Effects of Calcium Phosphate/Chitosan Composite on Bone Healing in Rats: Calcium Phosphate Induces Osteon Formation

Original Article

Effects of Calcium Phosphate/ChitosanComposite on Bone Healing in Rats:

Calcium Phosphate Induces Osteon Formation

AU1 c Tulio Fernandez,1,2 Gilberto Olave,2 Carlos H Valencia,2 Sandra Arce,3 Julian M.W. Quinn,1

AU2 c George A Thouas,1 and Qi-Zhi Chen, PhD1

Vascularization of an artificial graft represents one of the most significant challenges facing the field of bonetissue engineering. Over the past decade, strategies to vascularize artificial scaffolds have been intensivelyevaluated using osteoinductive calcium phosphate (CaP) biomaterials in animal models. In this work, we observedthat CaP-based biomaterials implanted into rat calvarial defects showed remarkably accelerated formation andmineralization of new woven bone in defects in the initial stages, at a rate of *60mm/day (0.8 mg/day), whichwas considerably higher than normal bone growth rates (several mm/day, 0.1 mg/day) in implant-free controls ofthe same age. Surprisingly, we also observed histological evidence of primary osteon formation, indicated byblood vessels in early-region fibrous tissue, which was encapsulated by lamellar osteocyte structures. These werelater fully replaced by compact bone, indicating complete regeneration of calvarial bone. Thus, the CaP bio-material used here is not only osteoinductive, but vasculogenic, and it may have contributed to the bone re-generation, despite an absence of osteons in normal rat calvaria. Further investigation will involve how thisstrategy can regulate formation of vascularized cortical bone such as by control of degradation rate, and use ofmodels of long, dense bones, to more closely approximate repair of human cortical bone.

Introduction

Bone repair is a subject of intensive investigation inorthopedic reconstruction because of the great clinical

need for effective approaches to enhance or direct bonehealing. Current approaches in bone reconstructive surgeryare dominated by autografts and allografts. However, bio-logical bone grafts all have shortcomings, such as donor siteshortage and morbidity in autografting, immune rejection,and the transmission of diseases (such as HIV and hepatitisvirus) in allografting and cross-contamination of animalviruses associated with xenografting.1 Over the past decade,scaffold-based tissue engineering strategies have been in-vestigated using various synthetic materials, including hy-droxyapatite (HA), calcium phosphates (CaPs), polyesters,chitosan, and their composites.2 Among these artificial bonematrices, CaP, because of its intriguing osteoinductivity,3,4

has been extensively evaluated in vivo.5,6 Among all animalmodels, rats and mice are the most widely used, primarilybecause of their relatively low cost compared with livestockand nonhuman primates. However, rat and mouse bones donot have Haversian canals,7 which are the essential struc-

tures carrying the vascular network in human cortical bone.This difference between rodent and human bone structure issignificant, bearing in mind that vascularization of bonetissue engineering scaffolds has been recognized as themajor obstacle to successfully achieving clinically viableartificial scaffolds.8 Therefore, the primary objective of thiswork is to address two questions: first, whether CaP im-plantation can result in formation of Haversian canals inrodents and, second, whether a rodent model can be used inevaluating bone engineering strategies, especially thoseaimed at promoting vascularization of cortical bone.

Osteoinduction is a process that results in heterotopic orectopic bone formation in vivo. Numerous growth factors(notably bone morphogenetic proteins; BMPs) and biomate-rials such as those containing CaP display an ability tostimulate osteoinduction.9 The reported osteoinductivity ofCaP is influenced by a number of physicochemical propertiesof the biomaterial, such as particle size,10 surface area,crystallinity, porosity, and composition.11,12 Particles of sizesranging from 80 to 300mm in diameter can induce ectopicbone formation, whereas particles greater than 500mm are notosteoinductive.13 In vivo studies also indicate that a specific

1Department of Materials Engineering, Monash Medical School, Monash University, Clayton, Australia.2School of Dentistry, University of Valle, Cali, Colombia b AU3.3Autonomous University of the Occident, Cali, Colombia.

TISSUE ENGINEERING: Part AVolume 00, Number 00, 2014ª Mary Ann Liebert, Inc.DOI: 10.1089/ten.tea.2013.0696

1

TEA-2013-0696-ver9-Fernandez_1P

Type: research-article

TEA-2013-0696-ver9-Fernandez_1P.3d 03/15/14 12:06pm Page 1

surface area above a threshold level of 1.0 m2/g is critical forCaP to exhibit osteoinduction.6 Moreover, biphasic CaP(mixtures of HA with CaP) generally demonstrate simulta-neously increased solubility and osteoinductivity,14 whereaspure HA or amorphous tricalcium phosphate (TCP) displayno osteoinductivity because of either having too high stabilityor too high a dissolution rate.15 Based on these consider-ations, it is likely that the osteoinductivity of CaP-basedbiomaterials is in essence controlled by their degradationkinetics, and other properties (e.g., particle size, surface area,crystallinity, porosity, and composition) influence the degra-dation rate and thus the osteoinductivity. Hence, a secondaryobjective of this work is to investigate whether the degrada-tion rate of CaP-based particles correlates with the bonegrowth rate induced by such materials.

Chitosan is another biomaterial being intensively inves-tigated for its ability to enhance osteoinductivity of CaP-based materials. Chitosan is a polysaccharide composed ofglucosamine and N-acetyl glucosamine and is naturally cat-ionic, so it can produce electrostatic unions with the glycos-aminoglycan anions, proteoglycans, and other negativelycharged molecules available in the extracellular matrix. Thisproperty has a positive influence on the bone healing processas cytokines and growing factors are attached to glycosami-noglycans such as heparin and heparin sulfate. Chitosan isreported to be able to enhance bone healing through pro-moting polymorphonuclear infiltration at the healing site andtheir ability to bind anionic molecules such as grown factorsand DNA.16 Thus, a graft of chitosan–glycosaminoglycanmay help increase the concentration of grow factors releasedby colonizing cells. For these reasons, the biomaterial used inthis study is a composite of b-tricalcium phosphate (b-TCP)and chitosan in its acetylated form.

In principle, the application of biomaterials as artificialbone substitutes is aimed at healing of large bone defectsthat cannot heal themselves and requires a relatively largebone graft. A critical-sized defect (CSD) is defined to be theminimal defect that would not heal, regardless of how muchtime it is given to heal.17,18 Clinically, the term CSD isgiven to a defect that has not healed within 8 months ofinjury.19 Various animal models of bone repair with theanatomical capacity to regenerate a CSD have been devel-oped for biomaterials research,20,21 but the 5-mm rat cal-varial bone defect is one of the most frequently usedmodels in in vivo studies.22–24 Very recently, a systematicreview on CSDs of the calvarial model25 indicated that only1.6% of such 5.0-mm defects completely heal with newlyformed bone. Therefore, we employed this model of an in-tramembranous bone healing process to investigate the limitof osteoinductive capacity of a CaP/chitosan composite, andthe suitability of this rat model for the evaluation of bonetissue engineering strategies. We were especially interested tosee whether osteoinduction involved the specific formation ofosteons or similar kinds of histological evidence of corticalbone remodeling.

