Skip to main content
NCBI home page
Science and Technology of Advanced Materials logoLink to Science and Technology of Advanced Materials
. 2009 Jul 10;10(2):025003. doi: 10.1088/1468-6996/10/2/025003

Porous hydroxyapatite and biphasic calcium phosphate ceramics promote ectopic osteoblast differentiation from mesenchymal stem cells

Lingli Zhang 1,2, Nobutaka Hanagata 2,3,4,✉, Megumi Maeda 3, Takashi Minowa 3, Toshiyuki Ikoma 2,4, Hongsong Fan 1, Xingdong Zhang 1
PMCID: PMC5090447  PMID: 27877290

Abstract

Because calcium phosphate (Ca–P) ceramics have been used as bone substitutes, it is necessary to investigate what effects the ceramics have on osteoblast maturation. We prepared three types of Ca–P ceramics with different Ca–P ratios, i.e. hydroxyapatite (HA), beta-tricalcium phosphate (β-TCP), and biphasic calcium phosphate (BCP) ceramics with dense-smooth and porous structures. Comprehensive gene expression microarray analysis of mouse osteoblast-like cells cultured on these ceramics revealed that porous Ca–P ceramics considerably affected the gene expression profiles, having a higher potential for osteoblast maturation. In the in vivo study that followed, porous Ca–P ceramics were implanted into rat skeletal muscle. Sixteen weeks after the implantation, more alkaline-phosphatase-positive cells were observed in the pores of hydroxyapatite and BCP, and the expression of the osteocalcin gene (an osteoblast-specific marker) in tissue grown in pores was also higher in hydroxyapatite and BCP than in β-TCP. In the pores of any Ca–P ceramics, 16 weeks after the implantation, we detected the expressions of marker genes of the early differentiation stage of chondrocytes and the complete differentiation stage of adipocytes, which originate from mesenchymal stem cells, as well as osteoblasts. These marker gene expressions were not observed in the muscle tissue surrounding the implanted Ca–P ceramics. These observations indicate that porous hydroxyapatite and BCP had a greater potential for promoting the differentiation of mesenchymal stem cells into osteoblasts than β-TCP.

Keywords: calcium phosphate ceramics, porous structure, osteogenesis, gene expression

Introduction

The widespread bone defects in clinics have emphasized the critical need for new materials as bone substitutes because autogenous bone grafts carry the risk of morbidity, and allograft bone poses the risk of transmitted diseases. The currently available bone substitutes have a wide variety of compositions and properties. Of these, inorganic calcium phosphate (Ca–P) ceramics have long been used in clinics. However, continuous efforts have been taken to develop ideal scaffolds for bone growth and osteogenic determination. The desirable properties of Ca–P ceramics include similarity in composition to the bone mineral, bioactivity, and osteoconductivity. When implanted in bone, Ca–P ceramics provide cell anchorage sites and structural guidance to form new bone [1]. Aside from the scaffold function, their potentials to promote osteogenic differentiation and gene expression were also demonstrated in vitro [2–5]. To disclose the material's function behind the camouflage of osteoconduction, several types of Ca–P ceramics are implanted into muscle, where the mechanical drive and growth environment are similar to those of bone, but the bone tissue is absent. Studies show that some specific Ca–P ceramics induce new bone formation in muscle without any help of added cells or growth factors externally [6–9]. This suggests that the material has its own role in promoting osteogenesis; however, the mechanism is yet unclear. Investigation of the cell determination, differentiation, and maturation behaviors in Ca–P ceramics is helpful to reveal the mechanism and identify ideal Ca–P ceramics for bone graft.

Osteoblasts differentiate from mesenchymal stem cells (MSCs), which are also the progenitors of other types of connective tissue, such as cartilage, adipose, and muscle [10]. The activation of progenitor cells along a specific pathway is due to either the expression of a master gene or a set of transcription factors. It has been identified that Cbfa1 (core-binding factor A1) and osterix are activators for osteogenesis, Sox9 (sex-determining region Y-box 9) for chondrogenesis, PPARγ (peroxisome proliferator-activated receptor-γ2) for adipogenesis, and MyoD for myogenesis [11]. Once MSCs are determined (or committed) toward one or several lineages, differentiation programs will proceed to give rise to a mature or functional population of cells. Differentiation pathways are therefore recognized by lineage-specific marker genes, like COL1A1 (collagen type I, alpha 1), ALP (alkaline phosphatase), SPARC (osteonectin), SPP1 (osteopontin), and BGLAP (osteocalcin) for osteoblasts [12], COL2A1 (collagen type II, alpha 1), COL9A1 (collagen type IX, alpha 1), and aggrecan for chondrocytes [13], LEP (leptin) and LPL (lipoprotein lipase) for adipocytes [14], and Mylpf (myosin light chain, phosphorylatable, fast skeletal muscle) for myocytes [15].

In this study, we investigated the osteoblast maturation behaviors affected by Ca–P ceramics at the transcriptional level in vitro and in vivo when implanted in rat muscle. We first analyzed the gene expression profiles of osteoblast-like cells cultured on dense-smooth and porous Ca–P ceramics with different Ca/P ratios using a DNA microarray technique. Next, we carried out in vivo evaluation of porous Ca–P ceramics. A new technique was used to make undecalcified cryosections of Ca–P ceramics to preserve the tissue morphology and RNA integrity at the same time. Tissues grown into the Ca–P ceramics were dissected accurately by laser microdissection, and reverse transcription polymerase chain reaction (RT-PCR) was performed to investigate the related gene expression pathways. We found that porous hydroxyapatite (Ca/P ratio=1.67) and biphasic calcium phosphate (BCP, Ca/P ratio=1.60) had a greater potential for promoting the osteoblast maturation than β-TCP, and osteoblasts could differentiate from mesenchymal stem cells in their pores. This study provided a molecular proof that Ca–P ceramics can promote the differentiation of mesenchymal stem cells into osteoblasts.

