Orthosilicic Acid Accelerates Bone Formation in Human Osteoblast-LikeCells Through the PI3K–Akt–mTOR Pathway

Hongming Zhou1,2& Guangjun Jiao1

& Meng Dong3& Hai Chi1 & Hongliang Wang1

& Wenliang Wu1& Haichun Liu1

&

Shanwu Ren1& Meng Kong1

& Ci Li1 & Lu Zhang1& Yunzhen Chen1,4

Received: 30 August 2018 /Accepted: 7 November 2018# Springer Science+Business Media, LLC, part of Springer Nature 2018

AbstractSilicon is one of the essential trace elements in the human body; the deficiency of which may lead to bone diseases. Numerous animalexperiments have shown that an appropriate increase in the intake of silicon is beneficial to enhancing bone density and toughness toprevent osteoporosis. However, the molecular mechanisms of the silicon-mediated osteogenesis process have not been sufficientlyclarified. In this study, we determined the possible osteogenesis-related mechanisms of orthosilicic acid at a molecular level. Wedetected the relevant pathway and osteogenic indicators by immunofluorescence (IF),Western blot, alkaline phosphatase (ALP) staining(using 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium [BCIP/NBT]), ALP enzyme labeling method, osteocalcin (OCN),andN-terminal propeptide of type 1 procollagen (P1NP) enzyme-linked immunosorbent assay (ELISA).We found that orthosilicic acidis capable of enhancing the expression of phosphatidylinositol-4,5-bisphosphate 3-kinase (PI3K), phospho-protein kinase B (P-Akt),phospho-mammalian target of rapamycin (P-mTOR), and related osteogenicmarkers (runt-related transcription factor 2 [RUNX2], typeI collagen [COL1], ALP, OCN, and P1NP). However, with the addition of PI3K–Akt–mTOR pathway-specific inhibitor LY294002,the expression of PI3K, P-Akt, P-mTOR, RUNX2, COL1, ALP, OCN, and P1NP decreased. The results indicated that the PI3K–Akt–mTOR pathway played a positive regulatory role in the process of orthosilicic acid–mediated osteogenesis in vitro.

Keywords Orthosilicic acid . Phosphatidylinositol-4,5-bisphosphate 3-kinase (PI3K) . Protein kinaseB (PKB/Akt) .Mammaliantarget of rapamycin (mTOR) . Osteoporosis

Introduction

Osteoporosis is a systemic bone disease characterized by re-duced bone tissue per unit volume and associated structuraldamage, which ultimately increase the risk of brittle fractures.In Europe, the prevalence of osteoporosis is estimated to be5.9–7.2% in men and 19.1–23.5% in women [1]. With society

aging rapidly, osteoporosis is becoming a primary risk factorin the subjectively healthy elderly population, after type 2diabetes and hypertension [2]. Therefore, it is important tostudy the causes and treatment of osteoporosis.

Silicon was believed to be harmful to the human body,particularly because the long-term inhalation of silica (SiO2)dust causes silicosis. However, it was not until the 1970s thatsilicon was understood to be a necessary element for bodilygrowth and development [3, 4]. Many foods including water,beer, coffee, and some vegetables naturally contain silicon [5,6]. Elemental silicon, though available in nature, can beabsorbed by humans and animals only in the form oforthosilicic acid, a substance with good water solubility [7].

It is well known that bone tissue is a hard connective tissuemade up of cells, fibers, andmatrix. The activity of osteoblasts ismainly reflected by the level of alkaline phosphatase (ALP);their fibrous tissue is made up of collagen, and their matrix isformed by large deposits of calcium salts. Previous studies haverepeatedly demonstrated the role of silicon in promoting theupregulation of ALP, type I collagen (COL1), and osteocalcin

* Yunzhen Chenqilucyz@yeah.net

1 Department of Orthopedics, Qilu Hospital of Shandong Universityand Spine and Spinal Cord Disease Research Center, ShandongUniversity, Jinan, Shandong, China

2 Department of Emergency Trauma Surgery, Linyi Central Hospital,Linyi, Shandong Province, China

3 Department of Orthopedics, Zibo Central Hospital, Zibo, ShandongProvince, China

4 Department of Orthopedics, Qilu Hospital of Shandong University,No. 107, Wen Hua Xi Road, Ji’nan 250012, Shandong Province,China

Biological Trace Element Researchhttps://doi.org/10.1007/s12011-018-1574-9

(OCN) markers in vitro and in vivo [8–12]. N-terminalpropeptide of type 1 procollagen (P1NP) can be detected inserum and is of great significance for guiding the treatment ofosteoporosis. Type 1 procollagen is COL1 precursor, which isdecomposed from both ends of the molecule by protease hydro-lysis to form two procollagen peptides, namely P1NP and P1CP[13]. Therefore, P1NP is used as an indicator of bone formation.The runt-related transcription factor (RUNX2), also known ascore-binding factor subunit alpha-1 (CBFα1) or AML3, is atranscription factor of the RUNX family. Previous studies havefound that RUNX2 plays an essential role in osteogenesis [14,15] by regulating the expression of various bone-related genesand is necessary for bone formation in vivo. Therefore, manyresearchers regard RUNX2 as an indicator of osteogenesis [16,17]. In particular, our team has found that orthosilicic acid canenhance the expression of ALP, COL1, and OCN through thebone morphogenetic protein 2 (BMP-2)/Smad/RUNX2 signal-ing pathway in vitro [9], demonstrating one of the osteogenicmechanisms of orthosilicic acid at the pathway level.