Materials and Methods

Biomaterial preparation

The biomaterial used in this work was a paste made fromsolid bioceramic powder and an aqueous chitosan solu-tion (2 wt%, pH = 4.5), which was purchased from Polimar

Cienciae Nutricao S.A. The bioceramic powder was amixture of b-TCP (Emprove�), calcium oxide, and zincoxide, all purchased from Merck�. The particle size ofeach powder was measured by laser granulometry using aMastersizer200 (Malvern Instruments). The analysis wasconducted using a laser diffraction liquid method on thefollowing suspensions: b-TCP dispersed in propenol andCaO and ZnO dispersed in water. The particle diameters ofb-TCP, CaO, and ZnO powders were 15, 3, and 9mm, re-spectively. CaO was incorporated mainly for adjustment ofpH of the composite, and ZnO was doped for its ability toinduce vascularization at early stages, as indicated by in-creased markers of osteoblast differentiation, matrix matu-ration, and bone mineralization in a previous work.26

A series of composites (100 samples) was systematicallyprepared from bioceramic powders of different b-TCP:CaO: ZnO ratios and mixed with the liquid chitosan solutionat various solid/(solid + liquid) percentages. The pH valuesof these composites were measured. The composites (20samples) with a pH value between 6.5 and 8.5 were con-sidered to be safe for biological environments, and the rest(80 samples) were discarded because of anticipated toxicity.Therefore, the ratio b-TCP: CaO: ZnO: chitosan of 38.4, 1.0,0.6, and 60 wt%, which had a pH value of *7.5, was thusused in the animal study.

The composite mixture was prepared as follows. Thethree weighted ceramic powders were gently mechanicallymixed for *5 min, and then the mixture was dried in amicrowave oven for *20 min. Each of these dried mixtureswas added to the chitosan solution according to the designedpercentage to produce a paste. The compressive strength ofthe paste was *0.5 MPa. The crystallinity of b-TCP was*30%, as provided by the supplier. The CaO and ZnOpowders were amorphous.

Animal model

Twelve male Wistar rats, which were 4-months old andweighed 300 g on average, were used. The rats were ran-domly divided into three groups of four rats, with the groupsto be examined respectively at 20-, 40-, and 60-day timepoints. Two bone defects (5 mm in diameter and full depth,which was *0.8 mm) were created in the rat skulls (controland experimental sites) ( b F1Fig. 1a). The experimental proto-cols and the animal care was approved and supervised by theAnimal Ethics Committee of the University of Valle (Cali,Colombia) and the University Autonoma de Occidente (Cali,Colombia).

Surgical procedure

Operations were performed on the rats using general an-esthesia, that is, ketamine (50 mg/mL; 0.7 mg/kg), xilacin(2%, 0.6 mg/kg), and acepromazinemaleate (10 mg/mL,0.6 mg/kg). Two circular bone defects were introduced us-ing a trephine bur with a dental implant surgical hand piece(400 rpm). After washing the rat skull with a physiologicalsaline solution, the right bone skull defect was filled withthe chitosan/ceramic paste, and the left one was left emptyas a control (Fig. 1b). The three groups of rats were sepa-rately sacrificed at 20, 40, and 60 days postimplantation. Tominimize the experimental differences between the rats, thesame experienced surgeon performed all the operations.

2 FERNANDEZ ET AL.


Histological and electron microscopysample preparation

The experimental bone defects were retrieved and fixed in2.5% glutaraldehyde Tris buffered saline (TBS) (0.01 M, pH7.4) solution over 48 h at 4�C. After washing with TBS(0.01 M, pH 7.4), control samples and two samples of eachexperimental group were decalcified in 15% ethylenediami-netetraacetic acid (EDTA) water solution (pH 7.3) at 4�C for2 months, and the other two samples of each experimentalgroup remained undecalcified. After rinsing in TBS (0.01 M,pH 7.4), the bulk decalcified and undecalcified samples were

dehydrated in increasing ethanol concentrations (70–100%)and embedded in paraffin wax for optical microscope his-tology or poly(methylmethacrylate) for electron microscopy.Then, 5-mm thick slides for histology and scanning electronmicroscopy (SEM) or 50–70 nm for transmission electronmicroscopy (TEM) were sectioned along the coronal plane(Fig. 1b) in cooling water with a microtome and an ultrathinmicrotome, respectively. Histological slides of decalcifiedbone were stained with hematoxylin-eosin (H&E), whereasthose of undecalcified bone was stained with Goldner’s

FIG. 1. (a) Surgical sites indicatedon the exposed rat calvaria and (b)diagram of the coronal view of ex-perimental site (ES-left) and controlsite (CS-right), which, with sur-rounding host bone, were retrieved atthe end of the experimental period.The histological samples were sec-tioned across the middle region ofeach sample along the coronal plane.Color images available online atwww.liebertpub.com/tea

FIG. 2. Low-power view of Goldner’s trichrome-stainedsections of the undecalcified defect tissue from rats that hadno implantation of biomaterial particles taken at (a) 20 (b)40, and (c) 60 days postsurgery. Almost no new bone for-mation was observed in the defects retrieved at day 20 and40 postsurgery. After 60 days, new bone was observed,occupying *20 Ar.% of the defect tissue. Color imagesavailable online at www.liebertpub.com/tea

FIG. 3. Low-power images of Goldner’s trichrome-stained sections of undecalcified CaP biomaterial-implantedtissue at (a) 20, (b) 40, and (c) 60 days after implantation.Tissues in Region I (see Fig. 5) mainly consisted of fibroustissues and CaP particles. Region II consisted of newlyformed bone. The blue/green stain observed in Region I andII was due to a high concentration of minerals produced bythe degradation of CaP particles. Region III consisted ofrelatively mature bone. The fields of the images coveredapproximately half of the defects around the center region.CaP, calcium phosphate. Color images available online atwww.liebertpub.com/tea

CALCIUM PHOSPHATE BIOMATERIAL INDUCES OSTEON FORMATION IN RAT 3


trichrome. H&E staining discriminates biomaterial particles,which stain purple or dark red; fibrous tissue, which stainslight pink, and decalcified bone, which stains orange or red.Goldner’s trichrome staining discriminates immature wovenbone (red) and mature lamellar bone (green/blue) in un-decalcified samples. Goldner’s trichrome stain is more sen-sitive to the level of mineralization in undecalcified samples,discriminating regions of a high or low mineral concentration,which stain green/blue or red, respectively.

