Periodontal regenerationfollowing implantationof cementumandperiodontalligament-derived cells

J. NuÇez1,2, S. Sanz-Blasco1,F. Vignoletti2, F. MuÇoz3,H. Arzate4, C. Villalobos1,L. NuÇez1, R. G. Caffesse2, M. Sanz21Institute of Molecular Biology and Genetics(IBGM), University of Valladolid and SpanishResearch Council (CSIC), Valladolid, Spain,2Faculty of Odontology, Complutense Universityof Madrid, Master in Periodontology, Madrid,Spain, 3Faculty of Veterinary, Universityof Santiago de Compostela, Lugo, Spain and4Biology and Molecular Laboratory, Facultyof Odontology, National Autonomous Universityof Mexico, Mexico

The periodontium is a complex organcomprising four mesenchymal tissuecomponents (cementum, bone, gingivaand periodontal ligament) that act as afunctional unit, providing the toothwith an attachment apparatus capableof withstanding masticatory forces (1).Amongst these components, the peri-

odontal ligament (PDL) has theimportant role of supporting the toothand maintaining the homeostasis of theperiodontium. This tissue is highlyvascularized and contains many celltypes, including fibroblasts and endo-thelial, epithelial, nerve and mesen-chymal progenitor cells. These latter

cells are mature multipotent mesen-chymal stem cells (MSC-PDL) (2) thatare able to di!erentiate into multiplecell lineages, as well as to self-renewand maintain their multipotent capac-ity (3–5). MSC-PDL have demon-strated the potential to di!erentiateinto osteogenic, adipogenic progenitor

Nunez J, Sanz-Blasco S, Vignoletti F, Munoz F, Arzate H, Villalobos C, Nunez L,

Ca!esse RG, Sanz M. Periodontal regeneration following implantation of cementum

and periodontal ligament-derived cells. J Periodont Res 2012; 47: 33–44. ! 2011

John Wiley & Sons A/S

Background and Objective: The periodontal regeneration of bone defects is often

unsatisfactory and could be largely improved by cell therapy. Therefore, the pur-

pose of this study was to evaluate the regenerative potential of implanting canine

cementum-derived cells (CDCs) and canine periodontal ligament-derived cells

(PDLDCs) in experimentally created periodontal intrabony defects in beagle dogs.

Material and Methods: Cells were obtained from premolars extracted from four

beagle dogs. Three-wall intrabony periodontal defects, 3 mm wide and 4 mm deep,

were surgically created in their second and fourth premolars and plaque was

allowed to accumulate. Once the defects were surgically debrided, periodontal

regeneration was attempted by random implantation of collagen sponges

embedded with 750,000 CDCs, 750,000 PDLDCs or culture medium. After 3 mo

of healing, specimens were obtained and periodontal regenerative outcomes were

assessed histologically and histometrically.

Results: The histological analysis showed that a minimal amount of new

cementum was formed in the control group (1.56 ± 0.39 mm), whereas in both

test groups, significantly higher amounts of new cementum were formed

(3.98 ± 0.59 mm in the CDC group and 4.07 ± 0.97 mm in the PDLDC group).

The test groups also demonstrated a larger dimension of new connective tissue,

resulting in a significantly more coronal level of histological attachment.

Conclusion: This proof-of-principle study suggests that cellular therapy, in com-

bination with a collagen sponge, promoted periodontal regeneration in experi-

mental intrabony periodontal defects.

Javier NuÇez, DDS, MS, Institute of MolecularBiology and Genetics (IBGM), Sanz y Fores 3,47003 Valladolid, SpainTel: +34 983 184821Fax: +34 983 184800e-mail: javilds@hotmail.com

Key words: cementoblasts; dental cementum;intrabony periodontal defects; mesenchymalstem cells; periodontal regeneration

