evaluation of bone repair of critical size defects treated with simvastatin-loaded...

http://jba.sagepub.com/Journal of Biomaterials Applications

http://jba.sagepub.com/content/early/2014/09/10/0885328214550897The online version of this article can be found at:

DOI: 10.1177/0885328214550897

published online 10 September 2014J Biomater ApplLorraine B Ferreira, Vivian Bradaschia-Correa, Mariana M Moreira, Natasha DM Marques and Victor E Arana-Chavez

acid) microspheres in rat calvaria-glycoliccoEvaluation of bone repair of critical size defects treated with simvastatin-loaded poly(lactic-

Published by:

http://www.sagepublications.com

can be found at:Journal of Biomaterials ApplicationsAdditional services and information for

http://jba.sagepub.com/cgi/alertsEmail Alerts:

http://jba.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://jba.sagepub.com/content/early/2014/09/10/0885328214550897.refs.htmlCitations:

What is This?

- Sep 10, 2014OnlineFirst Version of Record >>

at Library - Periodicals Dept on November 15, 2014jba.sagepub.comDownloaded from at Library - Periodicals Dept on November 15, 2014jba.sagepub.comDownloaded from

http://jba.sagepub.com/

http://jba.sagepub.com/content/early/2014/09/10/0885328214550897

http://www.sagepublications.com

http://jba.sagepub.com/cgi/alerts

http://jba.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav

http://jba.sagepub.com/content/early/2014/09/10/0885328214550897.refs.html

http://jba.sagepub.com/content/early/2014/09/10/0885328214550897.full.pdf

http://online.sagepub.com/site/sphelp/vorhelp.xhtml



XML Template (2014) [8.9.2014–5:31pm] [1–12]//blrnas3/cenpro/ApplicationFiles/Journals/SAGE/3B2/JBAJ/Vol00000/140063/APPFile/SG-JBAJ140063.3d (JBA) [PREPRINTER stage]

Article

Evaluation of bone repair of critical sizedefects treated with simvastatin-loadedpoly(lactic-co-glycolic acid) microspheresin rat calvaria

Lorraine B Ferreira, Vivian Bradaschia-Correa, Mariana M Moreira,Natasha DM Marques and Victor E Arana-Chavez

Abstract

Purpose: Statins are hypolipemiant drugs with osteoinductive effect. We evaluated the potential of simvastatin loaded

into poly(lactic-co-glycolic acid) (PLGA) microspheres to heal critical size defects in rat calvaria.

Methods: PLGA scaffolds (50:50 ratio) were synthesized as pure membranes or as microspheres loaded with 2.5%

simvastatin. Critical size defects (5-mm diameter) were created in the parietal bone of 3-month-old male Wistar rats;

they were either left filled with blood clot (C group), covered with a PLGA membrane (M group) or with PLGA

microspheres loaded with simvastatin (MSI group) or not (MM group), and then covered with the PLGA membrane.

The defects were evaluated after 30 or 60 days by light and electron microscopy, immunohistochemistry for osteopontin

(OPN), bone sialoprotein (BSP) and osteoadherin (OSAD), and immunocytochemistry for OPN.

Results: Scanning electron microscopy showed that the calvarial defects treated with MSI were almost completely

healed after 60 days, while groups M and C presented less bone formation, whereas the bone matrix formed into the

defects of MSI group was more organized and mature. The immunolabeling for OPN and BSP on the matrix in groups C

and M showed typical areas of primary bone unlike the MSI that presented weak labeling at the formed area. In the MSI

group, there was an intense immunostaining for OSAD in osteoid, as well as in osteocyte cytoplasm. The immunocyto-

chemistry showed intense labeling for OPN with homogeneous distribution in the interfibrillar spaces in all groups after

30 days and after 60 days; however, while C and M groups exhibited similar aspect, the MSI specimens showed weak

labeling. The ultrastructural evaluation showed the interaction between the biomaterial and the surrounding tissue

where some cells established intimate contact with microspheres.

Conclusions: The repair of critical size bone defects was accelerated and enhanced by the implantation of simvastatin-

loaded PLGA microspheres.

Keywords

Bone regeneration, osteoconduction, osteoinduction, PLGA, simvastatin

Introduction

Large bone defects in maxillofacial region are resultantfrom trauma, tumor removals, and other surgical pro-cedures. The regeneration of such defects is challengingand usually requires the employment of bone grafts,which are considered the gold standard for bone regen-eration. However, patients frequently suffer from donorsite morbidity, besides the limited supply of bone avail-able for grafting.1 To overcome these limitations,

Journal of Biomaterials Applications

0(0) 1–12

! The Author(s) 2014

Reprints and permissions:

sagepub.co.uk/journalsPermissions.nav

DOI: 10.1177/0885328214550897

jba.sagepub.com

Department of Biomaterials and Oral Biology, School of Dentistry,

University of Sao Paulo, Sao Paulo, Brazil

Corresponding author:

Victor E Arana-Chavez, Department of Biomaterials and Oral Biology,

School of Dentistry, University of Sao Paulo, Av. Prof. Lineu Prestes 2227,

05508-900 Sao Paulo, SP, Brazil.