Histomorphometric analysis

The above stained samples were imaged with an AperioScanCope� Turbo scanner (Aperio Technologies/Serial Num-ber AT1681). All scans were conducted at the same resolutionand magnification (i.e., 0.497 microns per pixel, 20 · ). Theimages of the region of interest were then processed withAdobe Photoshop� version CS2 (9.0) to obtain the masks. Inthis process, the selected features were designated in black, andthe rest of the examined area was designated in white, andthus, a black-and-white image (i.e., mask) was created. Im-ageJ� 1.46r software was used to measure the area percentageof the selected features in the examination field.27

SEM and TEM examination

The SEM samples were examined using a JEOL7001field emission gun SEM at 20 kV. Back-scattered electrondiffraction mode was used for elemental analysis. Energydispersive X-ray analysis was performed at 20 kV. The TEMfoils were examined with a Tecnai 20 microscope at 100–200 kV, depending on the materials present in the areas ofinterest.

Statistical analysis

All experiments were performed with at least six samplesper animal and 4 animals per experimental group, and the

statistical outputs are shown in the form of a mean withstandard error ( – SE). A one-way analysis of variance withTurkey’s post hoc test was performed to analyze the sig-nificant differences, and the significance levels were set at ap-value of less than 0.05.

Results

Histology and histomorphometric analysis

Analysis of undecalcified samples: trichrome stains. Allsurgical procedures were performed without complications.Histological examination (Goldner’s trichrome) revealedthat the control defects remained empty, with little new boneformed up to 40 days postimplantation ( b F2Fig. 2a, b), and theamount of new bone was considerable (*20 Ar.% of thedefects) only in the samples of the 60-day treatment group(Fig. 2c). These results are in agreement with previous workon bilateral calvarial defects that are 5.0 mm in diameter,which reported that the area percentage of new bone formedin the defects was 20.24% and 22.65% at 2 and 3 months,respectively.25

In contrast, the formation of new bone proceeded in allthe experimental sites implanted with the CaP-based bio-material, with defects being filled with newly formed softand/or hard tissue ( b F3Fig. 3). Undecalcified sections of the 20-and 40-day bone defect tissue samples stained predomi-nately red and green by Goldner’s trichome (Fig. 3a). Thethree regions (I, II, and III) were classified according to theirhistological characterization at high magnification, as shownin b F4Figure 4. Region I was occupied by a mixture of bio-material particles and fibrous tissue, with the former beingpredominant in the 20-day tissue samples (Fig. 4a), withfibrous tissue and blood vessels predominant in the 40-daysamples (Fig. 4b). In areas designated as Region I of the 40-day tissue samples, a high population of cells (fibroblasts)was observed in the brown area, with the nuclei stained dark

FIG. 4. High-power view of Re-gion I, II, and III of Figure 3 (un-decalcified). Region I was occupiedby the mixture of soft tissue andbiomaterial particles (BP)AU10 c , with BPbeing a particularly predominant tis-sue component in the 20-day samples(a) and fibrous tissue and blood ves-sels (BV) being the major componentin the 40-day samples (b). Region IIdisplayed a great deal of newlyformed bone characterized by pene-tration by many BV; large roundedosteocytes (OC) were also prominentand numerous (c). Region III (d)AU11 c

consisted of cortical bone (red). Thegreen/blue stain in the present sam-ples may have been caused by a highconcentration of minerals released byCaP particles, most of which lostquickly before precipitating intocrystalline bone hydroxyapatite. Thefour images have the same magnifi-cation. Color images available onlineat www.liebertpub.com/tea

4 FERNANDEZ ET AL.


brown (Fig. 4b). Region II was dominated by newly formedbone characterized by a high population of irregular-shapedcavities (Fig. 4c). Region III (red) was avascular corticalbone, characterized by canaliculi, with few irregular-shapedcavities (Fig. 4d). Almost no biomaterial particles could beobserved in Region II and, especially Region III under op-tical microscopy. The tissues of Regions I and II in the 40-

and 60-day samples were actually located above the originallevel of the skin and thus were soft and hard calli. The calluswould eventually be corrected on its own, and the normalbone contour would be restored.28

It should be mentioned that the green (blue) and red stainsin the present samples must be interpreted with a cautionbecause of the grafted CaP material. Goldner’s trichrome issensitive to the level of mineralization in bone and thus isused to discriminate newly formed relatively mature bone.With this staining method, nonmineralized (immature) bonestains red, and mineralized (mature) bone stains green/blue.29 The samples of the CaP-grafted group in the presentstudy, however, stained in the opposite manner. The newlyformed woven bone (Fig. 4c) and fibrous tissue (Fig. 4a, b),stained green/blue, whereas the relatively more mature bonein Region III stained red (Fig. 4d). This discrepancy couldbe attributable to the degradation of grafted CaP, whichwould have increased the local level of minerals in theearlier regions of repair in the implant sights. In the presentwork, the tissues in Region I and II stained green/blue be-cause of a high concentration of minerals newly released byCaP particles. It could be envisaged that if there had been noimplanted CaP, the woven bone of Region II would havestained red.

The red staining of the Region III bone at 40 and 60days (Fig. 3b and c) was possibly due to the quick loss ofCaP-associated minerals before the new avascular cortical

FIG. 5. Area percentages of Region I, II, and III afterretrieval at day 20, 40, and 60. Color images available on-line at www.liebertpub.com/tea

FIG. 6. Low-power views of H&E imagesof CaP biomaterial-implanted defects re-trieved at day 20 (a) and (b) day 40 post-implantation. Region I was characterized byfibrous tissue mingling with agglomerationsof BP. Region II was characterized by newlyformed bone that still had a relatively highporosity and a large amount of BV pene-tration. Region III was characterized bymature cortical bone. H&E, hematoxylin-eosin. Color images available online atwww.liebertpub.com/tea



bone was well calcified. Thus, the green/blue staining in thepresent work did not necessarily indicate a high degree ofbone maturation in Region I and II. In fact, the bone tissuein Region II (green/blue) was less mature than that in Re-gion III (red), as indicated by the histological characteristicsof sections stained by H&E (see the next section). However,the red staining of bone in Region III at 40 and 60 days (Fig.3b, c) indicated that the regenerated bone was still poorlymineralized.

F5 c Figure 5 shows the area percentages of Region I, II, andIII tissue (excluding the callus areas) as determined byhistomorphometric analysis using the samples stained by theGoldner’s trichrome method (undecalcified). Comparedwith the 20-day group, the area percentage of Region III(red) bone in the defects of the 40- and 60-day samplesincreased to nearly 100%, with tissues of Region I and IIpersisting as calli above the normal position of skin.