The phosphatidylinositol-4,5-bisphosphate 3-kinase (PI3K)–protein kinase B (PKB/Akt)–mammalian target of rapamycin(mTOR) pathway was first discovered in tumor cells and wasfound to be associated with multiple functions of such cells. Aktis a pivotal molecule downstream of the PI3-kinase signalingpathway. Many cytokines and growth factors lead to the activityof PI3-kinase, resulting in increases of PI(3,4)P2 and PI(3,4,5)P3in cells; both of which are required for the activation ofAkt. BothPIP2 and PIP3 bind directly to the pleckstrin homology domainof Akt to activate it. mTOR, a downstream effector of Akt, is aserine/threonine protein kinase composed of two different com-plexes (mTORC1 and mTORC2). The activation of mTORleads to metabolism of sugar, fat, and protein by promotingmRNA translocation. Several previous studies have explainedthat the PI3K–Akt–mTOR pathway is involved in the processof osteogenesis [16, 18, 19].

In conclusion, orthosilicic acid has been proven to have adefinite osteogenic effect. Meanwhile, the PI3K–Akt–mTORpathway has been proven many times to participate in the osteo-blast differentiation process. However, it is not clear whetherorthosilicic acid promotes osteogenesis through other non-classical BMP pathways, such as PI3K–Akt–mTOR. Therefore,this study aimed to discover the relationship between orthosilicicacid and the PI3K–Akt–mTOR pathway in terms of the promo-tion of osteogenesis, in order to provide a greater theoretical basisand new targets for the treatment of osteoporosis with silicon.

Materials and Methods

Cell Culture

In this study, we used osteoblast-like cell lines MG-63 andU2-OS as models of osteoblasts. Both of them were derived

from human osteosarcomas [8, 9, 20–23]. We cultured bothtypes of cells in Dulbecco’s Modified Eagle’s Medium(DMEM; HyClone, South Logan, UT, USA) with 10% fetalbovine serum (FBS; GIBCO brand, Thermo Fisher Scientific,Waltham, MA, USA) and 1% penicillin-streptomycin(100 mL; Beyotime Institute of Biotechnology, Nanjing,China) in a humidified atmosphere of 5% CO2 at 37 °C.

Addition and Measurement of Silicon

We prepared 1.0-mM stock solution of orthosilicic acid fromsodium silicate (1 L; Sigma–Aldrich Company Ltd.,Gillingham, UK) using ultra high purity water. The stock so-lution was filtered with a strainer (33 mm, 0.22 μm; EMDMillipore, Billerica, MA, USA) to ensure the sterility of thestock solution. Then, the sterile stock solution was added toserum-free DMEM at concentrations of 10 μM. The finalconcentrations of silicon in the culture medium were mea-sured in duplicate wells by inductively coupled plasma opticalemission spectrometry (ICP-OES; PerkinElmer, Waltham,MA, USA). Finally, the cell lines MG-63 and U2-OS werewashed twice with polybutylene succinate (PBS) andswitched to serum-free DMEM with either vehicle or solubleSi (10 μM) [8, 9].

Cellular Immunofluorescence

For the immunofluorescence (IF) assay, we cultured MG-63and U2-OS cells on Millicell EZ SLIDE 8-well glass slides(EMD Millipore), then incubated them with orthosilicic acid(0 and 10 μM for MG-63 and U2-OS, respectively) for 72 h.The cells were then incubated with primary antibodies againstP-Akt (1:400; Cell Signaling Technology [CST], Danvers,MA, USA) and P-mTOR (1:50; CST) overnight at 4 °C. Wewashed the cells three times with PBS and then incubatedthem with fluorescein isothiocyanate (FITC)–conjugated goatanti-rabbit immunoglobulin G (IgG; 1:200; AffinityBiosciences, Jiangsu, China) for 1 h at room temperature(RT). Then we added 4′,6-diamidino-2-phenylindole (DAPI;Beijing Solarbio Science & Technology Co., Ltd., Beijing,China), and after 5 min in the dark, we washed the DAPI awaywith PBS. We then added mounting medium (BeijingSolarbio) to the slide. Finally, we imaged the cells via IFmicroscopy (Olympus Corp., Tokyo, Japan) and analyzedthe images using Image-Pro Plus software version 6.0(Media Cybernetics, Inc., Rockville, MD, USA) to calculatethe ratio of IOD (sum)/area (sum) under the same measure-ment conditions.