Email: [email protected]

at Library - Periodicals Dept on November 15, 2014jba.sagepub.comDownloaded from



biomaterials with osteoinductive and osteoconductiveproperties represent a promising strategy to fulfill thedemand of bone grafts that would substantiallyincrease the amount and quality of bone formed inshorter time.

Statins are widely used lipid-lowering drugs withcapability to inhibit the 3-hidroxy-3-methylglutarylcoenzyme A reductase that is the rate-limiting enzymein the synthesis of endogenous cholesterol, reducingprenylation of guanosine-triphosphate hydrolase likeras and Rho, which are important in cellular integrity,cytoskeleton, and vesicle traffic. These drugs have pleio-tropic effects in bone and are able to increase thein vitro expression of bone morphogenetic protein-2and vascular endothelial growth factor by osteoblasticcells.2–4 However, when administered systemically, itshigh uptake by the liver5,6 reduces its bioavailability inbone.7 Indeed, some previous studies have shown thattopical administration of statins was able to acceleratethe repair of bone defects in vivo.8–10

The vehicle from which statins are delivered caninterfere on the speed of the drug releasing. Previousstudies that compared the effect of gels containing sim-vastatin in the repair of bone defects with the carrieralone have confirmed that it delayed bone regener-ation.8,11 This finding suggests that statins could havea superior osteoinductive effect if associated to a vehiclethat would not interfere in the process. Poly(lactic-co-glycolic acid) (PLGA) is a synthetic biodegradable, bio-compatible, and water-soluble polymer12 that allowsthe adhesion of cells to its surface in vitro13 andin vivo.14 Previous studies in which simvastatin-loaded PLGA scaffolds were placed into bone defectshave shown that this biomaterial is able to improvebone neoformation.10,15 Similar results have enhancedthe formation of long bones fracture bridging.9

In order to verify the effect of simvastatin on boneregeneration, simvastatin-loaded PLGA microsphereswere inserted into critical size defects created in therat parietal bone avoiding the involvement of the inter-parietal suture and consequently a possible interferenceof undifferentiated cells present in the periosteumand the connective tissue contained at the suturearea.16 The presence of neoformed bone matrix insidethe defect was first verified by scanning electron micros-copy (SEM), and its structure was evaluated by lightmicroscopy (LM) and transmission electron micros-copy (TEM). Aiming to verify how simvastatin inter-fered in bone maturation, the distribution of thenoncollagenous bone matrix proteins osteopontin(OPN) and bone sialoprotein (BSP), as well as the pro-teoglycan osteoadherin (OSAD) in the newly formedwas verified by immunohistochemistry; OPN wasalso ultrastructurally detected by colloidal goldimmunocytochemistry.

Materials and methods

Synthesis of PLGA copolymer

PLGA copolymers have been synthesized by adding in aglass ampoule, the cyclic diester of L-lactic acid (Sigma–Aldrich), and the cyclic diester of glycolic acid (Sigma–Aldrich) in relation 50/50, respectively, together withthe catalyst Sn (Oct) 2 (Sigma). The mixture was frozenin liquidN2, the vacuumperformed in the vial, whichwasthen sealed and immersed in an oil bath at 110�C for oneweek. After this period, the copolymer was dissolved inchloroform (Merck), precipitated in methanol (Merck),and dried in a vacuum oven at 60�C for 8 h.17

Fabrication of PLGA microspheres loaded withsimvastatin

The PLGA was synthesized in the laboratory ofBiomaterials at Pontific Catholic University ofSao Paulo at Sorocaba, as previously described.13

Briefly, 1.0 g of PLGA granules were first dissolved in10.0ml of chloroform (Merck) until their total dissol-ution, followed by the addition of 100ml of aqueous1% PVA at room temperature. The solution was stirredfor 3 h to allow the evaporation of the solvent. Then, thesolution was centrifuged for 5min at 3500 rpm. The pro-cedure for obtaining the PLGA microspheres with sim-vastatin was carried out by adding simvastatin (Merck)with 5% acetone solution to PLGA for reaching a finalconcentration of 2.5%. Subsequently, the solution poly-mer–drug was subjected to emulsification with PVA 1%following the methodology described above. The micro-spheres with or without simvastatin were collected andfrozen in liquid nitrogen for 48 h prior to storage.17

Fabrication of PLGA membranes

PLGA membranes were prepared by dissolving thePLGA copolymers in chloroform (Merck) at the con-centration 5% by weighting the monomers, relative tothe volume of chloroform (g/ml), at room temperature.After complete dissolution, the solution was poured ona glass plate in a previously saturated with solventvapor in connection to an air stream tank. After 24 h,the membranes were removed from the plate and placedin a vacuum oven at 60�C for 8 h for complete removalof the solvent. After drying, the membranes were storedin a desiccator under vacuum. The membranesobtained were presented as translucent material.17

Animals and surgical procedure

The experimental protocol for this study was approvedby the Ethical Committee for Animal Research of the

2 Journal of Biomaterials Applications 0(0)




University of Sao Paulo, Brazil. Principles of labora-tory animal care (NIH publication 85–23, 1985) andnational laws on animal use were observed for the pre-sent study. The animals were sheltered in ventilatedplastic cages at 22�C on a 12 h light/dark cycle andreceived standard rodent chow and water ad libitum.

