cells with Si(OH)4.This observation demonstrated that Si(OH)4could mediate potent stimulatory

effects on osteoblastic differentiation by suppressing TNFα-induced activity ofNF-κB.

4. DISCUSSION

Our previous studies [20] identified that β-CaSiO3 is more bioactive than α-CaSiO3 in inducing

hMSCs osteogenic differentiation. The two polymorphs of CaSiO3 have identical chemical

composition and stoichiometry but different crystal structures. This difference in crystal structure

results in the greater dissolution rate of silicate from β-CaSiO3 compared to α-CaSiO3 [19, 56].

17

Soluble silicate concentrations peaked at higher levels and earlier time points for β-CaSiO3 than

for α-CaSiO3, which may address the issue whether soluble silicate is bioactive to regulate bone

cell metabolism and bone formation. However, as pointed out in a critical review, it is not clear if

the positive effects can be purely attributed to silicate [31]. The aim of the present study was

therefore to test the hypothesis that Si(OH)4, the simplest soluble form of silicate, has a direct

effect on osteogenesis and bone formation in vitro and investigate the potential cellular

mechanisms by which Si(OH)4 might influence bone cell behaviors.

First, our data suggested that Si(OH)4 stimulated hMSCs proliferation and osteogenic

differentiation (Fig. 1 & 2). However, our data seemed to indicate an inhibitory role of the

soluble silicate in mineralization, which might be contrary to Carlisle, Schwarz, and Milne’s

dietary silicon deprivation studies in chickens and mice [1-3]. In the animals, silicon was taken

up via gastrointestinal absorption and finally concentrated and bonded within the osteoid layer of

growing bone. Therefore, the silanol groups, which have been known as nucleation sites for

apatite formation, could be successfully formed. The apatite is formed by the co-precipitation of

phosphate and Ca2+ in the matrix of collagen. Since soluble silicate is uncharged in neutral cell-

medium environment, perhaps, oligomers of soluble silicate [52] which have lower pKa than

Si(OH)4 [57], could de-pronate and attract Ca2+ in the solution to form ion clusters and decrease

the Ca2+ concentration close to collagen network (Fig. 5b). Although, Si(OH)4 has been shown in

previous work [34] and in the current study to increase the production of collagen matrix (Fig. 6a)

the limited concentration of Ca2+ might result in the lower apatite formation (Fig. 1). However, at

this point, further experimental evidence is needed to gain a complete understanding of all the

roles of the various silicon species in the process of apatite formation.

Additionally, within the current study, a mechanism has been described by which Si(OH)4

18

regulates the expression of miR-146a, which ultimately suppresses the activity of NF-κB in bone

cells to enhance osteoblastic differentiation.Several recent studies have shown that NF-

κBinfluences both osteoblast differentiation and activity. In one of the studies, it was shown

thatthe inhibition of NF-κB p65 by a mutant IKKγ (IKK-DN) and a super IκBα repressor (SR-

IκBα) enhanced osteoblastic differentiation of mesenchymal C2C12 cells in vitro[58]. NF-κB

p65 can interact with the Smad1-Smad5 complex in the nucleus and disrupt its binding to target

promoters [59]. Inhibiting this interaction of NF-κB with TGFβ-mediated Smad signaling can,

therefore, potently stimulate the differentiation and mineralization of primary murine bone

marrow stromal cells [60]. Runx2, the primary transcription factor required for commitment to

the osteoblast lineage, is NF-κB dependent [45]. In our studies, Si(OH)4 significantly

antagonized TNFα-induced NF-κB activity(Fig. 4) and activated Runx2 (Fig. 5) in MC3T3 cells.

Therefore, the effects of Si(OH)4 in osteoblasts could be suggested to block TNFα-dependent

canonical NF-κB signaling to activate Runx2pathway.

MiR-146a was previously identified as an immune system regulator that influences the

mammalian response to microbial infection. Further characterization of miR-146a revealed that it

is also NF-κB dependent, which has been shown to base-pair with sequences in the 3’ UTR of

TRAF6 (TNF receptor associated kinase) and IRAK1 (IL-1 receptor associated kinase) genes,

the key adaptor molecules downstream of toll-like and cytokine receptors in the NF-κB pathway

[61]. In the present study, miR-146a was found as a miRNA signature of osteogenic

differentiation of hMSCs treated with Si(OH)4.Data obtained from our in vitro experiments

revealed that inhibition of miR-146a function decreased osteogenic potential of pre-osteoblasts,

whereas Si(OH)4 treatment enhanced osteoblast differentiation (Fig. 3).Therefore, these findings

suggested a regulatory role of Si(OH)4, in which Si(OH)4 up-regulates the expression of miR-

19

146a gene that, upon processing and maturation, down-regulates NF-κB to reduce the activity of

Runx2 (Scheme 1). However, how Si(OH)4 up-regulates miR-146a in bone cellsis presently

unknown and remains to be investigated in our lab.