Experimental details

Preparation of dense and porous hydroxyapatite, beta-tricalcium phosphate, and biphasic calcium phosphate ceramics

Calcium phosphate powders were prepared by a wet method using Ca(NO3)2 and (NH4)2HPO4 as the starting materials. The reagents were mixed at Ca/P ratios of 1.67, 1.50 and 1.60 corresponding to hydroxyapatite, β-tricalcium phosphate (β-TCP), and biphasic calcium phosphate (BCP), respectively.

The powder mixtures were compressed into dense disks 20 mm in diameter and 2 mm in thickness under a uniaxial pressure of 40 MPa, and then sintered in air at 1100 °C for 2 h. The surfaces of the dense disks were polished with diamond abrasives to reduce the surface roughness. The dense-smooth disks were ultrasonically washed with distilled water three times, and then resintered at 800 °C for 2 h.

Powders of hydroxyapatite, β-TCP, and BCP were mixed with 5% H2O2 solution to prepare porous Ca–P ceramics. The mixed slurry was slowly heated in an oven from room temperature to 80 °C to make a foam. The dried porous green bodies were then sintered at 1100 °C for 2 h. The sintered porous ceramics were cut into flake like disks and then ultrasonically washed with distilled water three times. These porous ceramics were resintered at 800 °C for 2 h.

Characterization of calcium phosphate ceramics

The samples were analyzed by x-ray diffraction (XRD, X'Pert PRO MRD system, PANalytical, ALMELO, the Netherlands), using a CuKα radiation source at 40 kV and 30 mA over a 2θ range of 20–50°. The qualitative analyses of hydroxyapatite and β-TCP in the BCP ceramic were performed using integrated areas of hydroxyapatite and β-TCP diffraction peaks.

The surface morphologies of Ca–P ceramics were observed with a scanning electron microscope (SEM, JSM-5600 LV, JEOL, Tokyo, Japan). The surfaces were sputter-coated with platinum, and the accelerating voltage was 20 kV.

The nanostructures of the ceramics were observed with an atomic force microscope (AFM) (SPM-9500, Shimadzu Inc, Tokyo, Japan) in air. Silicon nitride probes mounted on cantilevers were applied in the dynamic mode (tapping mode) to avoid damaging the surface. Both the Z-range image and the root mean square (rms) roughness were analyzed.

Cell cultures

Osteoblast-like cells, MC3T3-E1, were subcultured on a tissue-culture–grade-polystyrene dish (TCPS) every 7 days in a proliferation medium consisting of α-Modified Eagle's Medium (α-MEM) with 10% (v/v) fetal bovine serum (FBS), 100 U ml−1 penicillin, and 100 μg ml−1 strep-tomycin. The subcultured cells were seeded at a density of 5600 cells cm−2 on all Ca–P ceramics and cultured in the proliferation medium. To promote osteoblastic differentiation, the medium was replaced after 9 days with a differentiation medium, which is a proliferation medium supplemented with 50 μg ml−1 of ascorbic acid and 2 mmol l−1β-glycerophosphate. The culture medium was changed every three days. All the cultures were grown at 37 °C in a humidified atmosphere containing 5% CO2.

DNA microarray hybridization and data analyses

The procedures for hybridizing the DNA microarray and analyzing the data have been described in [16]. RNA was extracted from the cells cultured for 20 days. The RNA was amplified using an Amino Allyl MessageAmpTM II aRNA Amplification Kit (Ambion Inc, TX, USA), and labeled with Cy3 or Cy5. The DNA microarray, AceGene-1 Chip Version-Mouse (Hitachi Software Engineering, Tokyo, Japan), was covered with a gap cover glass, and amplified RNAs labeled with Cy3 and Cy5 were injected. The DNA microarray was hybridized for 18 h at 50 °C. After hybridization, the microarray was washed and scanned using GenePix4000 (Axon Instruments, CA, USA). The microarray image was transferred into Dapple software (Washington University; http://www.cs.wustl.edu/∼jbuhler/dapple/) to automatically quantify the fluorescence intensity of each spot. The signal intensities of Cy3 and Cy5 were normalized using the global normalization method. The DNA microarray was hybridized twice in each combination. Reproducible spot data, whose signal intensity ratios in the two experiments were smaller than two, were used for further analysis. Hierarchical clustering was performed with the Pearson uncentered and average linkage cluster options of the TIGR multi-experimental viewer (MEV; http://www.tm4.org/mev.html).

The procedure for analyzing the expressions of ALP, osteocalcin and Ifitm5 genes using semiquantitative RT-PCR has been described in [16].