Sixty-six male Wistar rats (weight, 280–300 g each)were anesthetized with an intramuscular injection of0.86ml/kg Ketamin and 0.53ml/kg Xylazin. The surgi-cal site was shaved and disinfected with povidone–iodine, and a 20-mm long incision was created in thescalp to expose the parietal bones and the centralsuture line. The periosteum was reflected to expose theleft parietal bone in which a 5-mm diameter defect wascreated with a trephine drill under copious saline irriga-tion.18 The defects created were left filled by blood clot inthe control (C) group (n¼ 10), covered with PLGAmembrane (M) group (n¼ 20), or filled with 5mg ofsimvastatin-loaded PLGA microspheres and coveredwith PLGA membrane (MSI) group (n¼ 20). In orderto verify whether the PLGA scaffold alone would inter-fere in the wound healing, some animals (n¼ 6) weretreated only with PLGA microspheres without simvas-tatin and covered with membrane (MM) group (n¼ 10).The incision was sutured with 4-0 nylon suture. Thedefects were allowed to heal during 30 or 60 days.

Specimen obtaining, fixation, decalcification, andembedding

At the time points cited, the animals were anesthetizedas described and euthanized. The maxillae were fixed in0.1% glutaraldehyde and 4% formaldehyde buffered in0.1M sodium cacodylate, pH 7.4 under microwaveirradiation in a Pelco 3440 laboratory microwaveoven (Ted Pella, Redding, CA, USA) during threecycles with 100% potency and maximum temperatureof 37�C and then remained immersed in fresh fixativesolution overnight at 4�C.19 The specimens were decal-cified in 4.13% ethylenediaminetetraacetic acid for fourweeks and embedded in paraffin. Five-micrometer thicksections were obtained in a Micron HM360 microtomeand stained with hematoxylin and eosin. The slideswere examined in an Olympus BX60 light microscopeequipped with an Olympus DP72 CCD camera. Somesections were collected onto silane-coated slides andstored for further immunohistochemistry reactions.

Immunohistochemical detection of OPN/BSP/OSAD

The sections submitted to immunohistochemical detec-tion of OPN/BSP/OSAD were dewaxed and then trea-ted with H2O2/methanol solution (1:1) during 20min.The nonspecific binding sites were blocked during 1 hwith 10% nonimmune swine serum (Dako North

America Inc., Carpinteria, CA, USA) in bovine serumalbumin. Then, they were incubated with the primaryantibodies rabbit anti-mouse OPN (LF175 1:1200,Larry Fischer, NIH for 2 h), rabbit anti-human BSP(LF-84 1:1000, Larry Fischer, NIH for 2 h), or rabbitanti-rat OSAD, (1:500 for 12 h)20 at room temperaturewithin a humid chamber. After rinsing with buffer,detection was achieved using diaminobenzidine as sub-strate and counterstained with Harris’s hematoxylin.Samples were dehydrated and mounted in entellanand the slides examined in the Olympus BX60 lightmicroscope. Negative controls were done by omittingprimary or secondary antibodies.

TEM

To further explore the interactionbetween the biomaterialand cells, the specimens destined for ultrastructural ana-lysis were post fixed in 1%osmium tetroxide, dehydrated,and embedded in Spurr epoxy resin (EMS, Hatfield, PA,USA). Some specimens were left unosmicated andembedded in London Resin White acrylic resin forimmunocytochemistry analyses. Sections 80-nm thickwere obtained with a diamond knife on a Leica UltracutR ultramicrotome (Leica, Buffalo, NY, USA), collectedonto 200-mesh copper grids, stained with uranyl acetateand lead citrate, and examined in a Jeol 1010 transmissionelectron microscope operated at 80kV. The images weredigitally obtained with the GATAN imaging platformequipped with a SC1000 Orius CCD camera.

Immunocytochemistry for OPN

Some ultrathin sections from London Resin Whiteembedded specimens were collected onto 150-mesh par-lodion-coated nickel grids and incubated for OPN(LF175 1:20, Larry Fischer, NIH) for 12 h. The bindingsites were revealed by incubating the grid with proteinA-gold complex (1:25) for 40min.21 Negative controlswere done by omitting primary antibody or the proteinA-gold complex.

SEM

The calvariae from three animals of groups C, M, andMSI at 60 days were left undecalcified and immersed insodium hypochlorite at 2% for three periods of 20minfor removing the soft tissues. They were subsequentlywashed in distilled water, dehydrated in graded concen-trations of ethanol, and then treated with hexamethyl-disilazane for 10min and left to dry within a fume hoodsystem. The specimens were mounted on aluminumstubs, gold coated in a Bal-Tec SCD 050 apparatus,and examined in a 450 LEO scanning electron micro-scope operated at 15 kV.