In short, the present grafted biomaterial was largely de-composed and replaced by new bone after 40 days of im-plantation, and the defects were nearly filled by newly formedbone. The regenerated dense bone was cortical bone that wasmarkedly less vascularized than at earlier time points.

Analysis of decalcified samples: H&E. b F6Figure 6 demon-strates the histological images of samples retrieved at days20 and 40 postimplantation. In the 20-day sample (Fig. 6a),the tissues of Region I-III presented in an order that wasconsistent with the growing direction of new bone, that is,from the border to the center of the defect. Figure 6b is aconsecutive series section of the field shown in Figure 3b(Goldner’s trichrome). The histological characteristics of thethree regions (I, II, and III) revealed by H&E staining werein agreement with those shown by the Goldner’s trichromemethod (Fig. 4), that is, the tissue of Region I was non-osseous fibrous tissue mixed with biomaterial particles. Thetissue of Region II was newly formed bone populated withmany irregular-shaped cavities. The tissue of Region III wasdominated by avascularized cortical bone containing veryfew irregular-shaped cavities.

Region I stained purple (or red) because of the presence ofbiomaterial particles. Chitosan appeared completely absorbedand was replaced by the fibrous tissue at day 20 post-implantation. However, the density of biomaterial particles wasunevenly distributed in Region I in the 20-day samples (Fig.6a). Large agglomerates of biomaterial particles remained in

FIG. 7. Three representative high-powerviews of H&E stained tissue in Region Iobserved in samples of CaP-biomaterialimplanted bone defects retrieved at 20 dayspostimplantation. (a) Agglomerations of

AU10 c bioceramic particles (BP) were surroundedby fibrous tissue. (b) Reduced particles weresurrounded by fibrous tissue with prominentBV, where giant cells (GC), presumedmacrophage polykaryons consistent with aforeign body reaction were noted in prox-imity to the BP. (c) The zone was dominatedby fibrous tissue. Color images availableonline at www.liebertpub.com/tea

6 FERNANDEZ ET AL.


the central area of the defect (F7 c Fig. 7a), and few agglomerateswere observed in the zone immediately next to Region II (Fig.7c). In between, the amount of biomaterial was reduced interms of area percentage (Fig. 7b). Giant cells were observedsurrounding biomaterial particles in Region I (Fig. 7b, c).

Many blood vessels were also observed in Region I, especiallyin the regions represented by Figure 7b and c. Histomorphometricanalysis indicated that the area percentages of biomaterial parti-cles in the three typical regions represented by Figure 7a–c were*40%– 5%,*25%– 10%, and < 2%– 0.5%, respectively. Thenumber of blood vessels per unit area in the region representedby Figure 7c was determined to be *85– 18/mm2, which issignificantly ( p < 0.05) higher than that in corresponding regionsof the control samples at day 20 postsurgery, which was*55– 14/mm2.

F8 c Figure 8a demonstrates the transition zone between RegionI and II, where new woven bone was growing into the fibroustissue of Region I and cavities were forming around bloodvessels, as marked by the blue arrows. It is apparent that bloodvessels within the irregular-shaped cavities in Region II (Fig.8b) were originally formed in the fibrous tissue of Region I.These blood vessels and nearby undifferentiated (mesenchyme-like) cells were wrapped by new bone, resulting in the irreg-ular-shaped cavities in the primary bone of Region II. Themesenchyme-like cells within these cavities subsequently ap-peared to have differentiated into osteoblasts, which depositedconcentric lamellar bone around the outside of the cavities,resulting in the formation of what we believe to be structuresvery similar in appearance to primary osteons (F9 c Fig. 9a).

A comparison of Figure 9a with Figure 9b (Goldner’strichrome stain) reveals that the primary lamellar bone(Zone 1) stained blue, most likely because the structure oflamellar bone was more mature than in woven bone. Thewoven bone (Zone 2) was likely stained red because of boththe immature structure and reduced mineral concentration inlamellar bone in Zone 1. The distal woven bone (Zone 3)stained blue because of the high concentration of minerals.

The formation of putative primary osteons was moreadvanced in Figure 8c than in Figure 8b. In Figure 8c, asindicated by the concentrically lamellar bone, some putativeprimary osteons still had blood vessel cavities at its center(Fig. 9c), whereas no cavities could be observed in others(Fig. 9d). The growth pattern of woven bone, the formationof irregular-shaped primary blood vessel canals, and thegeneration of primary osteons revealed in Figure 8 areconsistent with the early stage of bone remodeling.30,31

In Figure 8c, biomaterial debris was observed. These irreg-ular-shaped purple areas in between primary osteons appearedto have a high concentration of mineral ions, as indicated by theintense purple staining of these areas in contrast with the redstaining of the bone matrix stained. These purple-stained areaswere apparently linked to the formation of primary osteons,which involved redistribution of bone minerals.

The defects of the 60-day samples were predominatelyfilled with virtually nonvascularized cortical bone (RegionIII) with a small amount of immature bone (Region II) andvery little soft tissue (Region I) (F10 c Fig. 10a). Although nobiomaterial debris could be observed in the 60-day samples,the purple staining due to the unevenly distributed mineralscould still be observed around primary blood vessel canalsin Region II (Fig. 10b), indicating that the remodelingprocess was still going on inside the new bone.

FIG. 8. High-power view of H&E-stained decalcified softtissues of CaP biomaterial-implanted bone defects retrievedat 20 and 40 days postimplantation. (a) The transition zonebetween Region I and II. The blue arrows indicate thegrowing directions of woven bone. (b) Many irregular-shaped cavities in Region II b AU12. (c) Primary osteons wereobserved in Region II. The three images have the samemagnification. Color images available online at www.liebertpub.com/tea



SEM examination

The mineral distribution in Region II was analyzed underSEM (F11 c Fig. 11). Three serial consecutive sections of the sameregion were histologically examined with Goldner’s tri-chrome (Fig. 11a), H&E staining (Fig. 11b), and in SEM(Fig. 11c), separately. The blue zone around blood vesselcavity (marked as BV2) in Figure 11a was stained lightpurple (Fig. 11b). The same zone was bright in back-scat-tered images (Fig. 11c). Back-scattered imaging, which issensitive to the atomic weights, indicated that the brightzone around BV2 in Figure 11c contains more metal ele-ments, which might be Zn. Elemental mapping suggestedthat the concentration of calcium (Fig. 11d) is some-what high in the histological (H&E) purple zone, but thesodium concentration is markedly high (Fig. 11e), whereasthe carbon concentration was lower (darker in Fig. 11f ).One possible reason could be the formation of HA,Ca10(PO4)6(OH)2, in the primary osteon, which pushesexcessive Na + ions out of the osteon, and it may formcomplexes with trace metals.