Recently, some studies have shown that soluble silicate inhibited osteoclast phenotypic gene

expression, osteoclast formation and function in vitro[62]. In agreement with their study, we also

demonstrated the inhibitory effects ofSi(OH)4 on osteoclast formation of hematopoietic stem

cells from 4-8 week old male C57/BLK-6 mice (Fig. 7a). MCSF/RANKL induced the

differentiation of mHSCs into osteoclasts[63], which were distinguished by their TRAP-positive

and multinucleated phenotype (Fig. 7a). Si(OH)4 treatment resulted in a significantly reduced

number of both osteoclasts and TRAP+ pre-osteoclast formation (Fig. 7a). In addition, the cells

treated with Si(OH)4 showed functional defects in osteoclast function, shown by decreased

number and average size of pits on BD BioCoatOsteologic bone cell culture system (Fig. 7b).

Cell viability assay demonstrated that this concentration of Si(OH)4 failed to mediate any toxic

effects on the viability or proliferation rates of mHSCs (Fig. S4). This suggests that the anti-

osteoclastogenic activities are related to suppression of differentiation along the osteoclast linage

rather than as a consequence of cytotoxicity. Furthermore, Si(OH)4 was shown to down-regulate

mRNA expression levels of osteoclast marker genes in mHSCs (Fig. 7c). NFATc1, the master

regulator of osteoclastogenesis[64], is induced potently by RANKL at early stage (day 2).

Notably, Si(OH)4 decreased mRNA expression levels of NFATc1 at both early stage (day 2) and

late stage (day 4). The transcription of DC-STAMP, a downstream target of NFATc1 to promote

cell-cell fusion of osteoclasts [65], was suppressed by the addition of Si(OH)4. The other

osteoclast-specific genes such as TRAP and CALCR, activated by transcriptional complex

containing NFATc1 [64], were also down-regulated by Si(OH)4 treatment.By using a Boyden

20

chamber co-culture system, various research groups concluded that the inhibition of

osteoclastogenesis of RAW264.7 pre-osteoclasts was due to the increase the level of

osteoprotegerin (OPG) released by SaOS-2 osteoblasts after exposed to silicate. They

suggestedan intercellularmechanism, by which silicate increased the levels of OPG, a soluble

decoy receptor for RANKL and inhibits the RANK-RANKL signaling pathway [62,

66].However, it has been recognized that silicate should have direct inhibiting effects on

differentiation and fusion of osteoclast precursors by interacting with important intracellular

signaling pathway [62].Since NF-κB is well known as one of the most important molecules in

osteoclastogenesis to activate osteoclast precursor cells, we hypothesized Si(OH)4 treatment

could apply the same pathway to up-regulation of miR-146a and potentially lead to suppression

of osteoclast formation and function. To explore this, we transfected miR-146a mimic into pre-

osteoclast RAW264.7 cells and osteoclastogenesis was induced by MCSF/RANKL treatment.

The number of TRAP-positive multinucleated cells was significantly reduced by transfection of

miR-146a mimic (Fig. 8a&b). Furthermore, we transfected RAW264.7 cells with the luciferase

reporter driven by tandem NF-κB consensus motifs, and then treated the cells with Si(OH)4 and

miR-146a mimic. Both Si(OH)4and miR-146a were found to suppress RANKL-induced NF-κB

translocation from cytosol to nucleus in RAW264.7 cells (Fig. 8c).These results clearly indicated

that Si(OH)4 might negatively regulate NF-κB via miR-146a to affect osteoclast formation, and

could play a role in the remodeling process of bone by inhibiting the bone physiological

resorption process [67, 68].