Ferreira et al. 3




Results

SEM

Scanning electron micrographs showed evident boneformation in the defect that was treated with simvas-tatin when compared with the defect filled with clotalone. The C defects seen after 60 days of healingpresented few areas of neoformed bone that wererestricted to the border of the defect (Figure 1(a)).In the M group, the newly formed bone matrix cov-ered the border of the defects that were not distin-guished; some short trabeculae are seen inside thedefects (Figure 1(b)). Differently, the defects of theMSI group were almost completely filled with bonematrix (Figure 1(c)). When the specimens were exam-ined at higher magnifications, the neoformed bonefrom C and M specimens, respectively, showed itssurface with some osteocyte lacunae and its minera-lized matrix with irregular aspect (Figure 1(d) and(e)). The neoformed bone in the defect of the MSIspecimens did not present osteocyte lacunae, whilethe mineralized matrix exhibited a smoother surfacewith lamellar aspect compared to the other groups(Figure 1(f)).

Hematoxylin and eosin staining

The histological analysis by LM showed all specimenswithout inflammation signs.

The bone defects created in the rat calvaria exam-ined 30 days after surgery showed, in the C group, athin layer of neoformed bone matrix adjacent to theborder of the defect and over the external surface ofthe calvaria (Figure 2(a)). At 60 days, there was a slightincrease in the amount of neoformed bone adjacent tothe border of the defect (Figure 2(b)).

The M group at 30 days after surgery presentedmore bone matrix formed into the defects than thatobserved in the C group (Figure 2(c)). At 60 days, how-ever, the neoformed bone into the defects appearedsimilar to that observed at the first time point (Figure2(d)).

Thirty days after surgery, the defects from the MSIgroup clearly showed more neoformed bone matrixinside the defect than that of the previous groups. Itwas surrounded by numerous empty small spaces thatcorresponded to the microspheres at both the internalside facing the duramater as well as at the outer surface,below portions of the membrane (Figure 2(e)). At 60days, the neoformed bone filled almost completely the

Figure 1. Scanning electron micrographs illustrating the calvarial surface in which the defects from representative specimens are

seen after 60 days of surgery. The C specimen presents, in (a), few areas in which neoformed bone (arrows) is seen at the border of

the defect. In the M group (b), the border of the defect is not distinguished; newly formed bone matrix is covering the border of the

defect and some trabeculae are seen inside the defect. The defect of the MSI group (c) is almost completely filled with bone matrix

(asterisk). In (d) and (e), higher magnifications of the surface of the neoformed bone from C and M specimens, respectively, show some

osteocyte lacunae (Lc) and irregular aspect of the mineralized matrix. The neoformed bone in the defect of the MSI specimen, in (f),

does not present osteocyte lacunae and the mineralized matrix presents a smoother surface with lamellar aspect compared to the

other groups. Bars: (a–c) 1 mm; (c–f) 20mm.





central portion of the defects where it appears thickand constituted by several trabeculae with lamellaraspect. Few microspheres were still identified into theconnective tissue located between the bone trabeculae(Figure 2(f)).

The bone defects from the MM group presentedneoformation similar to the C group at 30 days aftersurgery, despite the presence of numerous microspheres(Figure 2(g)). At 60 days, however, the amount of

neoformed bone around the border of the defects wasonly slightly greater than that of group C at this timepoint (Figure 2(h)).

Immunohistochemistry for OPN/BSP/OSAD

In all groups at 30 days, the OPN and BSP immunola-beling presented that usually observed in primary bone:it was seen in all the cement lines, i.e., in the cement line

Figure 2. Light micrographs of the border of the bone defects created in the rat parietal bone. At 30 days, the C group present

neoformed bone matrix (Nb) adjacent to the border of the defect (arrows) and over the external surface of the calvaria. The M group

presents more bone matrix formed into the defect than C. The defect of MSI group shows a large amount of neoformed bone matrix

inside the defect, surrounded by the numerous microspheres (ms). M, PLGA membrane. The MM group presents bone neoformation

similar to the C group at the same time point. At 60 days, there is a slight increase in the amount of neoformed bone adjacent to the

border of the defect in C; the amount of neoformed bone into the defect of M group is similar to that from the same group at the first

time point. The MSI group shows a thick cortical bone neoformed in the central portion of the defect with organized aspect. The

amount of bone formed in the MM group at this time point is superior compared to groups C and M at this time point. Bars: (a, b, g)

100 mm (c–f, h) 50mm.

Ferreira et al. 5




at the border of defect and in other appositional andreversal cement lines. These noncollagenous proteinswere distributed in some regions of the matrix generallyforming patches, as well as with disorganized appear-ance (Figure 3(a) to (c) and (g) to (i)).

At 60 days, the C and M groups both OPN and BSPpresented distribution similar to the first time point(Figure 3(d), (e), (j), and (k)); the MSI group, however,exhibited few areas of OPN and BSP immunolabeling;they were located between abundant lamellar bone

Figure 3. Immunohistochemistry for OPN, BSP, and OSAD. In all groups at 30 days, the OPN and BSP immunolabeling, in brown,

presented primary bone features; it is seen at the cement lines at the border of defect (BD) – neoformed bone interface, in other

appositional cement lines (arrows), as well as dispersedly distributed in some regions of the matrix with disorganized appearance

(arrowheads). At 60 days, the C and M groups presented OPN and BSP distribution similar to the first time point; the MSI group

presented few areas with primary bone OPN and BSP immunolabeling pattern, and abundant lamellar bone matrix weakly labeled for

these noncollagenous proteins is observed near the microspheres (ms). The OSAD immunolabeling appears strong at the connective

tissue in C and M specimens at 30 days, while the MSI connective tissue is weakly labeled. Specimens from all groups present

immunolabeling for OSAD at the osteoid and at osteocyte lacunae at this time point. At 60 days, the OSAD distribution in the connective

tissue and the osteoid and inside osteocyte lacunae was observed in all groups. Bars: (a)100 mm; (b–k) 50mm; (l) 200 mm; (m–r) 50mm.





matrix that was weakly labeled for these noncollagen-ous proteins (Figure 3(f) and (l)).