TEM examination

Although under optical microscope, few particles wereobserved in Region I immediately next to Region II in 20-and 40-day samples (Fig. 7c), TEM examination revealedthat many submicro-sized particles presented in putativegiant cells in Region I (F12 c Fig. 12a and b). These submicro-sized particles were broken into nano-sized particles (Fig.12c) and had an amorphous structure (Fig. 12d).

Summary of major observations

1. The regeneration process of bone in rat calvaria dis-played three major regions. Region I: fibrovasculartissue infiltrated into the biomaterial-grafted region(Fig. 7). Region II: new woven bone, that is, primary

bone, regenerated and replaced the tissue of Region I(including the fibrous tissue and degraded biomaterial)(Fig. 8). Region III: woven bone restructured intoprimary lamellar bone (Fig. 10).

2. The regenerated cortical bone is primarily avascular.3. The degradation of CaP particles primarily occurred in

Region I. Large agglomerates were first infiltrated andfragmented in the fibrous tissue (Fig. 7), and individualparticles degraded to become submicro-sized ( £ 1 mm)particles, which were then phagocytosed by giant cells.Inside the cells, submicro-sized particles were furtherbroken down to nano-sized particles by lysosomes(Fig. 12).

4. The degradation of CaP stimulated the formation ofblood vessels in Region I (Fig. 7). In Figure 7, CaPparticles degraded slightly in Figure 7a and nearlyfully dissolved in Figure 7c. Meanwhile, the density ofblood vessels was higher in Figure 7c than in bothFigure 7a and b.

5. The bone remodeling process in the biomaterial graftswas similar to that in the femur of rodents30,31; that is,the primary blood vessels that formed in Region Iremained in the newly woven bone during its gener-ation, resulting in many large, irregular-shaped cavi-ties in the woven bone of Region II, and the primaryblood vessel canals were then encapsulated by primaryosteons (Fig. 8). However, no secondary osteons andblood vessels (i.e., a Haversian system) were observed.

6. During remodeling, the redistribution of minerals oc-curred, with excessive sodium being pushed out ofprimary osteons (Fig. 11).

7. The 5-mm defect was nearly completely filled withnew bone tissue around day 40 postimplantation andcompletely healed with nonvascularized cortical boneafter 60 days postimplantation. The present CaP ma-terial was completely degraded and replaced by newbone.

FIG. 9. High-power histologicalimages of bones and BV cavities inRegion II. (a) The framed area ofFigure 8b, H&E stain, (b) Gold-ner’s trichrome stain, and (c, d) theframed areas of Figure 8c, H&Estain. The four images have thesame magnification. H&E stainingapplied for decalcified samples andGoldner’s trichrome staining forundecalcified samplesAU13 c . Colorimages available online at www.liebertpub.com/tea

8 FERNANDEZ ET AL.


Discussion

Control of healing rate of rat calvarial defects

Bone healing kinetics are controlled by a combination ofphysical factors (such as availability of physical scaffolds,availability of bone minerals, and degradation rate of im-planted biomaterials) and biological factors (such as species,age, and anatomic position). Histological analysis indicatedthat Region I of biomaterial repaired rat CSDs was nearlycompletely replaced by new bone by day 40 (Fig. 3). As-suming the bone was growing at a nominally constant rate,the average growing speed of the bone is thus estimated tobe 2.5 mm/40 days = 62.5 mm/day& 1.0 · 10 - 3 mm/s. Thisrate is remarkably higher than the growth rate of normal

remodeling in rats of the same age ( b F13Fig. 13). The data inFigure 13, which was retrieved from literature,32 indicatesthat the bone growth rate in 4-month-old rats could be aslow as several mm/day, which has also been reported byanother study.33

An alternative way to quantify growth rate is weight perday. The density of bone is *2 g/cm3. The weight of bonein each defect (5 mm-diameter and 0.8 mm-thickness) is*0.03 g. The bone growth rate of CaP-grafted defects isthus estimated to be 0.03 g/40 days& 0.8 mg/day. The bonecontrol defects are estimated to have grown at a rate of0.03 g · 20%/60 days& 0.1 mg/day, showing that the im-planted defects repair at significantly higher rates than thebone growth rate of normal modeling in rats of the controlgroup.

On the other hand, the present study has also demon-strated that the micro-sized (*14 mm) biomaterial parti-cles degraded in fibrous tissue into submicron particles,which were then phagocytosed (Fig. 12). The dissolutionrate of a CaP particle in fibrous tissue is thus estimated asfollows. The dissolution rate of a particle is given by thefollowing:

dM

dt¼ AD

h(Csurface�Csolvent) (1)

where M is mass of material dissolved, t is time, D is dif-fusion coefficient of CaP in soft tissue, (D of inorganic ionsin aqueous environment is used as the first approximation,which is of 10 - 5 cm2/s order), A is surface area of theparticle, h is the thickness of the liquid film, typically being1.25 mm,34 and Csurface and Csolvent are the concentration ofthe material in the liquid film on the surface of the particleand the bulk medium, respectively. On the surface, Csurface

is the supersaturated solubility of CaP (in the form of Ca2 +

and/or PO43 - ions) in body fluid, which is *5.2 · 10 - 5

g/cm3.35 Csolvent is the concentration of dissolved CaP (inthe form of calcium and phosphate ions) in normal bodyfluid, which is > 4.8 · 10 - 5 g/cm3.36

Eq. (1) can be modified as follows:

dV

dt¼ dM=q

dt¼ AD=q

h(Csurface�Csolvent) (2)

where q is the density of CaP. Given that dV = Adr, where ris the radium of the particle, Eq. (2) becomes the following:

dr

dt¼ D

hq(Csurface�Csolvent) (3)

Considering that each layer of particles dissolves on bothsides, the dissolving rate of each layer is the double of Eq.(3), that is

dr

dt¼ 2D

hq(Csurface�Csolvent) (4)

A calculation using the above data reveals that the max-imal dissolution rate of CaP in ceremonial solution is dr/dt& 2.0 · 10 - 3 mm/s. The actual rate could be lower be-cause the concentration of calcium and phosphate ions in thebody fluid is likely higher than the minimal level 4.8 · 10 - 5

FIG. 10. Histological (H&E stained) images of decalcifiedCaP-implanted defect tissue samples retrieved at day 60postimplantation. (a) The 5-mm defects were almost com-pletely filled with bone tissue but exhibited two distinctareas. These include areas of newly formed cortical bonedisplaying significant porosity with prominent BV, and highnumbers of large osteocytes. A distinct area delineated by acement line resembled relatively more mature, low porositycortical bone. (b) In the area of the less mature bone, bio-material particles could not be clearly observed, but someirregular areas of purple staining may indicate areas con-taining remnants of these particles. Color images availableonline at www.liebertpub.com/tea



g/cm3, not only because the body fluid (such as cerebro-spinal fluid and blood serum) is normally very much su-persaturated with Ca3(PO4)2

35 but also due to the release ofthese ions from the grafted biomaterial. Thus, it is reason-able to conclude that the dissolution rate of amorphous CaPparticles is approximately the same as the growth rate ofhealing bone (our measurement). This indicates that thehealing rate of new bone in the first 40 days of the experi-ment was controlled by the dissolution kinetics of CaPparticles. It is envisaged that the released Ca2 + and PO4

3 -

ions from degradation CaP induced the regeneration of thebone matrix, and the associated mechanism directly deter-mined the growth rate of new bone.