Taken together, Si(OH)4 may reveal a considerable biomedical potential with therapeutic and

prophylactic implementations, e.g. concerning osteoporosis. This disease is due to an imbalance

between bone formation and resorption, which is most common in women after menopause and

21

older men [69]. Therefore, Si(OH)4 could be used as a nutritional supplement in osteoporotic

persons [35, 70]. Until now, no reports are available that indicate either a passive or an active

influx of uncharged or charged Si(OH)4 into bone cells. In siliceous sponges and diatoms, silicate

may be transferred into cells through the membrane [71]. Experimental evidence indicates that

the membrane protein may use Si(OH)4 instead of HCO3- for the co-transport of Na+ (Na+-HCO3

-

co-transporter, NBC-1) [66, 71, 72]. The lack of Si-labelling technique makes it hard to

identifyhowSi(OH)4 enters bone cells and up-regulates the expression of miR-146a, which would

be investigated in the future.

5. CONCLUSION

We found that soluble Si(OH)4 can contribute to bone-formationin vitrobut reduce the activity

and number of osteoclasts. Si(OH)4 induces the expression of endogenous miR-146a to

antagonize activation of NF-κB. This inhibition of NF-κB could result in activation of Runx2,

the master transcription factor for osteoblast precursor differentiation, and suppresses the

expression of NFATc1, the key transcription gene for osteoclast precursor differentiation. To the

best of our knowledge, this is the first study showing soluble silicate’s possible cellular

mechanism. These data suggest that soluble Si(OH)4, along with miR-146a, can be used as

pharmaceuticals for bone fracture prevention and therapy in conditions such as localized bone

loss or generalized osteoporosis.

SUPPLEMENTARY INFORMATION AVAILABLE:The microRNA microarray heat map,

qRT-PCR of GAPDH and miR-146a, and viability of mHSCs to Si(OH)4.

ACKNOWLEDGMENTS

22

We are grateful to Dr. Kan He, Department of Pathology, University of Pittsburgh School of

Medicine, for excellent technical assistance and helpful discussion. Mr. Weilong Zhao in

University of Akron is gratefully thanked for assistance with illustration of the mechanism. The

Gene Expression and Genotyping Facility at Case Western Reserve University are acknowledged

for the help with microRNA microarray and data analysis. NF-κB Luc plasmid is the gift from Dr.

Thomas in Kent State University. Financial support was provided by NSF DMR (0906817) grant

and start-up funds from University of Akron, and Ohio Research Scholar Program to Dr. Nita

Sahai, PI.

23

FIGURE CAPTIONS

Figure 1. Effects of Si(OH)4 on the proliferation and osteoblastic differentiation of hMSC. (a)

hMSCs were grown in growth medium treated with different doses of Si(OH)4. Cell proliferation

was measured by the CyQuant proliferation assay kit and displayed as ng/mL DNA (n = 5 for all

experiments). (b) Cell viability and morphology of hMSCs after incubation with different doses

of Si(OH)4 for 24h. Live cells are stained fluorescent green and dead cells appear red. (c)

Osteoblastic different iation of hMSCs in osteogenic medium supplemented with different doses

of Si(OH)4. ALP activity was measured during the course of differentiation by normalizing to

total DNA (n = 4 for all experiments), *p < 0.05, **p < 0.01. (d) Alizarin Red staining were

performed at day 28. Mineralization was quantified by measuring the absorbance of orange-red

ARS (insert value). Data presented are from one of three independent experiments, with all three

exhibiting comparable results.

Figure 2. The specific osteogenic differentiation markers (Col1, Runx2, OCN, and OPN) were

assayed by qRT-PCR. Data presented are from one of three independent experiments, with all

three exhibiting comparable results. Values are presented as mean ± SD by normalized to

GAPDH (n = 3 for all experiments), **p < 0.01.

Figure 3. Expression of miR-146a at day 14 & day 21 in osteoblastic differentiation of hMSCs

treated with or without Si(OH)4. (a) The heat map was generated with the use of hierarchical

cluster analysis to show distinct miRNA expression patterns in hMSCs treated with or without

Si(OH)4 in osteogenic medium. The color bar was extracted to show the color contrast level of

the heat map. Red and blue indicate high expression levels and low expression levels,

respectively. (b) miR-146a transcription levels were determined by qRT-PCR during hMSCs

osteogenic differentiation treated in the presence or absence of Si(OH)4. Values are presented as

24

the mean ± SDs (n = 3). (c) Anti-miR-146a inhibited osteoblast differentiation on day 3, as

shown by qRT-PCR analysis of osteoblast-specific genes (Runx2, Col1, ALP,

andOCNnormalized to GAPDH). However, Si(OH)4 treatment promoted the transcription of

these genes even in the presence of anti-miR-146a. *p < 0.05;**p < 0.01, (n = 3 for all

experiments). (d) ALP staining at day 10 and Alizarin Red staining at day 21 were performed.