The OSAD proteoglycan appeared strongly immu-nolabeled in the connective tissue surrounding theforming bone in C and M specimens at 30 days(Figure 3(m) and (n)), while the MSI connective tissuewas weakly labeled. The osteoid and osteocyte lacunaeat this time point were immunolabeled for OSAD inspecimens from all groups (Figure 3(o)). At 60 days,the OSAD distribution in the connective tissue andthe osteoid, as well as in the osteocyte lacunae appearedmoderately labeled in all groups, similar to that of theMSI group at the first experimental period (Figure 3(p)to (r)).

TEM

Ultrastructural examination from the C group 30 daysafter surgery showed areas of bone formation in whichsecreting osteoblasts appeared adjacent to a layer ofosteoid with numerous typically banded collagen fibrils.At this time point, the mineralized matrix exhibited aglobular surface that revealed the progression of min-eralization (Figure 4(a)). Specimens from the MSIgroup at the same experimental period presented thePLGA microspheres containing a slightly flocculentmaterial often surrounded by cells with fine processes,while a collagenous matrix was seen between the micro-spheres (Figure 4(b)). Some areas of the primary bonetrabeculae from the MSI group were resorbed by typ-ical osteoclasts (Figure 5).

Sixty days after surgery, the specimens from the Cgroup presented osteoblasts secreting new bone matrixover reversal cement lines (Figure 6(a)). The adiposetissue adjacent to regions of bone formation containedtypical adipocytes containing empty fat globules(Figure 6(b)). The regions of primary bone from the

M group were characterized by poorly packed collagenfibrils with numerous interfibrillar spaces (Figure 6(c)).The specimens from the MSI group exhibited regions ofdegrading PLGA copolymer with an amorphousappearance, surrounded by collagen fibrils (Figure6(d)). In other regions, microspheres with a somewhatflocculent material were surrounded by cells with fineprocesses, as it was seen at the previous time point. Inthis period, however, more cell processes were notedaround the microspheres, some of them resemblingpodocyte processes; the surrounding matrix containedtypical collagen fibrils (Figure 6(e)). In addition, someregions at the center of the defect and below the mem-brane in which areas of matrix with strong electronopacity were also detected in this group; they are

Figure 4. Transmission electron micrographs showing, in (a), a region of bone formation from the C group 30 days after surgery.

Observe part of an osteoblast (Ob) on a layer of osteoid (Ost) containing numerous typically banded collagen fibrils. The mineralized

matrix (mm) exhibits a globular surface that reveals the progression of mineralization. In (b), a PGLA microsphere that contains a

slightly flocculent material (asterisk) appears surrounded by fine cell processes (arrows). A collagenous matrix (Col) surrounds the

microsphere. Bars: (a) 1mm and (b) 2mm.

Figure 5. Transmission electron micrograph of a resorbing

surface of a primary bone trabecula (Pb) from the MSI group.

A typical osteoclast (Ocl) with clear zone (arrows) and ruffled

border (asterisk) appears attached to a resorbing lacuna.

Bar: 5 mm.

Ferreira et al. 7




formed by densely packed collagen fibrils that weredevoid of the typical banding and were surroundedby regular collagen fibrils and by typical fibroblasts(Figure 6(f)).

Immunocytochemistry for OPN

The ultrastructural immunogold detection of OPN inthe neoformed bone at 30 days after surgery confirmedthe immunohistochemical labeling for this noncollagen-ous protein: specimens from all groups presented goldparticles on the cement lines, as well as forming patchesat the interfibrillar regions (Figure 7(a) to (c)). At 60

days, the labeling pattern in the C and M groupsremained over cement lines and interfibrillar regions(Figure 7(d) and (e)), while in the MSI group, thegold particles were visible dispersed through thematrix instead forming patches as in the previousperiod (Figure 7(f)).

Discussion

The present study showed that simvastatin loaded intoPLGA microspheres enhanced the bone formation intocritical size defects created in the rat parietal bone. Theneoformed bone in intimate contact with the degrading

Figure 6. Transmission electron micrographs showing representative specimens 60 days after surgery. In (a), part of an osteoblast

(Ob) with a newly deposited bone matrix (mm) over a reversal cement line (stars). In (b), a region of adipose tissue with a typical

adipocyte (Acy) surrounding an empty fat globule. In (c), a region of primary bone from the M group characterized by poorly packed

collagen fibrils surrounding an osteocyte lacuna (Ocy). In (d), a portion of degrading PGLA copolymer that appears amorphous

(asterisk), surrounded by collagen fibrils (Col) and a peripheral mineralized matrix (mm) from the MSI group. In (e), part of two PGLA

microspheres that contain a slightly flocculent material (asterisk). They are surrounded by fine cell processes (arrows), some of them

resembling podocyte processes (arrowheads). Collagen fibrils (Col) and fine cell processes surround the microspheres. In (f), a region

from the MSI group at the center of the defect, below the membrane. Observe a fibroblast (Fb) surrounded by few collagen fibrils in a

cross section (Col) adjacent to a region of densely packed matrix that acquires a strong electron opacity devoid of the typical

collagenous banding. Bars: (a, c, e) 1 mm; (b) 5 mm; (d, f) 2 mm.





copolymer, as well as the lamellar aspect of its miner-alized matrix suggested the combined osteoinductiveand osteoconductive effects of this biomaterial.