Accepted for publication June 27, 2011

J Periodont Res 2012; 47: 33–44All rights reserved

! 2011 John Wiley & Sons A/S

JOURNAL OF PERIODONTAL RESEARCH

doi:10.1111/j.1600-0765.2011.01402.x

cells (6) and precursors of the neuralcrest-like cells (7). It is not easy toidentify MSC-PDL because there is nosingle antigenic marker specific forthese cells (5); therefore, because MSC-PDL have many cell-surface moleculesin common with mature hematopoieticcells, a combination of markers areused for their identification, such asCD13, CD29, CD44, CD59, CD90,CD10 (8) and STRO-1 (9). Accordingto the International Society for Cellu-lar Therapy (ISCT), three minimalcriteria must be fulfilled for the in vitroidentification of MSCs: adherence toplastic, a specific surface-antigenexpression pattern (CD73+, CD90+,CD44+, CD105+, CD34), CD45),CD11b), CD14), CD19), CD79a) andHLA-DR)) and di!erentiation poten-tial (osteogenic, chondrogenic andadipogenic lineages) (10). Canine MSCsderived from PDL have only demon-strated expression of CD146 andSTRO-1 (11). MSCs have been isolatednot only from the PDL, but also fromthe dental pulp (12,13), in exfoliateddeciduous teeth (14) and in the apicaldental papilla (15).

MSCs have seldom been used ther-apeutically and most cell therapyattempts in oral and maxillofacial sur-gery applications have employed bonemarrow stromal mesenchymal stemcells (BMSMCs) because these cellsfavour osteogenic di!erentiation vs.other cell lineages (16). In caninemodels, BMSMCs vehicled in ceramichydroxyapatite/tricalcium phosphatesca!olds (17,18) promoted boneregeneration in alveolar defects andenhanced bone–implant contact whencompared with autogenous bonegrafts plus platelet-rich plasma (19).In bone-regenerative applications theBMSMCs have been tested in combi-nation with various sca!olds in sinuslift procedures (20) or for the treatmentof cleft palate defects (21,22). Thesestudies suggest that cell therapy maypromote the bone-regenerative capac-ity of di!erent bone grafts and hencereduce the donor site morbidity asso-ciated with autogenous bone grafting.

In periodontal regeneration the celltherapy approach has been rarelyexploited. The periodontium destroyedduring chronic periodontitis has lim-

ited capacity for self-regeneration. Infact, once the infection has beenremoved from the a!ected roots, thecells repopulating the wound willdetermine the nature of the tissueinterface, mainly through the estab-lishment of a long-junctional epithe-lium. This reparative process may beprevented when the epithelial and gin-gival connective tissue cells are excludedfrom the wound-healing process andcells from the PDL are allowed tocolonize the wound, thus achievingperiodontal regeneration (23). Thisperiodontal-regenerative potential hasbeen attributed to undi!erentiatedMSCs from the PDL that, togetherwith appropriate growth and di!eren-tiation factors, may promote the for-mation of new cementum, connectivetissue attachment and bone (1,24).

Even though the presence of MSCsin the PDL and the pulp is well estab-lished (9), it is not well known howthese cells behave in vivo or the factorsthat influence their di!erentiation intothe di!erent cell lineages. A fewinvestigations have attempted to iso-late and culture these cells, as well as tostudy their potential to form peri-odontal tissue (25–28). Preisig & Sch-roeder (29) cultured PDL cells in rootslices and developed an in vitro PDL-like tissue. In another experiment,demineralized roots with attachedPDL cells were re-implanted in vivo inmonkeys. Healing resulted in the for-mation of new collagen fibers perpen-dicularly oriented between the boneand the root. However, ankylosis androot resorption were also observed(30). These first attempts at regenera-tion demonstrated the need to isolateprogenitor cells and to use suitablesca!olds to deliver the cells and growthfactors to guide cell di!erentiationappropriately. Seo et al. (2) isolatedhuman MSCs from the PDL and usedhydroxylapatite/tricalcium phosphatesca!olds to implant these cells in arti-ficially created periodontal defects inthe mandibular molar of six immune-deficient rats. Eight weeks later theauthors found a thin layer of cemen-tum-like tissue with attached con-densed collagen fibers, resemblingSharpey!s fibers. Similarly, Zhao et al.(31) also implanted MSCs within a

polymer sponge in a fenestration-de-fect model in rats and reported di!er-entiation of mice cementoblasts withthe ability to form mineralized tissueafter 6 wk.