Treatment of animals

Seven-week-old male and 8-week-old female Sprague-Dawley rats (150–200 g) were anaesthetized by intraperitoneal injection of sodium phenobarbital at 30 mg kg−1 body weight. Porous Ca–P ceramics were implanted into the muscles of their hind limbs. At 4, 8, and 16 weeks after the surgery, the tissues harvested from the sacrificed animals were cryoembedded in super-cryoembedding medium (SCMC, Finetec Co Ltd, Tokyo, Japan) gel using hexane, then cooled with dry ice. The frozen tissue blocks were stored at −80 °C until further use. The experiment was approved by the local Animal Testing and Ethical Committee and veterinary authorities (No 15-2007-3).

Cryosection

The cryosections of the Ca–P ceramics were prepared by Kawamoto's method [17, 18]. In cryosectioning, the adhesive side of a film (LMD film, Finetec Co Ltd, Tokyo, Japan) was fastened to the cut surface of the sample to support the section. Serial cryosections of 8 μm thickness were cut slowly with a disposable tungsten carbide blade (TC-65, Leica Biosystems, Nussloch GmbH, Germany) in a cryostat (CM 1850, Leica Microsystems, Nussloch GmbH, Germany). The nonadhesive side of the film was then attached to the glass slide for support, with the section facing upward. Fixation was carried out by immersing the slides in precooled methanol for 1 min.

ALP and tartrate-resistant acid phosphatase (TRAP) staining

The ALP and TRAP activities were detected using an Alkaline Phosphatase Detection Kit (Chemicon International Inc, CA, USA) and a TRACP & ALP Double-staining Kit (TaKaRa Bio Inc, Shiga, Japan) according to the manufacturer's instructions. The pictures were captured using a Digital Microimaging Device (DMD108, Leica Microsystems, Nussloch GmbH, Germany).

Gene expression analysis from harvested tissue

The grown tissues in the pores of Ca–P ceramics were collected from the fixed cryosections with a laser microdissection system (LMD, LMD6000, Leica Microsystems, Nussloch GmbH, Germany) using a UV laser of 355 nm wavelength. The total RNA was extracted from the collected tissues using an RNeasy Mini or Micro Kit (Qiagen, Hilden, Germany) that included an on-column DNase I digestion step. Single-strand cDNA was synthesized using a PrimeScriptTM RT reagent kit (Takara Bio Inc, Shiga, Japan), and then RT-PCR was performed using Premix Taq® Hot Start Version (Takara Bio Inc, Shiga, Japan). The RT-PCR conditions were as follows: for Cbfa1, forward primer: 5′-ccacagagctattaaagtgacagtg-3′, reverse primer: 5′-aggtttagag tcatcaagcttctgt-3′, 35 cycles, 80 bp product; for osteocalcin (OC), forward primer: 5′-agcttcagctttggctactctc-3′, reverse primer: 5′-cctcatctggactttattttgga-3′, 30 cycles, 104 bp product; for Sox-9, forward primer: 5′-tactggtctgccagcttcct-3′, reverse primer: 5′-ctgaagggctacgactggac-3′, 35 cycles, 140 bp product; for aggrecan, forward primer: 5′-aaagtgtccaaggcatcca-3′, reverse primer: 5′-acacccctacccttgcttct-3′, 35 cycles, 124 bp product; for PPARγ, forward primer: 5′-actggcacccttgaaaaatg-3′, reverse primer: 5′-ccctggcaaagcatttgtat-3′, 30 cycles, 222 bp product; for LPL, forward primer: 5′-ctgaccagcggaagtaggag-3′, reverse primer: 5′-gcatttgagaaagggctctg-3′, 25 cycles, 370 bp product; for GAPDH, forward primer: 5′-gtcctcagtgtagcccagga-3′, reverse primer: 5′-agaacatcatccctgcatcc-3′, 25 cycles, 227 bp product. Each number of cycles was determined on the basis of the linear amplification range. The PCR products were analyzed on 2% agarose gels and visualized with ethidium bromide. The density of each band was quantified using Dolphin-Doc imaging system (Wealtec Corp, GA, USA) and a software of ImageMaster Total Lab (v2.01, Amersham Parmacia Biotech, NJ, USA). We determined the relative gene expression levels by dividing the densitometry value of the mRNA RT-PCR product by that of the GAPDH product. The RT-PCR values were generated from four samples for each group and from three replicate runs.

Results

Surface topography and phase composition of Ca–P ceramics

The surface topography of the dense-smooth ceramics was observed with AFM (figure 1(a)). The size of each grain boundary in hydroxyapatite was below 1 μm. The β-TCP had a larger grain boundary than hydroxyapatite, including nanopores of 400–500 nm appearing as black regions. Some surfaces of the grain boundary had a nano-bumpy structure. In BCP, grain boundaries with two different structures, attributed to the β-TCP and hydroxyapatite phases, were observed, including nanopores of 100–200 nm. The grain structure attributed to β-TCP had the same nano-bumpy structure of the pure β-TCP ceramic. The averaged surface roughness values (rms) of hydroxyapatite, β-TCP, and BCP were approximately 200, 2250 and 480 nm, respectively. The porous ceramics had both macropores (100–500 μm) and micropores on the macropore walls as shown in figure 1(b). The porosity of these porous ceramics was about 50%.

Figure 1.

Figure 1

Open in a new tab

Surface topography and x-ray diffraction patterns of Ca–P ceramics. (a) The surface of dense-smooth Ca–P ceramics was observed using an atomic force microscope (AFM). The bars are 1 μm in length. (b) The surface of porous Ca–P ceramics was observed using a scanning electron microscope (SEM). The bars are 1 mm in length. (c) X-ray diffraction patterns of porous hydroxyapatite (HA), β-TCP, and BCP ceramics prepared by the wet method from CaNO3 and (NH4)2HPO4. The weight ratio of hydroxyapatite/β-TCP in the BCP ceramic is 70/30. The open circles are attributed to the β-TCP phase and the closed circles to the hydroxyapatite phase.