A critical size defect was created in an only parietalbone in which the diameter (5mm) almost reached theboundaries of the bone, thus avoiding the involvementof the interparietal suture. It is well known that undif-ferentiated cells are present in the periosteum and theconnective tissue contained at the suture area.16

Therefore, the regeneration of a defect that includesboth parietal bones could be interfered by the suturaltissue. Thus, the formed bone in the M and MSI groupswas the result of the both strategies employed forenhancing the bone regeneration into the critical sizedefect. In this context, the great amount of formedbone, noted early at the first time point in the MSIspecimens, as well as its lamellar aspect observed atthe last experimental period revealed the osteoinductiveeffect of the biomaterial.

These results also could be the consequence of thetotal amount of drug delivered to the defect, whereas

previous studies reported that microspheres, which areprepared from solvent emulsification process, generallyshow an initial burst release; a subsequent medium orlong release depend on the drug–polymer interaction.22

Simvastatin, an inhibitor of the cholesterol synthesis,was topically applied to enhance the regeneration ofbone defects in the rat mandible using a gel as thevehicle.8 However, the stability of a gel is not guaran-teed in large defects as the one created in the presentinvestigation. The microspheres of PLGA copolymerplaced into the created defect were covered with a mem-brane in order to follow the largely applied procedureof guided tissue regeneration. The PLGA copolymercontaining simvastatin constituted a controlled systemfor releasing the simvastatin loaded at a concentrationof 2.5%8 into the bone defect. Indeed, it has beendemonstrated that statins are processed in the liverwhen they are systemically administered, thus remain-ing low amounts available for reaching bone23 that sup-ports the attempts to analyze its topical application. Inaddition, degradation of the PLGA copolymer depends

Figure 7. Ultrastructural immunocytochemical detection of OPN in the neoformed bone during the healing of calvarial defects. At

30 days, C and M group present gold particles localized in cement lines (arrows). The C and MSI specimens present abundant bold

particles in interfibrillar spaces (IFS) of the bone matrix. Ob, osteoblast. At 60 days, the C group presents cement lines with numerous

gold particles; the M group presents a bone matrix containing irregularly arranged collagen fibrils with some interfibrillar spaces

immunopositive for OPN. Lc, osteocyte lacune. MSI 60 days showed disperse gold particles in the mature bone matrix. Bars: (a, d)

2 mm; (b, c, e) 1mm; (f) 1000 nm.

Ferreira et al. 9




on the proportion of both PLA and PGA. The PLGAcopolymer at concentrations 50/50 allows an adequatedegrading speed, while new bone forms into the defect,as previously shown.24 In addition, the influence of thesize of the particles (microspheres) on the biomaterialdegradation was adequate for maintaining its osteocon-ductive properties, as showed when they were testedwithout simvastatin. Smaller particles (nanospheres)could degrade too fast due to their increased area sur-face24,25 yielding an altered drug releasing.

The bone defects filled with simvastatin-containingPLGA microspheres and covered with membraneshowed more newly formed bone not only in compari-son with the control specimens but even when com-pared with the defects covered only with membrane.These findings were demonstrated by all the approachesat the first time point. The SEM examination revealedthat the defects were filled by a mineralized matrix thatLM and immunohistochemistry for OPN and BSP fur-ther showed to be a primary bone with abundantpatches of noncollagenous proteins. On the otherhand, the small leucine-rich proteoglycan OSAD wasless intense in the connective tissue adjacent to regionsof lamellar bone in the MSI group. As OSAD appearsin more amounts at stages in which initial mineraliza-tion is taking place,26 it could also reflect some simvas-tatin effect by which bone remodeling could start earlywhen this drug is released.

Besides the osteoinductive effects of simvastatinshowed in previous studies,7,8,10,27 the PLGA copoly-mer in the form of microspheres could additionallyallow the spreading and migration of differentiatingosteogenic cells between the small particles of thedegrading copolymer. Indeed, it has been establishedthat migration and differentiation of cells require thatthe vehicle used is osteoconductive.28

Small areas of bone formation at central regions ofthe created defect showed the osteoinductive effect ofsimvastatin released by the PLGA microspheres.Whereas bone repairing of defects usually takes placeat the border regions, in the case of parietal bone, itoccurs preferentially at the inner surface of its borders,i.e., from the dura mater side that appears to containgreater osteogenic capability than that of the outer peri-osteum.29 Maciel-Oliveira et al.8 detected more boneformation at the side facing the periodontal ligamentrather than at the outer surface of noncritical defectscreated in the rat alveolar process as they tested a gel-containing simvastatin. Since the defects created in thepresent investigation were critical, it is conceivable thatthe areas of early bone formation at the central area ofthe defect could be triggered by the osteoinductiveeffect of this biomaterial.