Data presented are from one of three independent experiments, with all three exhibiting

comparable results.

Figure 4. Si(OH)4 treatment suppressed NF-κB activation in MC3T3 pre-osteoblast by

increasing miR-146a expression. (a) MC3T3 cells were transfected with NF-κB-responsive

luciferase reporter, and then co-transfected with anti-miR-146a in presence or absence of

Si(OH)4. TNFα was used to induce NF-κB activation in cells and luciferase activity was

quantitated 48 hours later. Data expressed as relative light units (RLU) normalized to total

protein (n = 4). (b) MC3T3 cells stimulated with TNFα, and then treated with miR-146a in

presence or absence of Si(OH)4 for 48 hrs. Western blot analysis was performed for NF-κB in

nuclear extract. (c) NF-κB contained in nuclear extract can bind specifically to the NF-κB

response dsDNA immobilized onto the wells. NF-κB subunit p65 of mouse was detected by

addition of a specific primary antibody and a secondary antibody conjugated to HRP. The data

represent the means ± SDs (n = 4). Data presented are from one of three independent

experiments, with all three exhibiting comparable results.

Figure 5. Runx2 staining of MC3T3 preosteoblasts after 14 days of culture with Si(OH)4.

MC3T3 cells treated with TNFαlacked the positive Runx2 stain; and Si(OH)4 can stimulated

Runx2 expression, even under the influence of TNFα. Data presented are from one of three

independent experiments, with all three exhibiting comparable results.

25

Figure 6. Soluble silicic acid increased extracellular matrix formation but decreased

mineralization in vitro. (a) Silicic acid increased the expression of collage, shown by Sirius

Red/Fast Green staining. Sirius Red specifically binds to the [Gly-X-Y]n helical structure of

fibrillar collagens and Fast Green binds to non-collagenous proteins. (b) The proposed

mechanism why calcium deposition was inhibited by soluble silicate. The soluble

silicateoligomers could be de-protonated in neutral buffer and attract Ca2+ to form ion cluster,

which could decreased Ca2+ concentration close to collagen network to suppress the

Ca2+deposition.

Figure 7. Si(OH)4 suppressed RANKL-inducedmousehematopoietic stem cells (mHSC)

differentiation to osteoclasts in vitro. 10 ppm of Si(OH)4 was used to evaluate osteoclast

formation of mHSCs. (a) Microphotographs are representative TRAP staining of cells treated

without or with Si(OH)4. Scale bars represent 100 µm. Top right is mature multinucleated

TRAP+ osteoclasts quantitated in the presence of M-CSF and RANKL. Value presented as the

mean ± SDs (n = 5). (b) Si(OH)4 decreased pit numbers and reduced pit areas in differentiated

osteoclasts from mHSCs. Scale bars represent 200 µm. Middle right is quantitative analysis of

pits areas. (c) The specific osteoclastic differentiation markers (NFATc1, DC-STAMP, TRAP, and

CALCR) were assayed by qRT-PCR. Values are presented as the mean ± SDs and are normalized

to GAPDH (n = 3) and representative of three independent experiments, **p < 0.01. Data

presented are from one of three independent experiments, with all three exhibiting comparable

results.

Figure 8. MiR-146a suppressed RANKL-induced osteoclastogenesis in RAW264.7 pre-

osteoclasts. (a,b) TRAP staining 5 days after induction of osteoclatogenesis. The number of

TRAP-positive multinucleated cells was decreased by miR-146a mimic transfection. (c)

26

RAW264.7 cells were transfected with NF-κB-responsive luciferase reporter, and then treated

with miR-146a or Si(OH)4. RANKL was used to induce NF-κB activation in cells and luciferase

activity was quantitated 48 hourspost transfection. Data are expressed as relative light units

(RLU) normalized to total protein (n = 4), **p < 0.01. Data presented are from one of three

independent experiments, with all three exhibiting comparable results.