Our research group has previouslyreported (32,33) a method for the iso-lation of MSCs derived from humanPDL and their di!erentiation intocementoblasts and fibroblasts by immu-nocytochemistry and RT-PCR. Thesehuman cementum-derived cells (humanCDCs) were positive for amelobla-stin, amelogenin, cementum protein 1(CEMP-1), cementum attachmentprotein (CAP) and osteocalcin. Oncethe methodology for the isolation andcharacterization of CDCs was estab-lished, we aimed to evaluate theircapacity to promote periodontal regen-eration in a preclinical experimentalmodel. Accordingly, the goal of thepresent investigation was to evaluatethe regenerative potential of implantedcanine CDCs and canine periodontalligament-derived cells (PDLDCs) with-in a collagen sponge (CS) used assca!old in experimentally created, crit-ical intrabony periodontal defects.

Material and methods

Cell isolation and culture

This experimental animal study wasapproved by the Ethical ResearchCommittee of the "Gomez Ulla! Cen-tral Military Hospital (Madrid, Spain).Four male, 1-year-old Beagle dogsweighing about 10 kg were selectedand subjected to the prescribed quar-antine period. Although the animalspresented healthy periodontal tissues,full-mouth prophylaxis using ultra-sonic scalers and manual curettes wascarried out prior to the experimentalphase. Canine CDCs and caninePDLDCs were isolated and character-ized using the method previouslyreported by our research group (32). Inbrief, first and third mandibular pre-molars were carefully extracted undergeneral anaesthesia (inhalatory anaes-thesia – isofluorane – induced withintravenous propofol) and immediatelyplaced in serum-free, alpha-modifiedEagle!s Medium (a-MEM/F12) nutrientmixture supplemented with 100 U/mL

34 Nunez et al.

of penicillin and 100 lg/mL of strep-tomycin. PDL was scraped from theroots using a 7/8 Gracey curette(Hu-Friedy, Chicago, IL, USA) andthe tissue obtained was digested byincubation, at 37"C, in medium con-taining 3 mg/mL of collagenase type Iand 4 mg/mL of dispase. After40 min, digested tissue was centrifugedfor 7 min at 200 g and washed fourtimes with fresh medium. Cells werethen passed through a 70-lm straineronto Petri dishes containing a-MEMsupplemented with 15% fetal bovineserum, 100 lM ascorbic acid 2-phos-phate, 2 mM glutamine, 100 U/mL ofpenicillin and 100 lg/mL of strepto-mycin, and the plates were placedinside an air incubator at 37"C with anatmosphere of 5% CO2/95%. Underthese conditions, a cluster of mesen-chymal cells grew after 4 d in vitro,forming the so-called colony-formingunit. Every 2 d or so, when the cul-tures became semiconfluent, cells weredetached and placed in a mediumcontaining trypsin-EDTA. The num-ber of PDLDCs doubled in a 2- to 3-dperiod and these cells were used forexperiments within the fourth and fifthpassages. The remaining root was leftin the same medium, containing 3 mg/mL of collagenase type I and 4 mg/mLof dispase, for 40 min at 37"C. Theroot was then washed four times withfresh medium and thin root layerswere scraped o! in an apical–coronaldirection using a 15c blade. Theselayers were washed and cut intosmaller pieces using microsurgeryscissors and then digested again byincubation, for 40 min at 37"C, inmedium containing 3 mg/mL of col-lagenase type I and 4 mg/mL of dis-pase. Small tissue pieces were thenplated in a Petri dish containinga-MEM supplemented with 15% fetalbovine serum, 100 lM ascorbic acid2-phosphate, 2 mM glutamine, 100 U/mL of penicillin and 100 lg/mL ofstreptomycin, and placed in a 5%CO2/95% air incubator at 37"C. After10 d of culture, a pool of canine CDCsgrew, forming discrete cell colonies.Every 4 d or so, when the culturesbecame semiconfluent, the cells weredetached using medium containingtrypsin-EDTA. The number of CDCs

doubled in a 3- to 4-d period, andthese cells were used for experimentswithin the fourth and fifth passages.