The phase compositions of the Ca–P ceramics prepared in this study were analyzed via their XRD patterns (figure 1(c)). The diffraction widths of all the ceramics were sharp, indicating good sinterability. All diffractions of the Ca–P ceramics prepared at a Ca/P ratio of 1.67 were attributed to the single phase of hydroxyapatite (JCPDS 9-432) without other phases of calcium oxide and α- and β-TCP. The β-TCP sintered at 1100 °C was a single phase of β-TCP corresponding to JSCPS 9-169. The BCP had hydroxyapatite and β-TCP phases without other phases, where the amount of hydroxyapatite and β-TCP, calculated from the (210) peak of hydroxyapatite and the (211) peak of β-TCP, was 70/30 in weight percent.

Gene expression profile of osteoblasts in vitro

We first compared the gene expression profiles of osteoblast-like MC3T3-E1 cells cultured on dense-smooth and porous Ca–P ceramics. The cells were cultured for 20 days on these materials and DNA microarrays were analyzed. The Ca–P ceramics were grouped on the basis of similarity in the gene expression profiles of osteoblast-like cells using hierarchical clustering (figure 2(a)). Hierarchical clustering indicated that the comprehensive gene expression profiles can be divided into two groups. The first included dense-smooth hydroxyapatite, β-TCP, and BCP; the second included porous β-TCP, hydroxyapatite, and BCP. This implied that the gene expression profile in porous Ca–P ceramics was different from that in dense-smooth Ca–P ceramics.

Figure 2.

Figure 2

Open in a new tab

Gene expression of osteoblast-like cells cultured on Ca–P ceramics. (a) Results of hierarchical clustering of differentially expressed genes obtained from DNA microarray analysis. The clustering image shows the similarities in the expressions of 4087 genes. The expression levels are indicated from green (lower expression than that of the control of the tissue-culture–grade-polystyrene dish) to red (higher expression than that of the control). The relationship between the Ca–P ceramics based on similarities in gene expression patterns is indicated by the tree. (b) The relative expression levels of marker genes for osteoblast maturation. The data were obtained in triplicate or quadruplicate for each RT-PCR from two independent 20-day-old cultures (mean±SE). HA is hydroxyapatite; OC is osteocalcin.

The expression levels of osteoblast-specific marker genes including ALP, osteocalcin, and interferon-induced transmembrane protein 5 (Ifitm5) genes were examined by RT-PCR (figure 2(b)). Ifitm5 is a novel marker gene involved in osteoblast maturation, which we found previously [16]. The expression levels of ALP and osteocalcin genes were significantly higher in porous hydroxyapatite, β-TCP and BCP ceramics than in dense-smooth ceramics. In addition, no expressions of the Ifitm5 gene were observed in the dense-smooth Ca–P ceramics, but the Ifitm5 gene was highly expressed in the porous ceramics. These results from DNA microarray and RT-PCR analyses suggest that the gene expression profiles in the porous Ca–P ceramics are different from those in dense-smooth Ca–P ceramics, and that the former has a higher potential for osteoblast maturation than the latter.

DNA microarray analysis also revealed the specific genes that are up- or down-regulated in porous Ca–P ceramics. Among the reproducible unique genes, 78 genes were up-regulated more than 2-fold (1-fold in log2 ratio) (Supplementary table 1), while 12 genes were down-regulated (Supplementary table 2).

Supplementary Table 1.

Genes up-regulated in porous Ca–P ceramics.

arachidonate 12-lipoxygenase, 12r type
zinc finger protein
bcl2-associated athanogene 3
lymphoblastomic leukemia
hoxa-4
mszf52; zinc finger
similar to deleted in lung and esophageal cancer 1 isoform DLEC1-N1 isoform 6
3-phosphoglycerate dehydrogenase
peptidoglycan recognition protein precursor
homolog to vesicle associated protein
solute carrier family 39 (zinc transporter), member 7
homolog to leucine-rich repeats containing f-box protein fbl3
similar to hypothetical protein dkfzp761d221
protein phosphatase 1, catalytic subunit, alpha isoform
ring finger protein 26
zinc finger, c3hc4 type (ring finger) containing protein
hls7-interacting protein kinase
cathepsin r
T cell receptor alpha chain v region
zinc finger CCCH-type containing 12D
kelch like protein
cadherin
rad53 homolog (s.cerevisiae)
olfactory receptor mor10-2
thymoma viral proto-oncogene 1
chaperonin subunit 7
myelin-oligodendrocyte glycoprotein precursor
coactivator-associated arginine methyltransferase 1
thymic stromal-derived lymphopoietin, receptor
cop9 complex subunit 6
U7 snRNP-specific Sm-like protein LSM11
cad protein
olfactory receptor mor255-1
myosin heavy chain
cortistatin
extracellular calcium sensing receptor precursor
potassium channel
small proline-rich protein 2d
spliceosomal protein gene 62
dihydropyridine sensitive l type, calcium channel alpha 2/de
ubiquinol-cytochrome c reductase core protein 1
sorting nexin 9 sh3 and px domain containing protein 1
small inducible cytokine a4
death associated protein 3
GLI-Kruppel family member HKR3
natural resistance associated macrophage protein nramp
ASF1 anti-silencing function 1 homolog B (S. cerevisiae)
prion protein dublet
ubiquitin conjugating enzyme 7 interacting protein 3
adrenergic receptor, alpha 1b
n-acetyltransferase similar to s. cerevisiae ard1
dehydrogenase/reductase (SDR family) member 1
ig mu chain c region
nucleoplasmin 3
ras gtpase activating like protein
endothelial cells scavenger receptor precursor acetyl ldl
olfactory receptor
proteasome (prosome, macropain) 26 s subunit, non-atpase, 9
eif5a
phospholipase adrab b precursor ec 3
40s ribosomal protein
membrane-associated tyrosine-and threonine-specific cdc2-inhibitory kinase
werner syndrome homolog (human) interacting protein
aconitase 2, mitochondrial
pro alpha1 (ix) collagen chain
transmembrane protein 127
tetracycline transporter-like protein
similar to RNA binding motif protein 7
Open in a new tab