Scheme 1: Suggested mechanism for the effect of Si(OH)4 on bone cells.The dashed red line

indicates that it is not known whether and how Si(OH)4 enter the cell through Na+-HCO3- co-

transporter (NBC-1).Within the cell Si(OH)4may induce the expression of endogenous miR-146a

to reduce activation of NF- κB. The inhibition of NF-κB results in activation of Runx2, the

master transcription factor necessary for osteoblast precursor differentiation. This deactivation of

NF-κB also inhibits the expression of NFATc1, the key transcription gene for osteoclast

precursor differentiation. Thus, Si(OH)4 can stimulate osteoblast differentiation and inhibit

osteoclast differentiation by antagonizing NF-κB activation via miR-146a, which implies a

potential role in bone remodeling.

27

REFERENCES

[1] Carlisle EM. Silicon: a possible factor in bone calcification. Science 1970;167:279-80.

[2] Carlisle EM. Silicon: an essential element for the chick. Science 1972;178:619-21.

[3] Schwarz K, Milne DB. Growth-promoting effects of silicon in rats. Nature 1972;239:333-4.

[4] Hench LL. Genetic design of bioactive glass. J Eur Ceram Soc 2009;29:1257-65.

[5] Ohura K, Nakamura T, Yamamuro T, Ebisawa Y, Kokubo T, Kotoura Y, et al. Bioactivity of

CaO·SiO2 glasses added with various ions. J Mater Sci: Mater Med 1992;3:95-100.

[6] Li R, Clark AE, Hench LL. An investigation of bioactive glass powders by sol-gel processing.

J Appl Biomater 1991;2:231-9.

[7] Vallet-Regí M, Izquierdo-Barba I, Salinas AJ. Influence of P2O5 on crystallinity of apatite

formed in vitro on surface of bioactive glasses. J Biomed Mater Res 1999;46:560-5.

[8] Xynos ID, Edgar AJ, Buttery LDK, Hench LL, Polak JM. Ionic Products of Bioactive Glass

Dissolution Increase Proliferation of Human Osteoblasts and Induce Insulin-like Growth Factor

II mRNA Expression and Protein Synthesis. Biochem Biophys Res Commun 2000;276:461-5.

[9] Reilly GC, Radin S, Chen AT, Ducheyne P. Differential alkaline phosphatase responses of rat

and human bone marrow derived mesenchymal stem cells to 45S5 bioactive glass. Biomaterials

2007;28:4091-7.

[10] Sayer M, Stratilatov AD, Reid J, Calderin L, Stott MJ, Yin X, et al. Structure and

composition of silicon-stabilized tricalcium phosphate. Biomaterials 2003;24:369-82.

[11] Gibson IR, Best SM, Bonfield W. Effect of Silicon Substitution on the Sintering and

Microstructure of Hydroxyapatite. J Am Ceram Soc 2002;85:2771-7.

[12] Porter AE, Patel N, Skepper JN, Best SM, Bonfield W. Effect of sintered silicate-substituted

hydroxyapatite on remodelling processes at the bone–implant interface. Biomaterials

28

2004;25:3303-14.

[13] Porter AE, Patel N, Skepper JN, Best SM, Bonfield W. Comparison of in vivo dissolution

processes in hydroxyapatite and silicon-substituted hydroxyapatite bioceramics. Biomaterials

2003;24:4609-20.

[14] Pietak AM, Reid JW, Stott MJ, Sayer M. Silicon substitution in the calcium phosphate

bioceramics. Biomaterials 2007;28:4023-32.

[15] Gibson IR, Best SM, Bonfield W. Chemical characterization of silicon-substituted

hydroxyapatite. J Biomed Mater Res 1999;44:422-8.

[16] De Aza PN, Fernández-Pradas JM, Serra P. In vitro bioactivity of laser ablation

pseudowollastonite coating. Biomaterials 2004;25:1983-90.

[17] De Aza PN, Luklinska ZB, Anseau MR, Hector M, Guitián F, De Aza S. Reactivity of a

wollastonite–tricalcium phosphate Bioeutectic® ceramic in human parotid saliva. Biomaterials

2000;21:1735-41.

[18] DeLong RK, Akhtar U, Sallee M, Parker B, Barber S, Zhang J, et al. Characterization and

performance of nucleic acid nanoparticles combined with protamine and gold. Biomaterials

2009;30:6451-9.