Western Blot Assays

We plated MG-63 and U2-OS cells in 6-well plates and stim-ulated them with orthosilicic acid (0 and 10 μM, respectively)

Zhou et al.

for 72 h and with PI3K–Akt–mTOR pathway–specific inhib-itor LY294002 (20 μM; MedChemExpress, MonmouthJunction, NJ, USA) for the last 48 h of this period. We thenwashed the cells three times with cold PBS, extracted cellprotein with radioimmunoprecipitation assay (RIPA) lysisbuffer (Beyotime), and centrifuged them at 15,000 rpm for20 min at 4 °C. We measured protein concentrations using abicinchoninic-acid (BCA) protein assay kit (Beyotime) and anxMark Microplate Absorbance Spectrophotometer (Bio-Rad,Hercules, CA, USA). Then we mixed equal amounts of pro-tein samples with 5 × sodium dodecyl sulfate-polyacrylamidegel electrophoresis (SDS-PAGE) loading buffer (Beyotime)and heated them at 95 °C for 10 min. We separated a sampleof 20 μg protein per well on 10% SDS-PAGE gel (Beyotime)and transferred them to polyvinylidene difluoride (PVDF)membranes (EMD Millipore). After blocking in 5% nonfatmilk in tris-buffered saline containing 0.1% Tween20(TBST) for at least 1 h, we probed the blots overnight at4 °C with primary antibodies. The blots were subsequentlywashed three times in TBST, incubated with appropriate sec-ondary antibodies for 1 h at RT, and washed again prior tosignal detection with a ChemiDoc Touch Gel Imaging Systemand Image Lab Touch software (Bio-Rad). Antibodies used inthis study were Akt (1:1000; CST), mTOR (1:1000; CST), P-Akt (1:2000; CST), P-mTOR (1:1000; CST), PI3K (1:1000;Abcam PLC, Cambridge, UK), RUNX2 (1:1000; Abcam),COL1 (1:1000; Abcam), and β-actin (1:1000; ZSGB-BIOCorp., Beijing, China).

Alkaline Phosphatase Staining

We plated MG-63 and U2-OS cells in 6-well plates with ap-proximately 90% uniform cell coverage per well. The nextday, we replaced the DMEM with two concentrations oforthosilicic acid (0 and 10 μM forMG-63 and U2-OS, respec-tively) for 72 h and with LY294002 (20 μM) for the last 48 hof this period. After 72 h, we washed the cells twice with PBSand fixed them with 4% paraformaldehyde for 30 min. Thenwe removed the paraformaldehyde and rinsed the cells threetimes with deionized water (5 min each time). Afterward, weadded 1 mL 5-bromo-4-chloro-3-indolyl phosphate/nitro bluetetrazolium (BCIP/NBT) color solution (BCIP/NBTAlkalinePhosphatase Color Development Kit; Beyotime) per well for1 h in the dark at RT. The reaction was terminated by quicklywashing the cells twice with deionized water. After acquiringthe images by microscopy (Olympus), we analyzed them withImage-Pro Plus software to calculate the IOD of a single sam-ple under the same measurement conditions.

Alkaline Phosphatase Assay

We grew MG-63 and U2-OS cells in 6-well plates and incu-bated them with two concentrations of orthosilicic acid (0 and

10 μM, respectively) for 7 days and with LY294002 (20 μM)for the last 2 days of this period. Next, we washed the cellstwice with cold PBS and lysed them with 1% Triton X-100(Solaibao Co., Beijing, China), then measured protein concen-tration via BCA assay (Beyotime). We determined ALP activ-ity in cells using the ALP assay kit (Jiancheng BioengineeringInstitute, Nanjing, China). First, we added 30 μL of cell lysateplus reaction buffers to each hole in 96-well plates and incu-bated the mixture for 15 min at 37 °C. Afterward, we mixedthe samples with 150 μL chromogenic agent from the ALPassay kit. We determined absorbance optical density (OD) ofthe individual wells using an xMark Microplate AbsorbanceSpectrophotometer (Bio-Rad) with wavelength modulated to520 nm. Finally, we calculated ALP activity in each group ofcells according to the formula in the instruction manual for theALP assay kit.

Enzyme-Linked Immunosorbent Assay

We grew MG-63 and U2-OS cells in 6-well plates (1.5 × 106)and stimulated them with two concentrations of orthosilicicacid (0 and 10 μM for MG-63 and U2-OS, respectively) for72 h and with LY294002 (20 μM) for the last 48 h of thisperiod. We then collected the supernatant after ending thestimulus. We determined OCN and P1NP concentration usingHuman Osteocalcin and P1NP ELISA Kit (ShanghaiEnzyme-linked Biotechnology Co., Ltd., Shanghai, China)per manufacturer’s instructions.

Statistical Analyses

The results were representative of experiments repeated atleast three times each. We determined differences by one-way analysis of variance (ANOVA) using Student’s t test asappropriate. The software we used for our analyses wasGraphPad Prism, version 6.0c (GraphPad Software, Inc.,San Diego, CA, USA). Results were expressed as the mean± standard deviation (SD), and P < 0.05 was considered sta-tistically significant.

