comparison of osteogenic potential of calvarial bone dust ... · arunesh gupta, md,* catherine...

Copyright @ 2009 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.

Comparison of Osteogenic Potential of CalvarialBone Dust, Bone Fragments, and Periosteum

Arunesh Gupta, MD,* Catherine Lobocki, MS,Þ Gopal Malhotra, MD,*and Ian T. Jackson, MD, FRCS, FACS*

Introduction: Bone dust is often used as a control when testing thepotential of a new reconstructive graft material. Under microscopicexamination, it would be expected to see the fully differentiated cel-lular components of bone, but instead only fusiform shapes charac-teristic of fibroblasts are mainly seen. This study aimed to comparethe osteogenic potential of cells obtained from calvarial bone dust,bone fragments, and periosteum using 3 assays: collagen, calcium,and alkaline phosphatase.Materials and Methods: Bone dust was harvested from the cal-varia of 5 euthanized rabbits by drilling burr holes. Small pieces ofintact, nondrilled bone, and periosteum were also obtained to serveas controls. The cells obtained from the bone dust, bone fragments,and periosteum were cultured for 5 weeks and then assayed forcollagen (type 1), calcium, and alkaline phosphatase.Results: Staining for calcium revealed that the greatest calciumdeposition was achieved with periosteum, followed by bone dust andthen bone fragments. Staining for alkaline phosphatase was similarfor bone dust and periosteum, followed by bone fragments. Collagenassay demonstrated the presence of collagen in similar concentra-tions in all 3 preparations.Conclusions: Bone dust has most of the necessary components forosteogenesis, including the presence of osteoprogenitor cells thathave the ability to lay down collagen type 1 and deposit calcium andcan differentiate to form bone. Further studies that can accuratelyquantify the percentage of surviving osteoblasts in various bonecomponents are needed.

Key Words: Bone dust, osteogenic potential

(J Craniofac Surg 2009;20: 1995Y1999)

An estimated 500,000 to 600,000 bone grafting procedures areperformed annually in the United States for replacing bone that

may have been lost because of atrophy, surgical intervention, orlong-term inflammation.1 Various filling materials are used for these

defects. These may be heterogenous or autogenous. Autogenous ma-terials are more biocompatible than heterogenous materials but havethe disadvantages of donor site morbidity and resorption. Bone dust,which is released on drilling of calvarial burr holes, can be collectedby a bone dust collector2,3 and used as an autograft for filling skulldefects.4Y8 In the various studies done on the properties of bone dust,there have been conflicting results. O’Broin et al9 used bone dustwith titanium mesh during cranioplasty. No new bone formation wasseen, and the bone dust was resorbed completely in their study.Contrary to their findings, Fukuta et al8 reported that bone dustproduced a great amount of bone. The questions on whether cellswithin the bone dust can survive the drilling procedure and whetherthese cells have any osteogenic potential remain to be determined. Inthis study, an attempt was made to identify the histochemicalproperties of bone dust to improve the understanding of itsphysiological behavior. A comparison of the osteogenic potentialof bone dust, bone fragments, and periosteum was made usingcollagen assay, alkaline phosphatase (ALP) assay (early markers ofosteoblastic phenotype), and the deposition of calcium phosphate(late markers of osteoblastic phenotype).

MATERIALS AND METHODSThis study was approved by the institutional research

committee of Providence Hospital. All funds for the study wereprovided by the Providence Hospital, Department of Patient CareResearch. All animals used in the study received humane care incompliance with the Guide for the Care and Use of LaboratoryAnimals published by the National Institutes of Health.

AnimalsFive New Zealand white rabbits (Covance Research Products,

Denver, PA) that were euthanized after an unrelated research projectwere used for this study. The calvaria of the rabbits had burr holesperformed with a neurosurgical burr under sterile conditions toobtain bone dust. At the same time, intact calvarial bone fragmentsand periosteum were harvested from the same rabbits as controls.The bone dust, bone fragments, and periosteum from each rabbitwere separately collected in a normal saline irrigation solution.