However, dissolved CaP did not appear to enhance furtherbone remodeling because the new bone in the defects re-mained immature during days 40–60 postimplantation, asindicated by the red staining in Figure 3. The slow miner-alization process after 40 days might be attributed to twopossible reasons. First, the Ca2 + and PO4

3 - ions releasedfrom CaP could have been significantly reduced in RegionIII, possibly through the canaliculi (Fig. 4d). The secondpossible reason is associated with the lack of an osteonsystem in rats, which can slowdown the rate of bone re-modeling.

Comparison of the present observationswith the normal remodeling process of calvariain rats and humans

The histological structures and remodeling process offlat bone in rodents show important differences comparedto those of large animals, such as the human and bovine. Inhumans, flat bones are sandwich structures, composedof two thin layers of cortical bone and cancellous bonein between. The cancellous bone layer, also called diploe,is the location of bone marrow.28 The cortical bone layer isvascularized via the Haversian system of osteons, whereblood vessels are contained within osteocyte canaliculi. Theaverage thickness of (adult) human calvaria is *6 mm, witheach compact bone layer and the diploe being *1.8 and2.4 mm in thickness, respectively. The bone growth rate inhuman is typically several mm per day.28

In rats, the skull bones are not sandwich structured, butrather composed only of cortical bone.37 It has also beenreported that the bones of rats and mice lack a Haversiancanal system and thus lack Haversian remodeling.7,31 Al-though a low level of Haversian canals have been reportedfor long bones in rats, it has been consistently reported thatthin flat bones (e.g., the skull bone) in rats do not have a

FIG. 11. Histological and scan-ning electron microscope images oftissue taken from CaP particle-im-planted defects at 60 days post-implantation. (a) Goldner’strichrome, (b) H&E, (c) electronback-scattered image, elementalmapping images of calcium (d),sodium (e), and carbon (f ). Colorimages available online at www.liebertpub.com/tea

10 FERNANDEZ ET AL.


Haversian system.37 An investigation into the microcirculationof parietal, scapula, and ileum in rats convincingly demon-strated that the skull roofs of rats, which were 0.4–0.8 mmthick, did not contain a bone microcirculation, whereas bonemarrow sinusoids and cortical vessels similar to that of longbones were observed in thicker flat bones (e.g., ileum).37 Themicrovascular pattern of bones in rats is strongly influenced by

the bone thickness, reflecting the mass transportation andmetabolic requirements of bone remodeling, with more oxygendiffusion needed as density increases. The skull bone tissue inrats is thin enough that it can survive by feeding directly off theperiosteal network (rather than by canaliculi), which is sharedby the adjacent muscular microvasculature.37

It is interesting to note that the above threshold thicknessof *0.8 mm for the development of intra-bone vascularnetworks is reasonably consistent with the values reportedfrom the field of tissue engineering. Under static tissue cul-ture conditions, the maximum thickness of engineered tis-sue is 0.1–0.2 mm.38 Dynamic cultivations with perfusion ofculture medium through the construct, which enhances theconvective-diffusive oxygen supply, yield tissues of up to*1 mm in thickness.38 Hence, it has been predicted that massdiffusion is unable to foster tissue survival without bloodcirculation if the tissue is thicker than *1 mm.38 This limi-tation has been believed to be a critical issue in the engi-neering of thick, vascularized tissue for clinical applications,especially in the case of high-density tissues like bone.39

In the present study, the thickness of rat calvaria was in therange of 0.7–0.8 mm. The regeneration of new bone in theCaP-grafted defects occurred in three regions: the formationof fibrovascular tissue, growth of woven bone with irregular-shaped blood vessel cavities, and the temporary formation ofprimary osteons around the blood vessels. The final productwas nonvascularized cortical bone, with no sign of secondaryosteon formation. Our histological observations are in agree-ment with the structure of natural calvaria of rats, and the

FIG. 12. Transmission electronmicroscope (TEM) images of (a)biomaterial particles inside cells inthe transition zone between RegionI (fibrous tissue) and II (new boneformation) in samples retrieved atday 20 postimplantation, (b) theparticles near the nucleus of thecells, (c) a micro-sized particle hadbroken down into nano particles,and (d) the high resolution TEMimage showing that the nano-sizedmaterial was amorphous. L indi-cates a probable lysosome and Nindicates the cell nucleus. Similarmorphologies were also observedin the transition zone between Re-gion I and II in samples retrieved atday 40 (not shown).

FIG. 13. Bone (tibia) growth rate versus age of rats. Theequation represents the curve that fits with the data of malerats.AU14 c Data were retrieved from.33



implanted CaP did not induce further Haversian remodeling inthe present model. In addition, the regenerated calvaria boneat 60 days postimplantation was avascularized cortical typewith no cancellous tissue in the defect region. These resultsare also in agreement with the natural counterpart. It must bementioned again that once the cortical bone formed aroundday 40, further mineralization in the cortical bone was slowand was apparently not being enhanced by the grafted CaP,which is likely to be due a range of factors, including passivediffusion, decrease in the rate of cell proliferation as thewound approaches full regeneration, and delayed replacementof collagen with bone matrix. Longer term follow-up of tissueregeneration beyond 60 days may be useful in determining theactual duration of remineralization.

Nevertheless, the grafted CaP material greatly acceleratedthe bone growth rate at the initial stage (up to 40 days),resulting in a rate of 62.5 mm/day or 0.8 mg/day in thepresent study, which is remarkably higher than in the normalmodeling process in rats of the same age (several mm/day or0.1 mg/day). This may have been assisted by the temporaryformation of vessels, which would have increased perfusionof the wound, and increased diffusion of CaP away from thebiomaterial, supporting its biodegradation.

Since their bones heal rapidly, rats and mice have beenwidely used to investigate bone formation, including angio-genesis/vascularization in artificial bone materials, such astissue engineering scaffolds and cements.40–46 We would rec-ommend, however, that the usage of these species, especiallythe calvarial model, should be limited to the investigation ofinitial regions of bone formation and/or small flat bones, ratherthan the more complex process of human osteon. Therefore,we do not recommend that rat or mouse models be suitable foranimal studies aimed at engineering of thick bone for clinicalapplications, in which vascularization of the artificial bonematrix is critical. Our result of complete restoration, however,with short-term vasculogenesis and complete material degra-dation, suggests ours to be an ideal tissue engineering strategyfor encouraging native bone to repopulate large defects.