Results

Silicon Concentration in Culture Medium

The baseline concentration of silicon in the serum-freeDMEM which was kept in the plastic containers was 1.46 ±0.8 μM (mean ± SD). Following the addition of 10 μMorthosilicic acid, the final silicon concentration measured intriplicate was 10.3 ± 1.8 μM.

Orthosilicic Acid Accelerates Bone Formation in Human Osteoblast-Like Cells Through the PI3K–Akt–mTOR...

The Effect of Orthosilicic Acid on P-Akt and P-mTORExpression in MG-63 and U2-OS Cells as Determinedby Immunofluorescence

Because it is not clear whether the PI3K–Akt–mTOR pathwayis changed when orthosilicic acid promotes osteogenesisin vitro, we used IF to detect fluorescent P-Akt and P-mTOR in MG-63 and U2-OS cells that had been exposed toorthosilicic acid for 72 h. As shown in Fig. 1a, there was anincrease in P-Akt and P-mTOR fluorescence in MG-63 andU2-OS cells compared with control cells. Then, using Image-Pro Plus software to analyze mean fluorescence intensity inboth cell lines, we found that relative fluorescence intensity ofthe Si group was significantly greater than that of the controlgroup (Fig. 1b; **P < 0.01 vs. control group).

Orthosilicic Acid and LY294002 Had Opposite Effectson Osteogenesis Via the PI3K–Akt–mTOR Pathwayin MG-63 and U2-OS Cells

As shown in Fig. 2, expression of PI3K, P-Akt, P-mTOR, andosteogenic indicators (COL1 and RUNX2) was significantlyincreased by stimulation with orthosilicic acid for 72 h com-pared with the control group. When we added LY294002 forthe last 48 h of this period, the abovementioned expressionwas significantly inhibited compared with the control and Sigroups. It is worth mentioning that expression of P-Akt and P-mTOR showed significant changes, while that of Akt and

mTOR did not change markedly (*P < 0.05 vs. control group;**P < 0.01 vs. control group; ##P < 0.01 vs. Si group).

The Effect of Orthosilicic Acid and LY294002on Alkaline Phosphatase Staining

BCIP/NBT is a common substrate for ALP. ALP catalysis hy-drolyzes BCIP to produce strong reactive products, which reactwith NBT to form insoluble blue-violet NBT-formazan. Asshown in Fig. 3a, this stainingmethod showed that after stimuluswith orthosilicic acid for 72 h, the number of cells with ALPstaining increased significantly in both cell lines. After stimuluswith LY294002 for the last 48 h of this period, these numbersdecreased. Using Image-Pro Plus software to analyze the meanintensity of ALP staining in both cell lines, we found that relativestaining intensity of the four groups was statistically significant(Fig. 3b; **P < 0.01 vs. control group; ##P < 0.01 vs. Si group).

The Effect of Orthosilicic Acid and LY294002on Alkaline Phosphatase Activity

ALP decomposes phenyldisodium phosphate and producesfree phenol and phosphoric acid. Phenol reacts with 4-amino-aminoantipyrine in alkaline solution and oxidizes bypotassium ferrocyanide to form red-quinone derivatives.Enzyme activity can be determined by the depth of redness.To further investigate the effect of prolonged orthosilicic acidexposure on ALP activity in both cell lines, we measured ALP

Fig. 1 IF analysis of P-Akt and P-mTOR expression in MG-63 and U2-OS cells. aNumber of P-Akt and P-mTOR cells increased in the Si groupcompared with the control group. Scale bar, 20 μm. b Relative

fluorescence intensity in the Si group was significantly greater than thatin the control group (**P < 0.01 vs. control group; n = 3)

Zhou et al.

activity using the standard ALP enzyme method. As shown inFig. 4, ALP activity obviously increased after exposure toorthosilicic acid for 7 days and decreased after exposure toLY294002 for the last 2 days of this period. These results,combined with those of ALP staining, indicated that ALPexpression increased after 72 h and 7 days of orthosilicic acidstimulation. This means that ALP is an important early- andmiddle-stage indicator of osteogenesis (*P < 0.05 vs. controlgroup; **P < 0.01 vs. control group; ##P < 0.01 vs. Si group).

The Effect of Orthosilicic Acid and LY294002 on OCNand P1NP

OCN and P1NP are secretory proteins produced and secretedby osteoblasts. We collected cell supernatants and determined

OCN and P1NP expression therein via the ELISA method. Asshown in Fig. 5a, b, orthosilicic acid for 72 h promoted OCNand P1NP expression, and LY294002 for the last 48 h of thisperiod reduced OCN and P1NP expression in both cell super-natants (*P < 0.05 vs. control group; **P < 0.01 vs. controlgroup; ##P < 0.01 vs. Si group).