Culture MediaExplants and cells obtained were maintained in Dulbecco

modified Eagle medium (DMEM) with 10% fetal bovine serum (FBS;Invitrogen, Carlsbad, CA). Osteogenic media contained DMEM plus10% FBS, 10-mmol/L A-glycerophosphate, 0.1-Kmol/L dexameth-asone (Sigma, St Louis, MO), and 50-Kmol/L ascorbic acid (WakoChemicals, Richmond, VA).

Culture and Isolation of CellsBone Dust

The explants were placed in 75-cm2 flasks, and these wereincubated at 37-C with 5% carbon dioxideY95% air in low-glucoseDMEM supplemented with 10% FBS as previously described.

ORIGINAL ARTICLE

The Journal of Craniofacial Surgery & Volume 20, Number 6, November 2009 1995

From the *Institute for Craniofacial and Reconstructive Surgery and†Department of Patient Care Research, Providence Hospital, Southfield,Michigan.Received March 28, 2009.Accepted for publication June 21, 2009.Address correspondence and reprint requests to Arunesh Gupta, MD,

Institute for Craniofacial and Reconstructive Surgery, 16001 W NineMile RdY3rd floorYFisher Center, Southfield, MI 48075; E-mail:[email protected]

Copyright * 2009 by Mutaz B. Habal, MDISSN: 1049-2275DOI: 10.1097/SCS.0b013e3181bd3010


Bone FragmentsThe soft connective tissue and periosteum were removed, and

the bone was cut into 2 � 2-mm3 pieces. The preparations were cen-trifuged at 400g for 10 minutes and washed twice with phosphate-buffered saline (PBS). The bone piece explants were placed in 75-cm2

flasks and were incubated at 37-C with 5% carbon dioxideY95% airin low-glucose DMEM supplemented with 10% FBS.

PeriosteumThe periosteum was stripped off the calvaria, washed with

PBS, and cut into 2 � 2-mm3 pieces. The periosteum was culturedwith the cambium side downward the culture medium as describedfor bone fragments.

Once confluent, 50,000 cells per slide were plated onto four9-cm2 slide flasks for each of the 3 cell lines with osteogenic me-dium; these were cultured for 5 weeks. The cell lines derived fromthe 5 rabbits were maintained separately. At the end of 5 weeks, 20slides for each cell line (periosteum, bone dust, and bone fragments)

were harvested and assayed for calcium, ALP, and collagen. Theassays are described in the following paragraphs.10Y13

Alkaline PhosphataseAlkaline phosphatase expression was assessed histochemi-

cally (Sigma Diagnostics, Procedure no. 85; Sigma) on cells grownon tissue culture slides. The slides were fixed in citrate-bufferedacetone and then incubated in a solution of naphthol AS-MX phos-phate. At sites of ALP activity, the naphthol AS-MX was releasedand reacted with a diazonium salt (fast violet B) forming insolublered granules. The slides were then counterstained with Mayerhematoxylin.

Collagen Type 1Collagen type 1 was detected by immunohistochemistry

using a mouse monoclonal antibody (Clone Col-1; Sigma). Briefly,endogenous peroxidase activity was reduced by a 5-minute in-cubation in 3% hydrogen peroxide. Slides were then blocked for10 minutes in 1.5% goat serum in PBS and then incubated with theprimary antibody at a 1:500 dilution for 1 hour. The Vectastain EliteABC Kit (Vector, Burlingame, CA) was used for detection withdiaminobenzidine as the chromogen. Nuclei were visualized using ahematoxylin counterstain.

FIGURE 1. Von Kossa stain for calcium: bonefragments Y nodular calcification pattern.

FIGURE 2. Alkaline phosphatase stain: bone fragments.

FIGURE 3. Von Kossa stain for calcium: periosteum Y nodularcalcification pattern.

FIGURE 4. Alkaline phosphatase stain: periosteum.

Gupta et al The Journal of Craniofacial Surgery & Volume 20, Number 6, November 2009

1996 * 2009 Mutaz B. Habal, MD


CalciumCalcium accumulation was assessed in cells grown on tissue

culture slides using the von Kossa staining method. The cells werewashed 3 times with PBS (without calcium or magnesium) and thenfixed with 10% formaldehyde for 10 minutes. After an additionalwash step with dH2O, the cells were incubated in a 2% silver nitratesolution under ultraviolet light for 1 hour. The slides were thenwashed with dH2O and treated with 5% sodium thiosulfate for5 minutes. After washing the slides, the cells were counterstainedwith nuclear fast red. Calcium salts appeared as black granules inthis reaction.