Summary

This article describes a histological investigation of thehealing process of CSDs in rat calvarial bone at both theoptical and electronic microscopic levels. The major con-clusions are summarized as follows: (1) Implanted CaPbiomaterials remarkably accelerated the bone growth rate ofthe defects at the initial stage, with a growth rate of*60mm/day or 0.8 mg/day, which is much higher than thebone growth rate (several mm/day or 0.1 mg/day) in non-implanted rats of the same age. (2) The CaP-based graftinduced histology similar to primary-osteon remodeling,which were eventually fully replaced by avascular corticalbone, similar to native at calvaria, indicating full defectrestoration. (3) The growth rate of new woven bone in theCaP-grafted defects was closely matched to the degradationrate of the biomaterial. Further work is required to explorethe combined osteoinductive and vasculogenic effects of ourbiomaterial on critical-sized defects in other species withvascularized compact bones.

Disclosure Statement

No competing financial interests exist.

References

1. Burg, K.J.L., Porter, S., and Kellam, J.F. Biomaterial de-velopments for bone tissue engineering. Biomaterials 21,2347, 2000.

2. Chen, Q.Z., Zhu, C.H., and Thouas, G.A. Progress andchallenges in biomaterials for tissue engineering. ProgBiomater 1, 2, 2012.

3. Geffre, C.P., Ochoa, J., Margolis, D.S., and Szivek, J.A.Evaluation of the osteogenic performance of calciumphosphate-chitosan bone fillers. J Invest Surg 23, 134,2010.

4. Thormann, U., Ray, S., Sommer, U., Elkhassawna, T.,Rehling, T., Hundgeburth, M., et al. Bone formation in-duced by strontium modified calcium phosphate cement incritical-size metaphyseal fracture defects in ovariectomizedrats. Biomaterials 34, 8589, 2013.

5. Bohner, M., Galea, L., and Doebelin, N. Calcium phosphatebone graft substitutes: failures and hopes. J Eur Ceram Soc32, 2663, 2012.

6. Chai, Y.C., Carlier, A., Bolander, J., Roberts, S.J., Geris,L., Schrooten, J., et al. Current views on calcium phosphateosteogenicity and the translation into effective bone re-generation strategies. Acta Biomater 8, 3876, 2012.

7. Ophoff, J., and Vanderschueren, D. The senile osteoporosismouse model SAMP-6: the ideal animal model for humanosteoporosis? BoneKEy-Osteovision 2, 26, 2005.

8. Laschke, M.W., Harder, Y., Amon, M., Martin, I., Farhadi,J., Ring, A., et al. Angiogenesis in tissue engineering:breathing life into constructed tissue substitutes. Tissue Eng12, 2093, 2006.

9. Habibovic, P., and de Groot, K. Osteoinductive biomaterials—properties and relevance in bone repair. J Tissue Eng RegenMed 1, 25, 2007.

10. Ignjatovic, N.L., Ajdukovic, Z.R., Savic, V.P., and Usko-kovic, D.P. Size effect of calcium phosphate coated withpoly-DL-lactide- co-glycolide on healing processes in bonereconstruction. J Biomed Mater Res B Appl Biomater 94,108, 2010.

11. Lu, J., Descamps, M., Dejou, J., Koubi, G., Hardouin, P.,Lemaitre, J., et al. The biodegradation mechanism of cal-cium phosphate biomaterials in bone. J Biomed Mater Res63, 408, 2002.

12. Lange, T., Schilling, A.F., Peters, F., Mujas, J., Wick-lein, D., and Amling, M. Size dependent induction ofproinflammatory cytokines and cytotoxicity of particu-late beta-tricalciumphosphate in vitro. Biomaterials 32,4067, 2011.

13. Balaguer, T., Boukhechba, F., Clave, A., Bouvet-Gerbettaz,S., Trojani, C., Michiels, J.-F., et al. Biphasic calciumphosphate microparticles for bone formation: benefits ofcombination with blood clot. Tissue Eng Part A 16, 3495,2010.

14. Yuan, H., van Blitterswijk, C.A., de Groot, K., and deBruijn, J.D. A comparison of bone formation in biphasiccalcium phosphate (BCP) and hydroxyapatite (HA) im-planted in muscle and bone of dogs at different time peri-ods. J Biomed Mater Res Part A 78A, 139, 2006.

15. Arinzeh, T.L., Tran, T., McAlary, J., and Daculsi, G. Acomparative study of biphasic calcium phosphate ceramicsfor human mesenchymal stem-cell-induced bone formation.Biomaterials 26, 3631, 2005.

16. Kim, I.Y., Seo, S.J., Moon, H.S., Yoo, M.K., Park, I.Y.,Kim, B.C., et al. Chitosan and its derivatives for tissueengineering applications. Biotechnol Adv 26, 1, 2008.

12 FERNANDEZ ET AL.


17. Schmitz, J.P., Schwartz, Z., Hollinger, J.O., and Boyan,B.D. Characterization of rat calvarial nonunion defects.Acta Anat (Basel) 138, 185, 1990.

18. Schmitz, J.P., and Hollinger, J.O. The critical size defect asan experimental model for craniomandibulofacial non-unions. Clin Orthop Relat Res 299, 1986.

19. Hollinger, J.O., and Kleinschmidt, J.C. The critical sizedefect as an experimental model to test bone repair mate-rials. J Craniofac Surg 1, 60, 1990.

20. Kochi, G., Sato, S., Fukuyama, T., Morita, C., Honda, K.,Arai, Y., et al. Analysis on the guided bone augmentationin the rat calvarium using a microfocus computerized to-mography analysis. Oral Surg Oral Med Oral Pathol OralRadiol Endod 107, e42, 2009.

21. Gosain, A.K., Song, L., Yu, P., Mehrara, B.J., Maeda, C.Y.,Gold, L.I., et al. Osteogenesis in cranial defects: reassess-ment of the concept of critical size and the expression ofTGF-beta isoforms. Plast Reconstr Surg 106, 360; discus-sion 72, 2000.

22. Mills, L.A., and Simpson, A.H. In vivo models of bonerepair. J Bone Joint Surg Br 94, 865, 2012.

23. Develioglu, H., Unver Saraydin, S., and Kartal, U. Thebone-healing effect of a xenograft in a rat calvarial defectmodel. Dent Mater J 28, 396, 2009.

24. Jones, L., Thomsen, J.S., Mosekilde, L., Bosch, C., andMelsen, B. Biomechanical evaluation of rat skull defects, 1,3, and 6 months after implantation with osteopromotivesubstances. J Craniomaxillofac Surg 35, 350, 2007.