[19] Zhang N, Molenda JA, Fournelle JH, Murphy WL, Sahai N. Effects of pseudowollastonite

(CaSiO3) bioceramic on in vitro activity of human mesenchymal stem cells. Biomaterials

2010;31:7653-65.

[20] Zhang N, Molenda JA, Mankoci S, Zhou X, Murphy WL, Sahai N. Crystal structures of

CaSiO3 polymorphs control growth and osteogenic differentiation of human mesenchymal stem

cells on bioceramic surfaces. Biomaterials Science 2013;1:1101-10.

[21] Clark AE, Hench LL, Paschall HA. The influence of surface chemistry on implant interface

29

histology: A theoretical basis for implant materials selection. J Biomed Mater Res 1976;10:161-

74.

[22] Hench LL, Splinter RJ, Allen WC, Greenlee TK. Bonding mechanisms at the interface of

ceramic prosthetic materials. J Biomed Mater Res 1971;5:117-41.

[23] Hench LL. Bioactive Ceramics. Ann N Y Acad Sci 1988;523:54-71.

[24] De Aza PN, Guitián F, De Aza S. Bioeutectic: a new ceramic material for human bone

replacement. Biomaterials 1997;18:1285-91.

[25] Izquierdo-Barba I, Ruiz-González L, Doadrio JC, González-Calbet JM, Vallet-Regí M.

Tissue regeneration: A new property of mesoporous materials. Solid State Sci 2005;7:983-9.

[26] Vallet-Regi M, Ruiz-Gonzalez L, Izquierdo-Barba I, Gonzalez-Calbet JM. Revisiting silica

based ordered mesoporous materials: medical applications. J Mater Chem 2006;16:26-31.

[27] Calvo-Guirado JL, Garces M, Delgado-Ruiz RA, Ramirez Fernandez MP, Ferres-Amat E,

Romanos GE. Biphasic β-TCP mixed with silicon increases bone formation in critical site

defects in rabbit calvaria. Clin Oral Implants Res 2015;26:891-7.

[28] Hench LL, Polak JM, Xynos ID, Buttery LDK. Bioactive materials to control cell cycle.

Mater Res Innovations 2000;3:313-23.

[29] Xynos ID, Hukkanen MVJ, Batten JJ, Buttery LD, Hench LL, Polak JM. Bioglass ®45S5

stimulates osteoblast turnover and enhances bone formation in vitro: implications and

applications for bone tissue engineering. Calcif Tissue Int 2000;67:321-9.

[30] Jugdaohsingh R. Silicon and bone health. J Nutr Health Aging 2007;11:99-110.

[31] Bohner M. Silicon-substituted calcium phosphates - A critical view. Biomaterials

2009;30:6403-6.

[32] Dvorak MM, Siddiqua A, Ward DT, Carter DH, Dallas SL, Nemeth EF, et al. Physiological

30

changes in extracellular calcium concentration directly control osteoblast function in the absence

of calciotropic hormones. Proc Natl Acad Sci U S A 2004;101:5140-5.

[33] Beck GR, Zerler B, Moran E. Phosphate is a specific signal for induction of osteopontin

gene expression. Proc Natl Acad Sci U S A 2000;97:8352-7.

[34] Reffitt DM, Ogston N, Jugdaohsingh R, Cheung HFJ, Evans BAJ, Thompson RPH, et al.

Orthosilicic acid stimulates collagen type 1 synthesis and osteoblastic differentiation in human

osteoblast-like cells in vitro. Bone 2003;32:127-35.

[35] Jugdaohsingh R, Calomme MR, Robinson K, Nielsen F, Anderson SHC, D'Haese P, et al.

Increased longitudinal growth in rats on a silicon-depleted diet. Bone 2008;43:596-606.

[36] Beertsen W, van den Bos T. Alkaline phosphatase induces the mineralization of sheets of

collagen implanted subcutaneously in the rat. J Clin Invest 1992;89:1974-80.

[37] Nuttelman CR, Benoit DS, Tripodi MC, Anseth KS. The effect of ethylene glycol

methacrylate phosphate in PEG hydrogels on mineralization and viability of encapsulated

hMSCs. Biomaterials 2006;27:1377-86.

[38] Zou S, Ireland D, Brooks RA, Rushton N, Best S. The effects of silicate ions on human

osteoblast adhesion, proliferation, and differentiation. Journal of Biomedical Materials Research

Part B: Applied Biomaterials 2009;90B:123-30.