Immunocytochemistry

Canine CDCs and PDLDCs (at pas-sages four to five) were fixed with 4%formaldehyde and analysed usingimmunocytochemistry. In brief, CDCswere incubated with primary antibodiesagainst osteocalcin (1 : 100 dilution;Alexis Biochemicals, Lausanne, Swit-zerland), CEMP-1 (1 : 100 dilution)and STRO-1 (1 : 100 dilution; SantaCruz Biotechnology, Heidelberg, Ger-many) in 10% goat serum for 90 min.PDLDCs were incubated with anti-bodies against osteocalcin (1 : 100 di-lution), STRO-1 (1 : 100 dilution),CEMP-1 (1 : 100 dilution) and CD44(1 : 100 dilution; AbCAM, Cambridge,UK) under the same experimental con-ditions. The samples were then incu-bated, for 60 min, with fluoresceinisothiocyanate (FITC)-labelled secon-dary antibodies (Invitrogen, Barcelona,Spain). Stained cells were identified byfluorescence microscopy.

Flow cytometry analysis

PDLDCs (at passages four to five) werecultured in a-MEM supplemented with15% fetal bovine serum, 2 mM gluta-mine, 100 U/mL of penicillin and100 lg/mL of streptomycin. Then, thecells were incubated with mouse anti-human serum to CD73, CD90, CD44,CD105, CD34, CD45, CD11b, CD80,CD19 and HLA-DR. Then, the cellswere incubated with a 1 : 100 dilutionof FITC-conjugated goat anti-mouseIgG and analyzed using an EPICS XLflow cytometer and the Expo 32 ACSsoftware# (Coulter Electronics, Brook-vale, Australia).

Mineralization and Von Kossastaining

Canine CDCs and PDLDCs (at pas-sages four to five) were plated at adensity of 104 cells/well in sterile six-well plates and cultured in a mineral-izing medium [Dulbecco!s modifiedEagle!s minimal essential medium(DMEM)/F12] containing 15% fetal

bovine serum and 1% penicillin/strep-tomycin (diluted 100-fold from a stocksolution containing 5000 U/mL ofpenicillin, 5000 U/mL of streptomycin,50 mg/mL of ascorbic acid, 10 mM

sodium b-glycerophosphate and 5 lMdexamethasone). The medium waschanged every 2 d and mineralizationwas tested after 3 wk using Von Kossastaining and bright-field microscopy.For Von Kossa staining, the cells werewashed with phosphate-bu!ered salineand fixed with 4% paraformaldehydefor 10 min. After washing with distilledwater, the cells were incubated in 5%silver nitrate at 4"C for 40 min. Sub-sequently, the cells were washed withwater and 5% sodium thiosulfatesolution was added at room tempera-ture for 5 min. After removal of thesodium thiosulfate solution and wash-ing, the cells were subjected to lightmicroscopy for analysis.

Experimental study

The same animals that provided theextracted teeth for the isolation ofCDCs and PDLDCs were used in theexperimental study. Intrabony peri-odontal defects were created undergeneral anaesthesia [inhalatory anaes-thesia with isofluorane and then in-duced with intravenous (i.v.) propofol].Buccal and lingual mucoperiostealflaps were raised in contralateral jawquadrants through intrasulcular inci-sions from the mesial aspect of thesecond premolar to the first molar.Using a hand-piece and fissureburs under copious irrigation, three-wall intrabony periodontal defects,measuring 3–4 mm in buccolingual,mesiodistal and apicocoronal direc-tions, were surgically created at themesial and distal aspects of secondpremolars and at the mesial aspect ofthe fourth premolars, thus generatingsix defects in each animal. Orthodonticwire ligatures were then placed andsecured around the a!ected teeth andinserted into the defects before theflaps were repositioned and suturedwith resorbable interrupted sutures(Ethicon, Summerville, NJ, USA). Thedogs were then fed with a soft dietto allow plaque accumulation andthus mimic chronic periodontitis. After