Discussion

Osteoporosis is a metabolic disease that cannot only decreasebone strength via deterioration in bone mass and quality butcan also increase the individual risk of fracture [24–26]. Bonemass and quality are affected by many factors, such as nutri-tion, physical activity, smoking, drinking, and genetics.

Fig. 2 Western blot analysis forthe expression of the PI3K–Akt–mTOR pathway and that ofosteogenic indicators (COL1 andRUNX2) in MG-63 and U2-OScells. Such expressionsignificantly increased afterstimulation with 10 μMorthosilicic acid for 72 h,compared with the control group.After the addition of LY294002for the last 48 h of this period, theexpression of PI3K, P-Akt, P-mTOR, and osteogenic indicators(COL1 and RUNX2) decreasedsignificantly when compared withthe control or Si group (*P < 0.05vs. control group; **P < 0.01 vs.control group; ##P < 0.01 vs. Sigroup; n = 3)

Orthosilicic Acid Accelerates Bone Formation in Human Osteoblast-Like Cells Through the PI3K–Akt–mTOR...

Osteoporosis is more common in postmenopausal women andolder men [27]. Currently, 10 million individuals are diag-nosed with osteoporosis in the USA, and more than half ofthe US population > 50 years is found to have low bone den-sity [28]. Meanwhile, about 2 million people in the USA havefractures due to osteoporosis [26]. In 2005, fragile fractures inthe USA caused a loss of $19 billion; unfortunately, the bur-den of osteoporosis in the USA is expected to increase bynearly 50% by 2025, with more than 3 million related frac-tures at an annual cost of about US $253 billion [28, 29].

Previous studies have shown that bioactive glass [30, 31]and ceramic [32, 33] have been proven to promote bone for-mation, but the components are too complicated to verify that

silicon elements play a definite role in the abovementionedprocess. However, bioactive glass and ceramic research hasshown that the silicate in the material can be released asorthosilicic acid [Si(OH)4], which has been repeatedly provento promote osteogenesis [8, 9, 34]. Our research differs fromthe abovementioned studies in that we used a solvent oforthosilicic acid more conducive to cell absorption as a stim-ulus. Our results will provide a theoretical basis for the sub-sequent study of oral-related silicon formulations for animalexperiments.

ALP is an indicator of bone formation and can regulatebone matrix mineralization [35]. It is often used as an earlymarker of osteogenic differentiation [36, 37]. OCN is present

Fig. 3 BCIP/NBTALP stainingmethod analysis for ALPexpression in MG-63 and U2-OScells. a Number of cells stainedpositive for ALP increasedsignificantly after stimulationwith orthosilicic acid for 72 h anddecreased after stimulation withLY294002 for the last 48 h of thisperiod. Scale bar, 500 μm. bRelative staining intensity of ALPincreased due to the action oforthosilicic acid and decreaseddue to inhibition by LY294002(**P < 0.01 vs. control group;##P < 0.01 vs. Si group; n = 3)

Fig. 4 Standard ALP enzymemethod for analyzing ALPactivity in MG-63 and U2-OScells. ALP activity obviouslyincreased after stimulation withorthosilicic acid for 7 days anddecreased after stimulation withLY294002 for the last 2 days ofthis period (*P < 0.05 vs. controlgroup; **P < 0.01 vs. controlgroup; ##P < 0.01 vs. Si group;n = 3)

Zhou et al.

during the late stages of osteoblast differentiation as a specificmarker of osteogenic differentiation, regulating matrix miner-alization [38]. RUNX2 is a key regulator of osteogenesis dur-ing osteogenic differentiation by stimulating osteoblast differ-entiation gene transcription [39, 40]. Meanwhile, previousstudies have found that RUNX2 is located upstream of OCNand plays a positive role in regulating it [41, 42]. As an organicsubstance in bones, collagen plays a very important role inbone toughness and is also an important indicator in the pro-cess of osteogenesis [43, 44]. In addition, P1NP is related tothe production of COL1 [45]. Previous studies by our teamhave found that orthosilicic acid has a significant osteogeniceffect on osteoblasts [8, 9, 34]. The current study also hasreached the same conclusion. Osteogenesis indicators, includ-ing RUNX2, COL1, ALP, OCN, and P1NP, all showed anincreased expression after the addition of orthosilicic acid(Figs. 2, 3, 4, and 5).

The PI3K–Akt–mTOR pathway plays a significant role inregulating a variety of cellular functions, including cell sur-vival, proliferation, differentiation, and angiogenesis [46–49].It has been found that the expression of P-Akt, P-mTOR, andosteogenesis indicators increased in different degrees whenexposed to different stimulants and depending on which stemcells they acted upon [16, 18, 19]. Our study showed the same