[39] Beck Jr GR, Ha S-W, Camalier CE, Yamaguchi M, Li Y, Lee J-K, et al. Bioactive silica-

based nanoparticles stimulate bone-forming osteoblasts, suppress bone-resorbing osteoclasts, and

enhance bone mineral density in vivo. Nanomed Nanotechnol 2012;8:793-803.

[40] Gaharwar AK, Mihaila SM, Swami A, Patel A, Sant S, Reis RL, et al. Bioactive Silicate

Nanoplatelets for Osteogenic Differentiation of Human Mesenchymal Stem Cells. Adv Mater

2013;25:3329-36.

31

[41] Ha S-W, Weitzmann MN, Beck GR. Bioactive Silica Nanoparticles Promote Osteoblast

Differentiation through Stimulation of Autophagy and Direct Association with LC3 and p62.

ACS Nano 2014;8:5898-910.

[42] Huang D-M, Hung Y, Ko B-S, Hsu S-C, Chen W-H, Chien C-L, et al. Highly efficient

cellular labeling of mesoporous nanoparticles in human mesenchymal stem cells: implication for

stem cell tracking. FASEB J 2005;19:2014-6.

[43] Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, et al. Multilineage

Potential of Adult Human Mesenchymal Stem Cells. Science 1999;284:143-7.

[44] Lian JB, Stein GS, van Wijnen AJ, Stein JL, Hassan MQ, Gaur T, et al. MicroRNA control

of bone formation and homeostasis. Nat Rev Endocrinol 2012;8:212-27.

[45] Inose H, Ochi H, Kimura A, Fujita K, Xu R, Sato S, et al. A microRNA regulatory

mechanism of osteoblast differentiation. Proc Natl Acad Sci U S A 2009;106:20794-9.

[46] Hobert O. Gene Regulation by Transcription Factors and MicroRNAs. Science

2008;319:1785-6.

[47] Pritchard CC, Cheng HH, Tewari M. MicroRNA profiling: approaches and considerations.

Nat Rev Genet 2012;13:358-69.

[48] He L, Hannon GJ. MicroRNAs: small RNAs with a big role in gene regulation. Nat Rev

Genet 2004;5:522-31.

[49] Eskildsen T, Taipaleenmäki H, Stenvang J, Abdallah BM, Ditzel N, Nossent AY, et al.

MicroRNA-138 regulates osteogenic differentiation of human stromal (mesenchymal) stem cells

in vivo. Proc Natl Acad Sci U S A 2011;108:6139-44.

[50] Li Z, Hassan MQ, Jafferji M, Aqeilan RI, Garzon R, Croce CM, et al. Biological Functions

of miR-29b Contribute to Positive Regulation of Osteoblast Differentiation. J Biol Chem

32

2009;284:15676-84.

[51] Vimalraj S, Selvamurugan N. MicroRNAs: Synthesis, Gene Regulation and Osteoblast

Differentiation. Curr Issues Mol Biol 2012;15:7-18.

[52] Iler RK. The chemistry of silica: solubility, polymerization, colloid and surface properties,

and biochemistry. 1979.

[53] Yang S, Wei D, Wang D, Phimphilai M, Krebsbach PH, Franceschi RT. In Vitro and In Vivo

Synergistic Interactions Between the Runx2/Cbfa1 Transcription Factor and Bone

Morphogenetic Protein-2 in Stimulating Osteoblast Differentiation. J Bone Miner Res

2003;18:705-15.

[54] Taganov KD, Boldin MP, Chang K-J, Baltimore D. NF-κB-dependent induction of

microRNA miR-146, an inhibitor targeted to signaling proteins of innate immune responses. Proc

Natl Acad Sci U S A 2006;103:12481-6.

[55] Selvamani SP, Mishra R, Singh SK. Chikungunya Virus Exploits miR-146a to Regulate NF-

κB Pathway in Human Synovial Fibroblasts. PLoS ONE 2014;9:e103624.

[56] Casey WH, Westrich HR, Banfield JF, Ferruzzi G, Arnold GW. Leaching and reconstruction

at the surfaces of dissolving chain-silicate minerals. Nature 1993;366:253-6.

[57] Tossell JA, Sahai N. Calculating the acidity of silanols and related oxyacids in aqueous

solution. Geochim Cosmochim Acta 2000;64:4097-113.