Stain Intensity and Percentage Area-PositiveScoring of Slides

For each of the assays, the intensity of the staining was gradedfrom 1 to 4 as observed under the microscope. The percentage area,which was positive for the stain on the slides, was graded from 1 to 4as follows:

1 = 5% to 25% area-positive2 = 26% to 50% area-positive3 = 51% to 75% area-positive4 = greater than 75% area-positive

A combined score was generated by multiplying the intensityby the percentage-positive area, and a final score of 1 to 16 wasawarded to each slide.

PhotographyOnce the slides were stained, they were photographed using an

Olympus 5000 microscope-mounted camera (Olympus America Inc,Center Valley, PA) and assessed visually.

RESULTSFor ALP, collagen type 1, and calcium stains, the final scores

from the 20 slides each for bone dust, periosteum, and bone frag-ments were used to generate a mean score. We used analysis ofvariance to compare the mean score among the 3 cell lines for eachof the staining methods.

Bone FragmentsThe cultures reached confluence after 7 days, and the cells

showed polygonal morphology. After 14 days in the osteogenicmedium, multiple white nodules that increased in size over timewere seen. The final score for the bone fragments ranged from 2 to 6for both calcium and ALP activity (Figs. 1 and 2).

FIGURE 5. Von Kossa stain for calcium: bone dustYdispersedcalcification pattern.

FIGURE 6. Alkaline phosphatase stain: bone dust.

FIGURE 7. Collagen stain: periosteum.

FIGURE 8. Collagen stain: bone dust.

The Journal of Craniofacial Surgery & Volume 20, Number 6, November 2009 Osteogenic Potential of Cell Types

* 2009 Mutaz B. Habal, MD 1997


PeriosteumThe cells reached confluence in 5 days. Initially, the cells

had a fibroblast-like morphology, but after placing them in theosteogenic medium, they attained a polygonal shape. White noduleformation was seen after 21 days in the osteogenic medium butthen increased exponentially over time. On gross examination, thenumber as well as the size of the white nodules appeared to begreater than for the bone fragments. The periosteal cell lines con-sistently showed a final score ranging from 9 to 16 for calcium andALP stains (Figs. 3 and 4).

Bone DustThe cells reached confluence after 4 days and exhibited a

polygonal morphology. No nodule formation was seen even after5 weeks of culture in the osteogenic medium. Bone dustYderivedcell lines showed a final score ranging from 6 to 12 for both calciumdeposition and ALP stains (Figs. 5 and 6). Collagen scores rangedfrom 2 to 4 for all 3 cell lines (Figs. 7, 8, and 9).

DISCUSSIONBone regeneration is a multistep process that involves mitotic

expansion of progenitor cells at the sites of bone formation and theirdifferentiation into functional osteoblasts. In our study, we tried todetermine whether bone dust could be used as a bone graft material,which helped in this multistep process of bone regeneration. We alsocompared the osteogenic potential of bone dust with periosteum and

bone fragments using various markers for the osteoblastic pheno-type. Any bone graft material must have the properties necessary tocreate an optimal microenvironment, which attracts potential bone-forming cells. Our study showed that when bone dust was placedin an adequate milieu of cytokines, nutrients, and hormones, theosteoprogenitor cells are activated and stain positive for both earlyand late markers of osteogenic potential.

Markers correlated with the osteoblastic phenotype includehigh ALP levels, expression of collagen type 1, and the presence ofcalcium.14 As parameters for the osteoinductive potential of bioma-terials, cells were evaluated on their expression levels of ALP, colla-gen synthesis (early markers), and calcium deposition (late markers).