Supplementary Table 2.

Genes down-regulated in porous Ca–P ceramics.

cd14 antigen
40s ribosomal protein
DnaJ (Hsp40) homolog, subfamily C, member 5 beta
rev protein
3110031B13Rik/cdna clone hypothetical protein
T-complex protein 11
b(2)-microglobulin
major intrinsic protein of eye lens fiber
etf-related factor-2a
melanoma antigen, family b, 3
pctaire-motif protein kinase 3
f4 4 novel protein similar to nucleolar protein 4
Open in a new tab

Expression level of genes related to osteogenic differentiation in vivo

Although the in vitro results demonstrated that the porous Ca–P ceramics had a higher potential for osteoblast maturation than the dense-smooth Ca–P ceramics, we could not obtain clear differences between hydroxyapatite, BCP, and β-TCP. Therefore, we next examined what effect the three porous Ca–P ceramics had on osteoblast maturation in vivo.

Porous Ca–P ceramics were implanted in the muscles of the hind limbs of rats. After 4, 8, and 16 weeks, the implanted Ca–P ceramics were harvested and immediately cryoembedded. ALP staining of the cryosections revealed that osteoblasts were present at the periphery of the materials and in the pores of all Ca–P ceramics at 16 weeks after the implantation (figure 3). There were more osteoblasts, predicted from ALP staining, in hydroxyapatite and BCP than in β-TCP. We also observed TRAP-stained osteoclasts that were tightly attached to the Ca–P ceramics (figure 3). Both ALP-stained osteoblasts and TRAP-stained osteoclasts could be observed even at 4 and 8 weeks (data not shown).

Figure 3.

Figure 3

Open in a new tab

Double staining of ALP and TRAP on cryosections prepared from porous Ca–P ceramics implanted in rat hind-limb muscle for 16 weeks. The left column shows osteoblasts stained by ALP (blue). Some of them are scattered in the matrix, others are arranged in a group along the edge of pores, which formed osteoid (pale pink). The right column shows osteoclasts stained with TRAP (red, arrowhead), which tightly attached to the ceramics. There were fewer osteoclasts than osteoblasts. The pictures were captured using a Digital Microimaging Device (DMD108). Scalebar=500 μm. HA is hydroxyapatite.

Next, we examined the transcriptional gene expressions of core binding factor alpha 1 (Cbfa1) and osteocalcin (figure 4). In Cbfa1 gene expression, which is an important transcription factor for differentiating osteoblasts, no significant differences were observed between hydroxyapatite, β-TCP, or BCP at either 8 or 16 weeks, although the expression levels at 16 weeks were higher than those at 8 weeks (figures 4(a) and (b)). In osteocalcin gene expression, which is a late differentiation marker of osteoblasts, although there were no differences in the expression levels at 8 weeks, the level was higher in hydroxyapatite and BCP than in β-TCP at 16 weeks (figures 4(a) and (b)). This suggests that hydroxyapatite and BCP have higher potential for in vivo osteoblast maturation than β-TCP. We analyzed these two gene expressions in muscle surrounding the implanted porous Ca–P ceramics, but no expressions were observed (figure 4(c)), suggesting the absence of osteoblast progenitors and osteoblasts in muscle tissue surrounding implanted Ca–P ceramics.

Figure 4.

Figure 4

Open in a new tab

Expression levels of osteogenic marker genes in tissue grown in pores of Ca–P ceramics implanted in rat hind-limb muscle: (a) Ethidium bromide staining of PCR products. (b) Band densitometry values of (a) normalized with the GAPDH expression level. The data were obtained in triplicate for each RT-PCR from two independent samples. (c) Expressions of Cbfa1 and osteocalcin (OC) in muscle where Ca–P ceramics were implanted. HA is hydroxyapatite.

Expression levels of genes related to other differentiation lineages of mesenchymal stem cells

The osteocalcin gene expression in the tissues grown in pores implied that porous Ca–P ceramics promoted the differentiation of osteoblasts. Osteoblasts originate from mesenchymal stem cells, and mesenchymal stem cells can also differentiate into chondrocytes and adipocytes. Therefore, we investigated the presence of chondrocytes and adipocytes in the tissues grown in pores using the marker genes expressions of these cell types (figure 5(a)). We detected the gene expression of Sox-9, but not aggrecan in any Ca–P ceramics, suggesting the presence of chondrocytes, which were in the early differentiation stage. We also observed the expressions of peroxisome proliferator-activated receptor gamma (PPARγ) and lipoprotein lipase (LPL) genes, which are markers of adipocytes. Although the expression levels of osteocalcin which is a marker of the late differentiation of osteoblasts, were higher in hydroxyapatite and BCP (figures 4(a) and (b)) than in β-TCP, no significant differences were observed in the expression levels of the marker genes of chondrocytes and adipocytes. No expressions of these marker genes were observed in muscle tissue surrounding the implanted Ca–P ceramics (figure 5(b)), which suggests that chondrocytes and adipocytes were absent in the muscle tissue.