result, except that we used human osteoblast-like cells. Afteradding the optimal concentration of orthosilicic acid, we de-tected the fluorescence density of P-Akt and P-mTOR by IFand found that P-Akt and P-mTOR showed higher fluores-cence intensity in the Si group than in the control group(Fig. 1). The result indicated that orthosilicic acid promotedosteogenesis and increased P-Akt and P-mTOR expression.Therefore, we added the inhibitor LY294002 [50] to studythe relationship between the PI3K–Akt–mTOR signalingpathway and the osteogenic effects of orthosilicic acid. Wedivided cells into four groups, treated them with orthosilicicacid and LY294002, and used Western blotting to detect thechanges in the PI3K–Akt–mTOR pathway and osteogenicindicators (RUNX2 and COL1). We found that orthosilicicacid increased the expression of the PI3K–Akt–mTOR path-way and that of osteogenesis indicators (RUNX2, COL1,ALP, OCN, and P1NP), while LY294002 led to a decreasein such expression (Figs. 2, 3, 4, and 5). The results indicatedthat orthosilicic acid positively regulated the osteogenesis pro-cess by promoting the expression of the PI3K–Akt–mTORpathway in vitro. That is, the PI3K–Akt–mTOR pathwayplays a significant role upstream of the osteogenesis process.In addition, we found that orthosilicic acid caused a significantincrease in P-Akt and P-mTOR, while Akt and mTOR did not

Fig. 5 ELISA analysis of OCN(a) and P1NP (b) expression inMG-63 and U2-OS cellsupernatants. In bothsupernatants, exposure toorthosilicic acid for 72 hpromoted OCN and P1NPexpression, and exposure toLY294002 for the last 48 h of thisperiod reduced OCN and P1NPexpression (*P < 0.05 vs. controlgroup; **P < 0.01 vs. controlgroup; ##P < 0.01 vs. Si group;n = 3)

Orthosilicic Acid Accelerates Bone Formation in Human Osteoblast-Like Cells Through the PI3K–Akt–mTOR...

show significant changes (Fig. 2). This is consistent with theresults of Li G et al. [18], but the difference is that we foundPI3K also showed a consistent trend with P-Akt and P-mTOR(Fig. 2). The difference may be related to stimulators, celltypes, and experimental conditions.

In conclusion, our results indicated that orthosilicic acidactivated the PI3K–Akt–mTOR pathway to promote osteo-genesis in vitro; when this pathway is inhibited, osteogenesiscaused by orthosilicic acid is significantly attenuated. Thefinding suggested that the PI3K–Akt–mTOR pathway con-tributed to the regulation of orthosilicic acid–mediated osteo-genesis in vitro. This study may provide clues to discoveringnew ways to treat osteoporosis.

Funding Information This study was supported by two grants from theDepartment of Science and Technology of Shandong Province (Grant No.2017GSF18160) and National Natural Science Foundation of China(NSFC) (Grant No. 81602361).

Compliance with Ethical Standards

Conflicts of Interest The authors declare that they have no conflict ofinterest.

References

1. Hernlund E, Svedbom A, Ivergard M et al (2013) Osteoporosis inthe European Union: medical management, epidemiology and eco-nomic burden. A report prepared in collaboration with theInternational Osteoporosis Foundation (IOF) and the EuropeanFederation of Pharmaceutical Industry Associations (EFPIA).Arch Osteoporos 8:136

2. Gutzwiller JP, Richterich JP, Stanga Z, Nydegger UE, Risch L,RischM (2018) Osteoporosis, diabetes, and hypertension are majorrisk factors for mortality in older adults: an intermediate report on aprospective survey of 1467 community-dwelling elderly healthypensioners in Switzerland. BMC Geriatr 18:115

3. Carlisle EM (1970) Silicon: a possible factor in bone calcification.Science 167:279–280

4. Schwarz K, Milne DB (1972) Growth-promoting effects of siliconin rats. Nature 239:333–334

5. Powell JJ, McNaughton SA, Jugdaohsingh R et al (2005) A provi-sional database for the silicon content of foods in the UnitedKingdom. Br J Nutr 94:804–812

6. Jugdaohsingh R, Anderson SH, Tucker KL et al (2002) Dietarysilicon intake and absorption. Am J Clin Nutr 75:887–893

7. Reffitt DM, Jugdaohsingh R, Thompson RP et al (1999) Silicicacid: its gastrointestinal uptake and urinary excretion in man andeffects on aluminium excretion. J Inorg Biochem 76:141–147

8. Reffitt DM, Ogston N, Jugdaohsingh R, Cheung HFJ, Evans BAJ,Thompson RPH, Powell JJ, Hampson GN (2003) Orthosilicic acidstimulates collagen type 1 synthesis and osteoblastic differentiationin human osteoblast-like cells in vitro. Bone 32:127–135

9. Dong M, Jiao G, Liu H, Wu W, Li S, Wang Q, Xu D, Li X, Liu H,Chen Y (2016) Biological silicon stimulates collagen type 1 andosteocalcin synthesis in human osteoblast-like cells through theBMP-2/Smad/RUNX2 signaling pathway. Biol Trace Elem Res173:306–315

10. Mieszawska AJ, Nadkarni LD, Perry CC, Kaplan DL (2010)Nanoscale control of silica particle formation via silk-silica fusionproteins for bone regeneration. Chem Mater 22:5780–5785