[58] Chang J, Wang Z, Tang E, Fan Z, McCauley L, Franceschi R, et al. Inhibition of osteoblastic

bone formation by nuclear factor-[kappa]B. Nat Med 2009;15:682-9.

[59] Yamazaki M, Fukushima H, Shin M, Katagiri T, Doi T, Takahashi T, et al. Tumor Necrosis

Factor α Represses Bone Morphogenetic Protein (BMP) Signaling by Interfering with the DNA

Binding of Smads through the Activation of NF-κB. J Biol Chem 2009;284:35987-95.

33

[60] Li Y, Li A, Strait K, Zhang H, Nanes MS, Weitzmann MN. Endogenous TNFα Lowers

Maximum Peak Bone Mass and Inhibits Osteoblastic Smad Activation Through NF-κB. J Bone

Miner Res 2007;22:646-55.

[61] Ma X, Becker Buscaglia LE, Barker JR, Li Y. MicroRNAs in NF-κB signaling. J Mol Cell

Biol 2011;3:159-66.

[62] Mladenović Ž, Johansson A, Willman B, Shahabi K, Björn E, Ransjö M. Soluble silica

inhibits osteoclast formation and bone resorption in vitro. Acta Biomater 2014;10:406-18.

[63] Kong Y-Y, Yoshida H, Sarosi I, Tan H-L, Timms E, Capparelli C, et al. OPGL is a key

regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature

1999;397:315-23.

[64] Asagiri M, Takayanagi H. The molecular understanding of osteoclast differentiation. Bone

2007;40:251-64.

[65] Yagi M, Miyamoto T, Sawatani Y, Iwamoto K, Hosogane N, Fujita N, et al. DC-STAMP is

essential for cell–cell fusion in osteoclasts and foreign body giant cells. J Exp Med

2005;202:345-51.

[66] Wiens M, Wang X, Schröder HC, Kolb U, Schloßmacher U, Ushijima H, et al. The role of

biosilica in the osteoprotegerin/RANKL ratio in human osteoblast-like cells. Biomaterials

2010;31:7716-25.

[67] Nielsen FH, Poellot R. Dietary silicon affects bone turnover differently in ovariectomized

and sham-operated growing rats. J Trace Elem Exp Med 2004;17:137-49.

[68] Hott M, de Pollak C, Modrowski D, Marie PJ. Short-term effects of organic silicon on

trabecular bone in mature ovariectomized rats. Calcif Tissue Int 1993;53:174-9.

[69] Khosla S, Amin S, Orwoll E. Osteoporosis in Men. Endocr Rev 2008;29:441-64.

34

[70] Jugdaohsingh R, Tucker KL, Qiao N, Cupples LA, Kiel DP, Powell JJ. Dietary silicon

intake is positively associated with bone mineral density in men and premenopausal women of

the Framingham Offspring cohort. J Bone Miner Res 2004;19:297-307.

[71] Schröder H-C, Perović-Ottstadt S, Rothenberger M, Wiens M, Schwertner H, Batel R, et al.

Silica transport in the demosponge Suberites domuncula: fluorescence emission analysis using

the PDMPO probe and cloning of a potential transporter. Biochem J 2004;381:665-73.

[72] Bhattacharyya P, Volcani BE. Sodium-dependent silicate transport in the apochlorotic

marine diatom Nitzschia alba. Proc Natl Acad Sci U S A 1980;77:6386-90.

35

Figure 1.

36

Figure 2

37

Figure 3

Figure 4

38

Figure 5

Figure 6

39

Figure 7

Figure 8

40

Scheme-1

41

Graphical abstract

SIGNIFICANT STATEMENT

Accumulating evidence over 40 years suggests that silicate is beneficial to bone formation.

However, the biological role(s) of silicate on bone cells are still unclear and controversial. Here,

we report that Si(OH)4, the simplest form of silicate, can stimulate human mesenchymal stem

cells osteoblastic differentiation. We identified that miR-146a is the expression signature in bone

cells treated with Si(OH)4. Further analysis of miR-146a in bone cells reveals that Si(OH)4 up-

regulates miR-146a to antagonize the activation of NF-κB. Si(OH)4 was also shown to de-

activate the same NF-κB pathway to suppress osteoclast formation. Our findings are important to

the development of third-generation cell-and gene affecting biomaterials, and suggest silicate and

miR-146a can be used as pharmaceuticals for bone fracture prevention and therapy.