In our study, all cell lines showed deposition of calcium,signifying mineralization, but the pattern of calcium deposition wasdifferent in bone dust cell lines when compared with the bonefragments and periosteal cell lines. Bone dust cell lines showed adispersed calcification pattern, whereas bone fragments and peri-osteal cell lines showed foci of well-defined nodules. This differencein mineralization can be explained by the following. Cells derivedfrom periosteum and calvarial bone fragments are a heterogeneouspopulation of cells, possibly having a small amount of bone-formingcells (osteoprogenitors and colony-forming unitVosteoblasts). Invitro culture of these cells derived from periosteum and calvarialbone fragments results in the formation of well-distinct mineralizednodules representative of the clonogenic expansion of the colony-forming unitVosteoblasts (from calvarial bone fragments) or theosteoprogenitors (from the periosteum). Bone dustYderived cells arehomogeneous cell populations hence have most of the components

FIGURE 9. Collagen stain: bone fragments.

FIGURE 10. Mean final score and SE for calcium (stainintensity and percentage area-positive) in periosteum, bonedust, and bone fragments (n = 20).

FIGURE 11. Mean final score and SE for ALP (stain intensityand percentage area-positive) in periosteum, bone dust, andbone fragments (n = 20).

FIGURE 12. Mean final score and SE for collagen (stainintensity and percentage area-positive) in periosteum, bonedust, and bone fragments (n = 20).

Gupta et al The Journal of Craniofacial Surgery & Volume 20, Number 6, November 2009

1998 * 2009 Mutaz B. Habal, MD


for forming and secreting minerals leading to the dispersed patternof calcification.

For the calcium assay, the mean scores were significantly dif-ferent among bone dust, bone fragments, and periosteum (P e 0.001).The mean for periosteum (13.2) was higher than the mean for bonedust (10) or bone fragments (3.9), implying that the periosteal celllines were most active in laying down calcium (Fig. 10).

Bone-specific ALP is considered to be the single most accu-rate marker of bone formation.15 Although efforts have been madeto develop analytical techniques that can differentiate between thedifferent isoenzymes of ALP, most techniques are restricted to bonemetabolism of human origin. Consequently, tissue-nonspecific ALPis measured in osteoblast-like cell cultures in vitro. Thus, the pres-ence of ALP in bone dustYderived cell lines shows osteoblasticactivity. For the ALP assay, there was no statistical difference inthe mean score between the periosteum (10.5) and bone dust (10.2;P = 0.813). However, there was a statistical difference observedbetween the mean scores of bone dust and periosteum in comparisonto bone fragments (4.0; P e 0.001; Fig. 11).

Positive collagen staining confirms the matrix-laying capacityof these cell lines. For collagen stains, there was no significantdifference between the mean scores of all 3 cell lines (P = 0.650;Fig. 12). Stringa16 established a striking correlation between the rateof vascular penetration into the bone implant and its ability to sur-vive. He also showed that blood vessels penetrate cancellous bonemuch more readily than cortical bone because the cortex presents aphysical impediment to invading blood vessels. Bone dust is com-pletely permeable to new blood vessels coming from the surround-ing tissue, thus having a better neovascularization potential. Withthe full complement of cellular activity, bone dust is osteogenic,osteoinductive, osteosynthetic, and is easily incorporated into thedefect in which it has been placed.13,17 In our study, the markersused were an indirect measure of osteoblast activity.

Further studies that can quantify the exact percentage ofsurviving osteoblasts by radiolabeled DNA marking techniques willhelp in validating our results.

In our study, the period chosen for studying the osteogenicproperties of the osteoblasts was 5 weeks. Various studies10,11,14

have chosen time points varying from 2 to 6 weeks for determiningthe osteogenic potential. These varying time points were decided onby investigators depending on the aspect of osteoblasts beingstudied. As we were investigating the overall response of theosteoblasts from all sources, the 5-week end point seemed to be areasonable conclusion.

Bone dust does not offer structural support but is well suitedfor filling bone defects and cavities. As such, bone dust can be usedas an effective bone graft material on its own if the defect is small orin combination with other materials, including bone graft. This canminimize the amount of graft needed to fill a defect, hence, limitingdonor site size and morbidity.7,18Y20

In our study, maximum mineralization and ALP formationwas seen with the periosteum. This observation leads us to concludethat if bone defects are filled with bone dust and the periosteum isused for cover, then the healing process will be expedited.