Figure 5.

Figure 5

Open in a new tab

Expression levels of marker genes for chondrocyte and adipocyte. (a) Expressions of marker genes for chondrocytes (Sox-9 and aggrecan) and adipocytes (PPARγ and LPL) in tissue grown in the pores of Ca–P ceramics implanted into rat hind-limb muscle for 16 weeks. (b) Expressions of marker genes for chondrocytes, adipocytes, and skeletal muscle (myosin light chain, phosphorylatable, fast skeletal muscle, Mylpf) in muscle where Ca–P ceramics were implanted.

Discussion

We prepared dense-smooth and porous ceramics of hydroxyapatite, β-TCP, and BCP and first evaluated these Ca–P ceramics as bone-replacement materials by an in vitro experiment. Although rats were used in the experiments, we analyzed the comprehensive gene expression profiles of murine osteoblast-like MC3T3-E1 cells using mouse DNA microarray, since the annotation of genes in rat DNA microarray is not sufficient to obtain more information on gene functions. MC3T3-E1 cells represent a clonal cell line considered to be an in vitro model for osteoblast maturation [19, 20]. Analysis of the DNA microarrays made it clear that the comprehensive gene expression pattern of osteoblast-like cells cultured on porous Ca–P ceramics differed from that of cells cultured on dense-smooth Ca–P ceramics. ALP, osteocalcin, and Ifitm5 genes, which are osteogenic markers, were more highly expressed in porous Ca–P ceramics than in dense-smooth materials. These in vitro analyses suggested that porous Ca–P ceramics had a greater potential for osteoblast maturation than dense-smooth ceramics. We found specific genes that are up- or down-regulated in porous ceramics from comprehensive gene expression analysis (Supplementary tables 1 and 2). The important unknown genes involved in osteoblast maturation may be included in these specific genes. In a scaffold for cell and tissue growth, porosity is one of the most crucial factors affecting biological activity. It has been shown that a pore size of approximately 100 μm diameter is required for the ingrowth of bone cells, and 100–600 μm facilitates osteoconductivity [21–24].

The in vivo experiment that followed clearly revealed that porous hydroxyapatite and BCP have a greater potential for osteoblast differentiation than β-TCP. The ALP staining of cryosections suggested that osteoblasts appeared inside the pores of porous Ca–P ceramics implanted in muscle. Although we could not determine the suitable pore sizes for tissue growth, it seems that 200–500 μm diameter is sufficient for tissue growth and osteoblast differentiation. Cbfa1 revealed the existence of osteoblast progenitors. Osteocalcin is a matrix protein involved in calcification, and it has been used as a specific marker of the late stage of osteogenesis. There were no clear differences in the levels of Cbfa1 expressions among these three Ca–P ceramics. On the other hand, higher expression levels of osteocalcin were observed in hydroxyapatite and BCP than in β-TCP at 16 weeks, although their expression levels at 8 weeks varied slightly among Ca–P ceramics. There were more ALP-stained osteoblasts in hydroxyapatite and BCP than in β-TCP at 16 weeks, which may be attributed to the higher expression level of the osteocalcin gene in hydroxyapatite and BCP than in β-TCP. In porous Ca–P ceramics, hierarchical clustering obtained by DNA microarray analysis revealed that the comprehensive gene expression profiles in hydroxyapatite and BCP were similar, but the profile in β-TCP was dissimilar to those in hydroxyapatite and BCP. This in vitro result may help explain the difference in the effects of porous β-TCP from porous hydroxyapatite and porous BCP on osteoblast differentiation in the in vivo experiment.

The chemical composition based on the Ca/P ratio first affects the dissolution properties of Ca–P ceramics [2,5, 2,6]. β-TCP whose Ca/P ratio is 1.50 is completely soluble after long-term implantation, but hydroxyapatite with a Ca/P ratio of 1.67 is insoluble in vivo. The activities of osteoblasts and osteoclasts are sensitive to the calcium ion concentration [2,7]. It is likely that the microenvironment in the pores was created by the dissolution of Ca–P ceramics, which further affected osteogenesis. The microenvironment and the greater dissolution properties of β-TCP might inhibit osteoblast differentiation. We observed osteoclasts in the tissue grown in the pores of all Ca–P ceramics by TRAP staining. Osteoclasts contribute to the degradation of Ca–P ceramics [2,8, 2,9] and they also provide factors that stimulate osteoblast lineage cells to replace resorbed bone [30].