Regeneration by cellular implantation 35

1 mo of plaque accumulation, the dogswere sedated with 80 lg/kg of mede-tomidine, 20 lg/kg of butorfanol and100 lg/kg of atropine sulfate and theligatures were removed together withthe accumulated plaque and calculususing ultrasonic scalers. Seven dayslater (5 wk after the creation of thedefects), regenerative surgery was car-ried out using the same anestheticprotocol described earlier. After raisingbuccal and lingual mucoperiostealflaps, the intrabony periodontal defectswere carefully debrided and the rootswere planed with curettes. The follow-ing intrabony measurements wererecorded using a North Carolina peri-odontal probe (Hu-Friedy, Chicago,IL, USA): the distance between thealveolar bone crest and the cemento–enamel junction (ABC–CEJ), the dis-tance from the ABC to the base of thedefect (ABC–BD) and the bucco–lingual dimension of the defect. Allmeasurements were rounded down tothe nearest millimeter. With the use ofa round bur, a reference notch wasmade at the base of the defect. Thethree modalities of treatment werealways applied in the same sequence inorder to avoid repeating the treatmentin the defects. The randomization oftreatments was performed by the toss-ing of a coin, selecting which treatmentto start at the mesial intrabony peri-odontal defect of the second premolar(Table 1). The regenerative treatmentstested consisted of the implantation ofa CS (Xemax Surgical Products Inc.,Napa, CA, USA) containing (a)> 750,000 CDCs (test), (b) > 750,000PDLDCs (test) or (c) 50 lL of culturemedium (control) into the periodontaldefect. The flaps were then reposi-

tioned and sutured with resorbableinterrupted sutures (Ethicon). To en-sure the presence of viable cells in thesca!old (CS) we previously assessedthe maintenance of living cells byincubating canine CDCs for 24 h at37"C in a 5% CO2/95% air atmosphereand then after 5 h in a refrigerator(4"C) to mimic the conditions to whichcells are exposed during transit fromthe laboratory to the operating room.Cell viability was assessed by fluores-cence microscopy after staining thesca!old with fluorescein diacetate(50 lg/mL, 3 min) and propidiumiodide (20 lg/mL, 30 s) (Fig. 1).

After the experimental surgery, allanimals received antibiotics (amoxicil-lin, 15 mg/kg intramuscularly, every48 h for 7 d) and anti-inflammatorymedication (meloxicam, 0.2 mg/kg sub-cutaneously, every 24 h for 3 d) andwere fed with a soft diet for 2 wk toreduce potential mechanical trauma tothe surgery area. A chlorhexidine glu-conate (0.12%) spray was topicallyapplied every third day for a 3-moperiod.

Histological study

Three months after surgery, the dogswere killed by an overdose of sodiumpentobarbital and the tissues were fixedby vascular perfusion through the car-otid arteries with Karnovsky fixativesolution (34). The lower jaw was thenremoved and immersed in Karnovskysolution for 1 wk. The experimentalteeth were hemisected and tissue blockscontaining the treated root and its sur-rounding soft and hard tissues weredissected and processed for groundsectioning according to the methoddescribed by Donath & Breuner (35).The tissue blocks were sectioned at themost central aspect of the defect using adental radiograph to ensure properorientation. One mesiodistal sectionwas prepared from each tissue blockand reduced to a thickness of approxi-mately 20 lm using a microtome(Exakt#; Apparatebau, Norderstedt,Germany). The sections were stainedusing the Levai–Laczko technique (36).All specimens were analyzed histologi-cally and histometrically under a lightmicroscope (Eclipse E800; Nikon Inc.,

Tokyo, Japan) equipped with acomputerized image-analysis system(NIS Elements BR, Nikon DS-Ri1;Amstelveen, the Netherlands). A well-trained examiner (J.N.), blinded to thegroups, carried out all histologicalevaluation and histometric analysis. Toensure reproducibility and calibrationbefore the histometric analysis, a subsetof 10 samples was measured in dupli-cate after 1 wk. This analysis resultedin 100% agreement between the dupli-cate measurements, within a range of±0.3 mm. The following measure-ments were recorded in each section: thedistance from the CEJ to the notch(CEJ-Notch); the distance between thegingival margin (GM) and the apicalend of the junctional epithelium (JE)(GM-JE); the distance from the CEJ tothe GM (CEJ-GM); the distancebetween the CEJ and the apical end ofthe JE (CEJ-JE); the distance betweenthe apical extent of the JE and the mostcoronal cementum (JE-CC); the dis-tance between the most coronal extentof newly formed cementum to the notch(CC-Notch); and the distance betweenthe most coronal osseous crest and thenotch (OC-Notch).