CONCLUSIONSThis study shows that bone dust has most of the necessary

components for osteogenesis. It also establishes that bone dust has

mineral deposition, ALP production, and collagen laying down prop-erties. Owing to its porous structure, bone dust is easily neovascu-larized, and because it is autogenous, there is no immunologicreaction, thus getting easily incorporated in the defect. Although theperiosteum was most active in laying down calcium, bone dust,owing to its easy availability in bulk and paste-like consistency, isbest suited for filling and molding into bone defects.

REFERENCES1. Bucholz R. Nonallograft osteoconductive bone graft substitutes.

Clin Orthop Relat Res 2002;395:44Y522. Jackson IT, Sullivan C, Shiele UU. A simple bone dust collector.

Plast Reconstr Surg 1988;81:967Y9693. Munschler GF, Lane JM. Orthopedic surgery. In: Habal MB, Reddi AH,

eds. Bone Grafts and Substitutes. Philadelphia, PA: WB Saunders,1992:375Y407

4. Jackson IT, Pellet C, Smith JM. The skull as a bone graft donor site.Ann Plast Surg 1983;11:527Y532

5. Shehadi SI. Skull reconstruction with bone dust. Br J Plast Surg1970;23:227

6. Hunter PD, Pelofsky S. Classification of autogenous skull grafts incranial reconstruction. J Craniomaxillofac Trauma 1995;1:8Y15

7. Matsumoto K, Kohmura E, Kato A, et al. Restoration of small bonedefects at craniotomy using autologous bone dust and fibrin glue.Surg Neurol 1998;50:344Y346

8. Fukuta K, Har-Shai Y, Collares MV, et al. Comparison of inorganicbovine bone mineral particles with porous hydroxyapatite granules andcranial bone dust in the reconstruction of full thickness skull defects.J Craniofac Surg 1992;3:25Y29

9. O’Broin ES, Morrin M, Breathnach E, et al. Titanium mesh and bonedust calvarial patch during cranioplasty. Cleft Palate Craniofacial J1997;34:354Y356

10. Declercq H, Van den Vreken N, De Maeyer E, et al. Isolation,proliferation and differentiation of osteoblastic cells to study cell/biomaterial interactions: comparison of different isolation techniquesand source. Biomaterials 2004;25:757Y768

11. Aronow MA, Gerstenfeld LC, Owen TA, et al. Factors that promoteprogressive development of the osteoblast phenotype in culturedfetal rat calvaria cells. J Cell Physiol 1990;143:213Y221

12. de Pollak C, Arnaud E, Renier D, et al. Age-related changes in boneformation, osteoblastic cell proliferation and differentiation duringpostnatal osteogenesis in human calvaria. J Cell Biochem1997;64:128Y139

13. Ho ML, Chang JK, Chuang LY, et al. Characteristics of primaryosteoblast culture derived from rat fetal calvaria. Kaohsiung JMed Sci 1999;15:248Y255

14. Declercq HA, Verbeeck RM, De Ridder LI, et al. Calcification as anindicator of osteoinductive capacity of biomaterials in osteoblasticcell cultures. Biomaterials 2005;26:4964Y4974

15. Allen MJ. Biochemical markers of bone metabolism in animals:uses and limitations. Vet Clin Pathol 2003;32:101Y113

16. Stringa G. Studies on the vascularization of bone grafts. J BoneJoint Surg 1957;39:395

17. Gotoh Y, Fujisawa K, Satomura K, et al. Osteogenesis by humanosteoblastic cells in diffusion chamber in vivo. Calcif Tissue Int1995;56:246Y251

18. Sheikh BY. Simple and safe method of cranial reconstruction afterposterior fossa craniectomy. Surg Neurol 2006;65:63Y66

19. Jackson IT. Analysis and treatment of hemifacial microsomia.Eur J Plast Surg 2004;27:159Y170

20. Jackson IT. Cranial bone grafting for orbital reconstruction: is it stillbest. J Craniofac Surg 2005;16:181Y185

The Journal of Craniofacial Surgery & Volume 20, Number 6, November 2009 Osteogenic Potential of Cell Types

* 2009 Mutaz B. Habal, MD 1999

comparison of osteogenic potential of calvarial bone dust ... · arunesh gupta, md,* catherine...

Documents