We previously demonstrated that porous BCP has superior potential for bone formation in the soft tissue of dogs and goats than hydroxyapatite [31, 3,2], but there were no differences in the induction of osteoblast differentiation between BCP and hydroxyapatite in this study. This discrepancy might be attributed to the difference in the sintering temperature of hydroxyapatite and/or animal species. Our hydroxyapatite was sintered at 1100 °C, which was lower than the standard sintering temperature of around 1250 °C. Ca–P ceramics sintered at relatively lower temperatures result in a smaller crystal size, which is favorable for cell attachment, proliferation, and bone-tissue formation [3,3]. The formation of ectopic bone by Ca–P ceramics was reported in dogs [6], sheep [7], and rabbits [9], and the formation was dependent on the animal species [8]. There were few reports on ectopic-bone formation by Ca–P ceramics in rats. Dependence on animal species is attributed to the difference in the number of circulating growth factors in each species [3,4].

Where did the osteoblasts in the pores come from? We observed the expressions of Sox-9, PPARγ, and LPL genes but not aggrecan gene with no significant differences among Ca–P ceramics. These results indicated that not only osteoblasts but also early differentiated chondrocytes and fully differentiated adipocytes were present in the pores. These cell types originate from mesenchymal stem cells. We could not detect marker gene expressions of osteoblasts, chondrocytes, or adipocytes in muscle tissue where Ca–P ceramics had been implanted. This suggests that mesenchymal stem cells but not differentiated cells had migrated into the pores. However, what triggers mesenchymal stem cells to migrate into the pores of Ca–P ceramics? Further studies on this issue are required. The expression levels of the osteocalcin gene in hydroxyapatite and BCP at 16 weeks after implantation were higher than that in β-TCP, and the more ALP-positive osteoblast cells in hydroxyapatite and BCP could be attributed to the higher expression level of this late-stage marker gene. However, the osteocalcin gene expression level in β-TCP at 8 weeks after the implantation was almost the same as those in hydroxyapatite and BCP. This suggests that the differentiation into osteoblast lineage from MSCs still continued in hydroxyapatite and BCP, while in β-TCP, the differentiation ceased after 8 weeks. We observed a higher expression level of bone morphogenetic protein 2 (BMP-2) gene in the pores of hydroxyapatite and BCP than in β-TCP at 16 weeks (data not shown). Because BMP-2 is a critical factor for osteoblast differentiation, the higher expression level of BMP-2 might contribute to the differentiation into osteoblast lineage from MSCs in porous hydroxyapatite and BCP. This study provides clues on the mechanism of ectopic bone formation using implanted Ca–P ceramics.

Conclusions

This study provided a molecular-level proof that Ca–P ceramics promote osteogenic commitment and osteoblast maturation. The surface structures of Ca–P ceramics had a greater effect on osteoblast maturation than the chemical composition. When implanted in muscle, porous HA and BCP had the potential for promoting osteoblast maturation in the hind-limb muscle of rats, and osteoblasts could differentiate from mesenchymal stem cells in tissue grown in the pores of HA and BCP.

Acknowledgments

The authors wish to thank Dr Y Shirai for feeding and breeding of the experimental animals, Ms. K. Izumi for her technical guidance with the cryosection, and Drs R Nishiyama and S Yamada for their assistance with the laser microdissection. This work was partially supported by grants from the Research Promotion Bureau of the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan, under contract No 16-794.