Statistical analysis

Means and standard deviations werecalculated for each clinical and histo-logical parameter in both experimentaland control groups, with the beagledog serving as the unit of analysis(n = 4). When only two means werecompared, the Student!s t-test wasused. For more than two groups, sta-tistical significance of the data wasassessed using analysis of variance andcompared using Bonferroni!s multiple-comparison test. Di!erences wereconsidered significant at p < 0.05.

Results

Culture and characterization ofcanine CDCs and PDLDCs

Isolation and culture of experimentalcells obtained from the canine root-rendered cells with the distinct abilityto form cell colonies and demonstrat-ing the characteristic CDC andPDLDC phenotypes, was carried out.

Table 1. Distribution of the treatmentsused to treat intrabony periodontal defects

Area Control Test 1 Test 2

Mesial PII 4 2 2Distal PII 2 3 3Mesial PIV 2 3 3

The values given represent the number oftreatments.Control, culture medium; Test 1, caninecementum-derived cells (CDCs); Test 2,canine periodontal ligament-derived cells(PDLDCs).

36 Nunez et al.

PDLDCs started to spread out fromthe explants after 4 d in vitro (Fig. 2A),whereas CDCs did the same after 10 din vitro (Fig. 2B). After 60 d in vitro,both cell types formed colonies withclear deposition of mineralized tissue

(Fig. 2C) and amorphous brown/blackprecipitates after Von Kossa staining(Fig. 2D). Immunocytochemical anal-ysis was performed on fourth-passage(60 d in vitro) CDCs using anti(a)-CEMP-1, a-STRO-1 and a-osteocalcin

and on fourth-passage PDLDCs usinga-STRO-1, a-CD44, a-CEMP-1 anda-osteocalcin. Figure 3A shows positiveimmunofluorescence staining in CDCsfor a-CEMP-1 (78% of the CDCpopulation) and a-osteocalcin (66%),

Fig. 1. Cell viability in the carrier [collagen sponge (CS)]. The CS was impregnated with canine cementum-derived cells (CDCs) in 50 lL of

culture medium and incubated for 24 h at 37"C in a 5% CO2/95% air incubator and for 5 h in a refrigerator (4"C). The left image shows

transmitted light of the cells in the CS. The center image shows dead cells stained with propidium iodide. The right image shows live cells

stained with fluorescein diacetate. Bar = 5 lm, the magnification was the same in all panels. The results shown are representative of three

experiments.

A

B

C

D

Fig. 2. Canine cementum-derived cells (CDCs) and canine periodontal ligament-derived cells (PDLDCs) form cell colonies and mineral.

Teeth from young beagle dogs were used as a source of PDL. (A) Periodontal ligament was scraped from the tooth and digested using

collagenase and dispase. After 4 d in vitro, the cells started to form colonies, termed fibroblast colony-forming units (CFU-F). At 60 d in vitro,

the culture had completed four passages. (B) The external layers from root surfaces were removed and digested using collagenase and dispase.

Cells from the cultured tissue explants started to spread out after 10 d in vitro and to form colonies and a mineralized tissue. A representative

example at 60 d in vitro is shown. CDCs and PDLDCs cultured for 3 wk with dexamethasone, ascorbic acid and b-glycerophosphate were

able to form mineralized tissue, as shown by light microscopy (C) and Von Kossa staining (D). Bar = 10 lm in all panels. Data are

representative of three independent experiments.

Regeneration by cellular implantation 37

but not for a-STRO-1. Figure 3Bshows positive immunofluorescencestaining in PDLDCs for a-CD44 (6%)and a-STRO-1 (12%) but not fora-CEMP-1 and a-osteocalcin. Weattempted to further characterize thesecells by flow cytometry. We found thatcanine PDLDCs expressed CD105, aspecific surface antigen of MSCs(Fig. 3). The canine-derived cells were

negative for a number of characteristicmarkers of MSCs when antibodiesagainst human antigens were used(data not shown), implying either thatthe cells do not express these markersor, more likely, that the antibodies forhuman markers do not identify theantigens in canine cells. Unfortunately,canine-specific antibodies are notavailable at present.