In addition, some areas resembling regions of miner-alized matrix were detected in the proximities of the

forming bone. When mineralized areas are decalcifiedduring processing for TEM, they appear denselypacked and most of them without the typical bandingof regular collagen type I fibrils. It is because whencollagen fibrils mineralize, the mineral crystals breaktheir regular arrangement.30 Interestingly, no osteo-blasts were observed surrounding these areas, contrar-ily to what is observed at the onset of osteogenesiswhere a continuous layer of osteoblasts compartmen-talizes the forming and mineralizing bone matrix.31

Since these areas were only observed in the MSIgroup, it is possible that simvastatin could influencethe deposition of mineral crystals in some processesindependently of cell regulation.

The neoformed bone observed in the defects fromthe MSI group exhibited a lamellar aspect and numer-ous cement lines, especially those from the reversaltype. It indicates that bone remodeling took placebetween the examined time periods. Examination atthe later time point revealed that the repairing boneeven at central regions of the defect was composed bydensely packed collagen fibrils with few and sparse non-collagenous matrix proteins between them. Thus, it ispossible that remodeling took place more rapidly thanin the C and M groups. Indeed, typical active osteo-clasts were observed from the first experimental timepoint in the MSI group.

The ultrastructural analysis confirmed that someregions of adipose tissue were adjacent to the repairingarea in control specimens. Areas of adipose tissue areoften present adjacent to regions of forming bone.32

However, we were able to discern that areas in theMSI group that appeared under LM resembling adi-pose tissue, actually corresponded to degrading micro-spheres. In fact, the degrading microspheres weresurrounded by cells that embraced them by fine cellprocesses that sometimes resembled the podocyte pro-cesses present at the glomerular capillaries in the renalcorpuscles.33 The formation of adipose tissue in areasof repairing bone is considered an impairing factor inosteogenesis; on the other hand, simvastatin has beenshown to induce osteoblastic differentiation, while inhi-bits adipocytic differentiation.34 The ultrastructure alsorevealed the presence of collagenous matrix between thedegrading microspheres confirming the osteoconduc-tive properties of the biomaterial, as previouslyobserved.23 The osteoinductive effect was demonstratedby the high amount of repairing bone in the MSI group.

In summary, our data suggest that the local deliveryof simvastatin stimulates the earlier regeneration of ratparietal bone, associated with the improvement of bonequality. In addition, topical application could be a con-venient procedure in surgery, since the material used inthe present study showed favorable properties for boneregeneration with high biocompatibility.





Acknowledgments

The authors thank Drs. Eliana R. Duek and Newton Maciel-

Oliveira for their help with the synthesis of the simvastatin/PGLA microspheres.

Declaration of conflicting interests

None declared.

Funding

This study was supported by CAPES/Brazil and partially sup-ported by CNPq and FAPESP, Brazil.

References

1. Scaglione S, Lazzarini E, Ilengo C, et al. A compositematerial model for improved bone formation. J Tissue

Eng Regen Med 2010; 4: 505–513.2. Garrett IR, Gutierrez G and Mundy GR. Statins and

bone formation. Curr Pharm Des 2001; 7: 715–736.

3. Mundy G, Garrett R, Harris S, et al. Stimulation of boneformation in vitro and in rodents by statins. Science 1999;286: 1946–1949.

4. Davignon J and Laaksonen R. Low-density lipoprotein-independent effects of statins. Curr Opin Lipidol 1999; 10:543–559.

5. Jadhav SB and Jain GK. Statins and osteoporosis: Newrole for old drugs. J Pharm Pharmacol 2006; 58: 3–18.

6. Oxlund H and Andreassen TT. Simvastatin treatmentpartially prevents ovariectomy-induced bone loss while

increasing cortical bone formation. Bone 2004; 34:609–618.

7. Skoglund B and Aspenberg P. Locally applied simvasta-

tin improves fracture healing in mice. BMCMusculoskelet Disord 2007; 27: 98.

8. Maciel-Oliveira N, Bradaschia-Correa V and Arana-

Chavez VE. Early alveolar bone regeneration in ratsafter topical administration of simvastatin. Oral SurgOral Med Oral Pathol Oral Radiol Endod 2011; 112:170–179.

9. Tai IC, Fu YC, Wang CK, et al. Local delivery of con-trolled-release simvastatin PLGA/HAp microspheresenhances bone repair. Int J Nanomed 2013; 8: 3895–3904.

10. Naito Y, Terukina T, Galli S, et al. The effect of simvas-tatin-loaded polymeric microspheres in a critical sizebone defect in the rabbit calvaria. Int J Pharm 2014;

461: 157–162.11. Stein D, Lee Y, Schmid MJ, et al. Local simvastatin

effects on mandibular bone growth and inflammation.