25. Vajgel, A., Mardas, N., Farias, B.C., Petrie, A., Cimoes, R.,and Donos, N. A systematic review on the critical sizedefect model. Clin Oral Implants Res 2013 [Epub ahead ofprint]; DOI: 10.1111/clr.12194.

26. Fielding, G., and Bose, S. SiO and ZnO dopants in three-dimensionally printed tricalcium phosphate bone tissueengineering scaffolds enhance osteogenesis and angiogen-esis in vivo. Acta Biomater 9, 9137, 2013.

27. Egan, K.P., Brennan, T.A., and Pignolo, R.J. Bone histo-morphometry using free and commonly available software.Histopathology 61, 1168, 2012.

28. Thompson, J.C. Basic Science, 2nd edition. Saunders, 2009AU4 c .29. Spicer, P.P., Kretlow, J.D., Young, S., Jansen, J.A., Kasper,

F.K., and Mikos, A.G. Evaluation of bone regeneration usingthe rat critical size calvarial defect. Nat Protoc 7, 1918, 2012.

AU5 c 30. JD, C. Bones: Structure and Mechanism. 2002AU6 c .31. Martiniakova, M., Grosskopf, B., Vondrakova, M.,

Omelka, R., and Fabis, M. Observation of the microstruc-ture of rat cortical bone tissue. Scr Med 1, 45, 2005.

32. Hansson, L.I., Menander-Sellman, K., Stenstrom, A., andThorngren, K.G. Rate of normal longitudinal bone growthin the rat. Calcif Tissue Res 10, 238, 1972.

33. Raman, A. Appositional growth rate in rat bones using the tet-racycline labelling method. Acta Orthop Scand 40, 193, 1969.

34. Thompson, P.C. Dissolution equations, calculations andrelationships—a brief survey. Pharmaceutics 31, 2004AU7 c .

AU8 c 35. Emmett Holt, L., and Gittleman, I. The Solubility of ter-tiary calcium phosphate in cerebrospinal fluid. J Biol Chem64, 23, 1925.

36. Pincus, J.B., and Kramer, B. Comparative study of theconcentration of various anions and cations in cerebrospi-nal fluid and serum. J Biol Chem 463, 1923AU7 c .

37. Pannarale, L., Morini, S., D’Ubaldo, E., Gaudio, E., andMarinozzi, G. SEM corrosion-casts study of the microcir-culation of the flat bones in the rat. Anat Rec 247, 462,1997.

38. Carrier, R.L., Rupnick, M., Langer, R., Schoen, F.J., Freed,L.E., and Vunjak-Novakovic, G. Perfusion improves tissuearchitecture of engineered cardiac muscle. Tissue Eng 8,175, 2002.

39. Petrie Aronin, C.E., Sadik, K.W., Lay, A.L., Rion, D.B.,Tholpady, S.S., Ogle, R.C., et al. Comparative effects ofscaffold pore size, pore volume, and total void volume oncranial bone healing patterns using microsphere-basedscaffolds. J Biomed Mater Res A 89, 632, 2009.

40. Zanetti, A.S., Sabliov, C., Gimble, J.M., and Hayes, D.J.Human adipose-derived fstem cells and three-dimensionalscaffold constructs: a review of the biomaterials andmodels currently used for bone regeneration. J BiomedMater Res Part B Appl Biomater 101B, 187, 2013.

41. Tavassol, F., Kampmann, A., Schumann, P., Lindhorst, D.,Kokemueller, H., Essig, H., et al. A novel approach forstudying microcirculation in bone defects by intravitalfluorescence microscopy. Tissue Eng Part C Methods 17,1151, 2011.

42. Chen, M., Song, K., Rao, N., Huang, M., Huang, Z., andCao, Y. Roles of exogenously regulated bFGF expressionin angiogenesis and bone regeneration in rat calvarial de-fects. Int J Mol Med 27, 545, 2011.

43. Wernike, E., Montjovent, M.O., Liu, Y., Wismeijer, D.,Hunziker, E.B., Siebenrock, K.A., et al. VEGF incorpo-rated into calcium phosphate ceramics promotes vascular-isation and bone formation in vivo. Eur Cells Mater 19, 30,2010.

44. Roldan, J.C., Detsch, R., Schaefer, S., Chang, E., Kelantan,M., Waiss, W., et al. Bone formation and degradation of ahighly porous biphasic calcium phosphate ceramic inpresence of BMP-7, VEGF and mesenchymal stem cells inan ectopic mouse model. J Craniomaxillofac Surg 38, 423,2010.

45. Okuda, T., Uysal, A.C., Tobita, M., Hyakusoku, H., andMizuno, H. Prefabrication of tissue engineered bone graftsan experimental study. Ann Plast Surg 64, 98, 2010.

46. Naito, Y., Nagata, T., Tachibana, S., Okimoto, M., Ohara,N., Hakamatsuka, Y., et al. Locally applied TCP inhibitstumor growth via possible activation of macrophages. JBiomed Mater Res Part A 92A, 542, 2010.

b AU9Address correspondence to:Qi-Zhi Chen, PhD

Department of Materials EngineeringMonash Medical School

Monash UniversityClayton

Victoria 3800Australia

E-mail: [email protected]

Received: November 11, 2013Accepted: January 22, 2014

Online Publication Date:



AUTHOR QUERY FOR TEA-2013-0696-VER9-FERNANDEZ_1P

AU1: Please review all authors’ surnames for accurate indexing citations.

AU2: Please mention the authors’ degrees.

AU3: Please mention department, if any, in authors’ affiliations 2 and 3.

AU4: In Ref. 28, please mention the publisher’s location.

AU5: In Ref. 30, please expand author name JD.

AU6: In Ref. 30, please mention other publication details.

AU7: In Refs. 34 and 36, please mention the volume number.

AU8: In Ref. 35, author name ‘‘L. Emmett Holt JaWttaoIG’’ has been changed to ‘‘Emmett Holt, L., and Gittleman, I’’.

Please confirm.

AU9: Please confirm corresponding author’s address.

AU10: Please fix the expansion of the acronym BP: ‘‘biomaterial particles’’ or ‘‘bioceramic particles’’?

AU11: Label ‘‘(b)’’ has been changed to ‘‘(d).’’ Please confirm.

AU12: Please mention what the ‘‘dotted square’’ indicates in Fig. 8b.

AU13: Please mention what the ‘‘1, 2, and 3’’ indicate in Fig. 9.

AU14: The sentence ‘‘Data were retrieved.’’ appears incomplete. Please check.


Effects of Calcium Phosphate/Chitosan Composite on Bone Healing in Rats: Calcium Phosphate Induces Osteon Formation

Documents