11. Varanasi VG, Leong KK, Dominia LM, Jue SM, Loomer PM,Marshall GW (2012) Si and Ca individually and combinatoriallytarget enhanced MC3T3-E1 subclone 4 early osteogenic markerexpression. J Oral Implantol 38:325–336

12. Shi M, Xia L, Chen Z, Lv F, Zhu H,Wei F, Han S, Chang J, Xiao Y,Wu C (2017) Europium-doped mesoporous silica nanosphere as animmune-modulating osteogenesis/angiogenesis agent. Biomaterials144:176–187

13. Brown JE, Sim S (2010) Evolving role of bone biomarkers incastration-resistant prostate cancer. Neoplasia 12:685–696

14. Komori T, Yagi H, Nomura S et al (1997) Targeted disruption ofCbfa1 results in a complete lack of bone formation owing to mat-urational arrest of osteoblasts. Cell 89:755–764

15. Otto F, Thornell AP, Crompton T, Denzel A, Gilmour KC, RosewellIR, Stamp GWH, Beddington RSP, Mundlos S, Olsen BR, SelbyPB, Owen MJ (1997) Cbfa1, a candidate gene for cleidocranialdysplasia syndrome, is essential for osteoblast differentiation andbone development. Cell 89:765–771

16. Luo G, Xu B, Huang Y (2017) Icariside II promotes the osteogenicdifferentiation of canine bone marrow mesenchymal stem cells viathe PI3K/AKT/mTOR/S6K1 signaling pathways. Am J Transl Res9:2077–2087

17. Liu X, Wang Y, Cao Z, Dou C, Bai Y, Liu C, Dong S, Fei J (2017)Staphylococcal lipoteichoic acid promotes osteogenic differentia-tion of mouse mesenchymal stem cells by increasing autophagicactivity. Biochem Biophys Res Commun 485:421–426

18. Li G, Hu J, Chen H, Chen L, Zhang N, Zhao L, Wen N, Yang Y(2017) Enamel matrix derivative enhances the proliferation andosteogenic differentiation of human periodontal ligament stem cellson the titanium implant surface. Organogenesis 13:103–113

19. Tong Y, Feng W, Wu Y, Lv H, Jia Y, Jiang D (2015) Mechano-growth factor accelerates the proliferation and osteogenic differen-tiation of rabbit mesenchymal stem cells through the PI3K/AKTpathway. BMC Biochem 16:1

20. Lajeunesse D, Kiebzak GM, Frondoza C, Sacktor B (1991) Regulationof osteocalcin secretion by human primary bone cells and by the humanosteosarcoma cell line MG-63. Bone Miner 14:237–250

21. Lajeunesse D, Frondoza C, Schoffield B et al (1990) Osteocalcinsecretion by the human osteosarcoma cell line MG-63. J BoneMiner Res 5:915–922

22. Orimo H, Goseki-Sone M, Hosoi T, Shimada T (2008) Functionalassay of the mutant tissue-nonspecific alkaline phosphatase geneusing U2OS osteoblast-like cells. Mol Genet Metab 94:375–381

23. Niu X, Chen Y, Qi L, Liang G, Wang Y, Zhang L, Qu Y, Wang W(2018) Hypoxia regulates angeogenic-osteogenic coupling processvia up-regulating IL-6 and IL-8 in human osteoblastic cells throughhypoxia-inducible factor-1alpha pathway. Cytokine. https://doi.org/10.1016/j.cyto.2018.06.022

24. Bessette L, Ste-Marie LG, Jean S, Shawn Davison K, Beaulieu M,Baranci M, Bessant J, Brown JP (2008) Recognizing osteoporosisand its consequences in Quebec (ROCQ): background, rationale,and methods of an anti-fracture patient health-management pro-gramme. Contemp Clin Trials 29:194–210

25. Diab DL,Watts NB (2013) Diagnosis and treatment of osteoporosisin older adults. Endocrinol Metab Clin N Am 42:305–317

26. Bukata SV, Digiovanni BF, Friedman SM et al (2011) A guide toimproving the care of patients with fragility fractures. GeriatrOrthop Surg Rehabil 2:5–37

27. Khosla S, Amin S, Orwoll E (2008) Osteoporosis in men. EndocrRev 29:441–464

28. Ekman EF (2010) The role of the orthopaedic surgeon in minimiz-ingmortality andmorbidity associated with fragility fractures. J AmAcad Orthop Surg 18:278–285

Zhou et al.