J Periodontol 2005; 76: 1861–1870.12. Mano JF, Sousa RA, Boesel LF, et al. Bioinert, bio-

degradable and injectable polymeric matrix compositesfor hard tissue replacement: state of the art and recent

developments. Compos Sci Tech 2004; 64: 789–817.13. Barbanti SH, Santos AR Jr, Zavaglia CA, et al. Porous

and dense poly(L-lactic acid) and poly(D,L-lactic acid-co-

glycolic acid) scaffolds: in vitro degradation in culturemedium and osteoblasts culture. J Mater Sci MaterMed 2004; 15: 1315–1321.

14. PietroL, SilvaDR, doCarmoAlberto-RinconM, et al. Themicroscopical characterization of membranes poly(L-gly-colic-co-lactic acid) with and without added plasticizer: an

in vivo study. J Mater Sci Mater Med 2008; 19: 1069–1071.15. Chang PC, Chung MC, Lei C, et al. Biocompatibility of

PDGF-simvastatin double-walled PLGA (PDLLA)microspheres for dentoalveolar regeneration: a prelimin-

ary study. J Biomed Mater Res A 2012; 100: 2970–2978.16. Opperman LA. Cranial sutures as intramembranous

bone growth sites. Dev Dyn 2000; 219: 472–485.

17. Garrett IR, Gutierrez GE, Rossini G, et al. Locally deliv-ered lovastatin nanoparticles enhance fracture healing inrats. J Orthop Res 2007; 25: 1351–1357.

18. Cameron JA, Milner DJ, Lee JS, et al. Employing the biol-ogy of successful fracture repair to heal critical size bonedefects. Curr Top Microbiol Immunol 2013; 367: 113–132.

19. Massa LF and Arana-Chavez VE. Ultrastructural pres-ervation of rat embryonic dental tissues after rapid fix-ation and dehydration under microwave irradiation. EurJ Oral Sci 2000; 108: 74–77.

20. Couble ML, Bleicher F, Farges JC, et al.Immunodetection of osteoadherin in murine tooth extra-cellular matrices. Histochem Cell Biol 2004; 121: 47–53.

21. Arana-Chavez VE and Nanci A. High-resolutionimmunocytochemistry of noncollagenous matrix proteinsin rat mandibles processed with microwave irradiation.

J Histochem Cytochem 2001; 49: 1099–1109.22. Nath SD, Linh NT, Sadiasa A, et al. Encapsulation of

simvastatin in PLGA microspheres loaded into hydrogelloaded BCP porous spongy scaffold as a controlled drug

delivery system for bone tissue regeneration. J BiomaterAppl 2014; 28: 1151–6323.

23. Imbronito AV, Scarano A, Orsini G, et al. Ultrastructure

of bone healing in defects grafted with a copolymer ofpolylactic/polyglycolic acids. J Biomed Mater Res A 2005;74: 215–221.

24. Berkland C, King M, Cox A, et al. Precise control ofPLG microsphere size provides enhanced control ofdrug release rate. J Control Release 2002; 82: 137–147.

25. Sinha VR, Bansal K, Kaushik R, et al. Poly-epsilon-caprolactone microspheres and nanospheres: an over-view. Int J Pharm 2004; 278: 1–23.

26. Ramstad VE, Franzen A, Heinegard D, et al.

Ultrastructural distribution of osteoadherin in rat boneshows a pattern similar to that of bone sialoprotein.Calcif Tissue Int 2003; 72: 57–64.

27. Jeon JH, Piepgrass WT, Lin YL, et al. Localized intermit-tent delivery of simvastatin hydroxyacid stimulates boneformation in rats. J Periodontol 2008; 79: 1457–1464.

28. Brown JL, Kumbar SG and Laurencin CT. Bone tissueengineering. In: Ratner BD, Schoen J, Hoffman AS andLemons JE (eds) Biomaterials Science: An Introduction toMaterials in Medicine, 3th ed., Chapter II.6.7, Oxford:

Elsevier, 2013, pp. 1194–1197.29. Adeeb N, Mortazavi MM, Tubbs RS, et al. The cranial

dura mater: a review of its history, embryology, and anat-

omy. Childs Nerv Syst 2012; 28: 827–837.30. Rey C, Combes C, Drouet C, et al. Bone mineral: update

on chemical composition and structure. Osteoporos Int

2009; 20: 1013–1021.

Ferreira et al. 11




31. Arana-Chavez VE, Soares AM and Katchburian E.Junctions between early developing osteoblasts of rat cal-varia as revealed by freeze-fracture and ultrathin section

electron microscopy. Arch Histol Cytol 1995; 58: 285–292.32. He F and Ye J. In vitro degradation, biocompatibility,

and in vivo osteogenesis of poly(lactic-co-glycolic acid)/calcium phosphate cement scaffold with unidirectional

lamellar pore structure. J Biomed Mater Res A 2012;100: 3239–50.

33. Brinkkoetter PT, Ising C and Benzing T. The role of thepodocyte in albumin filtration. Nat Rev Nephrol 2013; 9:328–336.

34. Song C, Guo Z, Ma Q, et al. Simvastatin induces osteo-blastic differentiation and inhibits adipocytic differenti-ation in mouse bone marrow stromal cells. BiochemBiophys Res Commun 2003; 308: 458–462.




evaluation of bone repair of critical size defects treated with simvastatin-loaded...

Documents