References

  1. El-Ghannam A. Expert Rev. Med. Devices. 2005;2:87. doi: 10.1586/17434440.2.1.87. [DOI] [PubMed] [Google Scholar]
  2. Wang C, Duan Y, Markovic B, Barbara J, Howlett C R, Zhang X. and Zreiqat H. Biomaterials. 2004;25:2507. doi: 10.1016/j.biomaterials.2003.09.035. [DOI] [PubMed] [Google Scholar]
  3. Tan Y, Wang G, Fan H, Wang X, Lu J and Zhang X. 2007. J. Biomed. Mater. Res. A 82 152. [DOI] [PubMed] [Google Scholar]
  4. Sun H, Ye F, Wang J, Shi Y, Tu Z, Bao J, Qin M, Bu H. and Li Y. Transplant. Proc. 2008;40:2645. doi: 10.1016/j.transproceed.2008.07.096. [DOI] [PubMed] [Google Scholar]
  5. Müller P.et al2008J. Cell. Mol. Med. 12281. 10.1111/j.1582-4934.2007.00103.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Kondo N.et al2006Biomaterials 274419. 10.1016/j.biomaterials.2006.04.016 [DOI] [PubMed] [Google Scholar]
  7. Le Nihouannen D, Daculsi G, Saffarzadeh A, Gautheir O, Delplace S, Pilet P. and Layrolle P. Bone. 2005;36:1086. doi: 10.1016/j.bone.2005.02.017. [DOI] [PubMed] [Google Scholar]
  8. Yang Z, Yuan H, Tong W, Zou P, Chen W. and Zhang X. Biomaterials. 1996;17:2131. doi: 10.1016/0142-9612(96)00044-0. [DOI] [PubMed] [Google Scholar]
  9. Kurashina K, Kurita H, Wu Q, Ohtsuka A. and Kobayashi H. Biomaterials. 2002;23:407. doi: 10.1016/S0142-9612(01)00119-3. [DOI] [PubMed] [Google Scholar]
  10. Dennis J E. and Charbord P. Stem Cells. 2002;20:205. doi: 10.1634/stemcells.20-3-205. [DOI] [PubMed] [Google Scholar]
  11. Silva W A Jr, Covas D T, Panepucci R A, Proto-Siqueira R, Siufi J L, Zanette D L, Santos A R. and Zago M A. Stem Cells. 2003;21:661. doi: 10.1634/stemcells.21-6-661. [DOI] [PubMed] [Google Scholar]
  12. Kulterer B, Friedl G, Jandrositz A, Sanchez-Cabo F, Prokesch A, Paar C, Scheideler M, Windhager R, Preisegger K H. and Trajanoski Z. BMC Genomics. 2007;8:70. doi: 10.1186/1471-2164-8-70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. de Crombrugghe B, Lefebvre V, Behringer R R, Bi W, Murakami S. and Huang W. Matrix Biol. 2000;19:389. doi: 10.1016/S0945-053X(00)00094-9. [DOI] [PubMed] [Google Scholar]
  14. Schilling T, Nöth U, Klein-Hitpass L, Jakob F. and Schütze N. Mol. Cell. Endocrinol. 2007;271:1. doi: 10.1016/j.mce.2007.03.004. [DOI] [PubMed] [Google Scholar]
  15. Dezawa M, Ishikawa H, Itokazu Y, Yoshihara T, Hoshino M, Takeda S, Ide C. and Nabeshima Y. Science. 2005;309:314. doi: 10.1126/science.1110364. [DOI] [PubMed] [Google Scholar]
  16. Hanagata N, Takemura T, Monkawa A, Ikoma T and Tanaka J. 2007. J. Biomed. Mater. Res. A 83A 362 10.1002/jbm.a.31240 [DOI] [PubMed] [Google Scholar]
  17. Kawamoto T. Arch. Histol. Cytol. 2003;66:123. doi: 10.1679/aohc.66.123. [DOI] [PubMed] [Google Scholar]
  18. Kawamoto T. and Shimizu M. Histochem. Cell Biol. 2000;113:331. doi: 10.1007/s004180000149. [DOI] [PubMed] [Google Scholar]
  19. Kodama H, Amagai Y, Sudo H, Kasai S. and Yamamoto S. Japan J. Oral Biol. 1981;23:899. [Google Scholar]
  20. Sudo H, Kodama H A, Amagai Y, Yamamoto S. and Kasai S. J. Cell Biol. 1983;96:191. doi: 10.1083/jcb.96.1.191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. von Doernberg M C, von Rechenberg B, Bohner M, Grünenfelder S, van Lenthe G H, Müller R, Gasser B, Mathys R, Baroud G. and Auer J. Biomaterials. 2006;27:5186. doi: 10.1016/j.biomaterials.2006.05.051. [DOI] [PubMed] [Google Scholar]
  22. Daculsi G. and Passuti N. Biomaterials. 1990;11:86. [PubMed] [Google Scholar]
  23. Karageorgiou V. and Kaplan D. Biomaterials. 2005;26:5474. doi: 10.1016/j.biomaterials.2005.02.002. [DOI] [PubMed] [Google Scholar]
  24. Mastrogiacomo M, Scaglione S, Martinetti R, Dolcini L, Beltrame F, Cancedda R. and Quarto R. Biomaterials. 2006;27:3230. doi: 10.1016/j.biomaterials.2006.01.031. [DOI] [PubMed] [Google Scholar]
  25. Ducheyne P, Radin S. and King L. J. Biomed. Mater. Res. 1993;27:25. doi: 10.1002/jbm.820270105. [DOI] [PubMed] [Google Scholar]
  26. LeGeros R Z, Lin S, Rohanizadeh R, Mijares D. and LeGeros J P. J. Mater. Sci. Mater. Med. 2003;14:201. doi: 10.1023/A:1022872421333. [DOI] [PubMed] [Google Scholar]
  27. Rice J M, Hunt J A. and Gallagher J A. Calcif. Tissue Int. 2003;72:726. doi: 10.1007/s00223-002-2045-y. [DOI] [PubMed] [Google Scholar]
  28. Wenisch S, Stahl J P, Horas U, Heiss C, Kilian O, Trinkaus K, Hild A and Schnettler R. 2003. J. Biomed. Mater. Res. A 67 713 10.1002/jbm.a.10091 [DOI] [PubMed] [Google Scholar]
  29. Yamada S, Heymann D, Bouler J M. and Daculsi G. Biomaterials. 1997;18:1037. doi: 10.1016/S0142-9612(97)00036-7. [DOI] [PubMed] [Google Scholar]
  30. Martin T J. and Sims N A. Trends Mol. Med. 2005;11:76. doi: 10.1016/j.molmed.2004.12.004. [DOI] [PubMed] [Google Scholar]
  31. Yuan H, Yang Z, De Bruij J D, De Groot K. and Zhang X. Biomaterials. 2001;22:2617. doi: 10.1016/S0142-9612(00)00450-6. [DOI] [PubMed] [Google Scholar]
  32. Yuan H, Yang Z, Li Y, Zhang X, De Bruijn J D. and De Groot K. J. Mater. Sci. Mater. Med. 1998;9:723. doi: 10.1023/A:1008950902047. [DOI] [PubMed] [Google Scholar]
  33. Wilson C E, Kruyt M C, de Bruijn J D, van Blitterswijk C A, Oner F C, Verbout A J. and Dhert W J. Biomaterials. 2006;27:302. doi: 10.1016/j.biomaterials.2005.06.041. [DOI] [PubMed] [Google Scholar]
  34. LeGeros R Z. Chem. Rev. 2008;108:4742. doi: 10.1021/cr800427g. [DOI] [PubMed] [Google Scholar]

Articles from Science and Technology of Advanced Materials are provided here courtesy of National Institute for Materials Science and Taylor & Francis

RESOURCES