29. Burge R, Dawson-Hughes B, Solomon DH, Wong JB, King A,Tosteson A (2007) Incidence and economic burden ofosteoporosis-related fractures in the United States, 2005-2025. JBone Miner Res 22:465–475

30. Xynos ID, Edgar AJ, Buttery LD et al (2000) Ionic products ofbioactive glass dissolution increase proliferation of human osteoblastsand induce insulin-like growth factor II mRNA expression and pro-tein synthesis. Biochem Biophys Res Commun 276:461–465

31. Reilly GC, Radin S, Chen AT, Ducheyne P (2007) Differentialalkaline phosphatase responses of rat and human bone marrowderived mesenchymal stem cells to 45S5 bioactive glass.Biomaterials 28:4091–4097

32. Zhang N, Molenda JA, Fournelle JH, MurphyWL, Sahai N (2010)Effects of pseudowollastonite (CaSiO3) bioceramic on in vitro ac-tivity of human mesenchymal stem cells. Biomaterials 31:7653–7665

33. Zhang N, Molenda JA, Mankoci S, Zhou X, Murphy WL, Sahai N(2013) Crystal structures of CaSiO3 polymorphs control growthand osteogenic differentiation of human mesenchymal stem cellson bioceramic surfaces. Biomater Sci 1:1101–1110

34. Zhou X, Moussa FM, Mankoci S, Ustriyana P, Zhang N,Abdelmagid S, Molenda J, Murphy WL, Safadi FF, Sahai N(2016) Orthosilicic acid, Si(OH)4, stimulates osteoblast differenti-ation in vitro by upregulating miR-146a to antagonize NF-kappaBactivation. Acta Biomater 39:192–202

35. Hessle L, Johnson KA, Anderson HC, Narisawa S, Sali A, GodingJW, Terkeltaub R, Millan JL (2002) Tissue-nonspecific alkalinephosphatase and plasma cell membrane glycoprotein-1 are centralantagonistic regulators of bone mineralization. Proc Natl Acad SciU S A 99:9445–9449

36. Cao FY, Fan JX, Long Y, Zeng X, Zhang XZ (2016) A smartfluorescence nanoprobe for the detection of cellular alkaline phos-phatase activity and early osteogenic differentiation. Nanomedicine12:1313–1322

37. Anderson HC, Sipe JB, Hessle L, Dhamyamraju R, Atti E,Camacho NP, Millán JL (2004) Impaired calcification around ma-trix vesicles of growth plate and bone in alkaline phosphatase-deficient mice. Am J Pathol 164:841–847

38. Zappitelli T, Chen F, Aubin JE (2015) Up-regulation of BMP2/4signaling increases both osteoblast-specific marker expression and

bone marrow adipogenesis in Gja1Jrt/+ stromal cell cultures. MolBiol Cell 26:832–842

39. Kwun IS, Cho YE, Lomeda RA et al (2010) Zinc deficiency sup-presses matrix mineralization and retards osteogenesis transientlywith catch-up possibly through Runx 2 modulation. Bone 46:732–741

40. Kumar Y, Kapoor I, Khan K, Thacker G, Khan MP, Shukla N,Kanaujiya JK, Sanyal S, Chattopadhyay N, Trivedi AK (2015) E3ubiquitin ligase Fbw7 negatively regulates osteoblast differentiationby targeting Runx2 for degradation. J Biol Chem 290:30975–30987

41. Schroeder TM, Jensen ED, Westendorf JJ (2005) Runx2: a masterorganizer of gene transcription in developing and maturing osteo-blasts. Birth Defects Res C Embryo Today 75:213–225

42. Ziros PG, Basdra EK, Papavassiliou AG (2008) Runx2: of boneand stretch. Int J Biochem Cell Biol 40:1659–1663

43. Burr DB (2002) The contribution of the organic matrix to bone’smaterial properties. Bone 31:8–11

44. Wang X, Bank RA, TeKoppele JM et al (2001) The role of collagenin determining bone mechanical properties. J Orthop Res 19:1021–1026

45. Takahashi S (2013) Bone metabolic markers for evaluation of bonemetastases. Clin Calcium 23:391–400

46. Bertacchini J, Heidari N, Mediani L, Capitani S, Shahjahani M,Ahmadzadeh A, Saki N (2015) Targeting PI3K/AKT/mTOR net-work for treatment of leukemia. Cell Mol Life Sci 72:2337–2347

47. Follo MY, Manzoli L, Poli A, McCubrey JA, Cocco L (2015) PLCand PI3K/Akt/mTOR signalling in disease and cancer. Adv BiolRegul 57:10–16

48. Manfredi GI, Dicitore A, Gaudenzi G et al (2015) PI3K/Akt/mTORsignaling in medullary thyroid cancer: a promising molecular targetfor cancer therapy. Endocrine 48:363–370

49. Guo X, Wei S, Lu M, Shao Z, Lu J, Xia L, Lin K, Zou D (2016)Dose-dependent effects of strontium ranelate on ovariectomy ratbone marrow mesenchymal stem cells and human umbilical veinendothelial cells. Int J Biol Sci 12:1511–1522

50. Gharbi SI, Zvelebil MJ, Shuttleworth SJ, Hancox T, Saghir N,Timms JF, Waterfield MD (2007) Exploring the specificity of thePI3K family inhibitor LY294002. Biochem J 404:15–21

Orthosilicic Acid Accelerates Bone Formation in Human Osteoblast-Like Cells Through the PI3K–Akt–mTOR...