Regenerative experimental study byapplication of a CS imbibed in CDCs,PDLDCs or culture medium

A critical step in experimental celltherapy is to test the e"ciency of thecell carrier. We chose to test a CS as acell carrier. Therefore, we tested theviability of CDCs or PDLDCs (atpassages four to five) imbibed andcultured within the CS. The CS wasable to maintain cell viability for 24 hof incubation and for 5 h in a coldenvironment (Fig. 1), two conditionsrequired to fill the CS with cells and totransport them from the cell culturefacility to the surgery room. CS-con-taining cells were applied to thedefects, as described in the Materialand methods, and the e!ects on dif-ferent parameters were tested usinghistometry. Table 2 depicts the intra-surgical measurements recorded at thesites where the CS containing CDCs,PDLDCs or culture medium wasapplied. Postoperative healing occurreduneventfully in all animals and at 4 wkthe periodontal tissues were healthywithout any sign of infection. Speci-mens for histological analysis wereobtained for all defects in the fourexperimental animals. Their gingivaltissues had similar histological charac-teristics in the three study groups,depicting a parakeratinized epitheliumcovering the gingival margin with anunderlying connective tissue rich incollagen fibers and fibroblasts, exceptfor a band of inflammatory infiltrateconfined to the sulcular and junc-tional epithelium (Fig. 4). The perio-dontal attachment apparatus, however,showed distinct histological character-istics when the groups were compared.The healing of the defects in both testgroups demonstrated histologicalcharacteristics of periodontal regener-ation, including formation of new cel-lular cementum and a significantlylarger connective tissue attachmentrelative to the control group (Fig. 5).Conversely, in the control group,healing mostly occurred by repair, withlimited formation of new cellularcementum coronal to the referencenotch. No signs of root resorption orankylosis were observed in eithergroup. In the PDL between the newly

Fig. 3. Characterization of cell markers in canine cementum-derived cells (CDCs) and canine

periodontal ligament-derived cells (PDLDCs). (A) CDCs were subjected to immunocyto-

chemistry, using, as probes, anti-cementum protein 1 (a-CEMP-1), a-STRO-1 and a-osteo-calcin, and positive cells were revealed with fluorescein-conjugated second antibody. Left,

immunofluorescence images. Right, bright-field images. (B) PDLDCs were subjected to

immunocytochemistry, using, as probes, a-CD44, a-STRO-1, a-osteocalcin and a-CEMP-1.

Positive cells were revealed with fluorescein-labelled a-CD44 and a-STRO-1 as secondary

antibodies. Left, immunofluorescence images. Right, bright-field images. The bar represents

10 lm, the left panels have the same magnification as the corresponding right panels. (C)

Flow cytometry analysis of expression of CD105 in PDLDCs. Cells were treated with

antibodies against human CD105 and positive cells were tested by flux cytometry.

38 Nunez et al.

formed bone and cementum, rich cap-illary vessels were observed (Fig. 5).Although new-bone formation wasgreater in the PDLDC group than inthe CDC and control groups, the dif-ferences did not reach statistical sig-nificance.

The results from the histometricalanalysis in the experimental (CDCsand PDLDCs) and control groups areshown in Fig. 6. The dimension of theepithelium (GM-JE) was not signifi-cantly di!erent (p > 0.05) between thecontrol group (2.45 ± 0.20 mm) and theexperimental groups (2.18 ± 0.26 mmfor the CDCs and 1.83 ± 0.64 mm forthe MSCDCs), representing 44, 39 and33% of the defect, respectively. Thedimension of the histological attach-ment (CEJ-JE), however, was signifi-cantly di!erent (p < 0.05) between thecontrol group (3.14 ± 0.57 mm) andthe experimental groups (1.64 ±0.28 mm for the CDCs and 1.68 ±0.50 mm for the PDLDCs), representing56, 31 and 33% of the defect, respec-tively. The dimension of the gingivalconnective tissue adhesion to the rootsurface between the new cementumand the junctional epithelium (JE-CC)was 0.80 ± 0.30 mm in the control,0.30 ± 0.10 mm in the CDCs and0.15 ± 0.08 mm in the PDLDCs;these di!erences were not statisticallysignificant (p > 0.05).

The dimension of the newly formedcementum (CC-Notch) was also sig-nificantly di!erent (p < 0.05) betweenthe control group (1.56 ± 0.39 mm)and the experimental groups (3.98 ±0.59 mm in the CDCs and 4.07 ±0.97 mm in the PDLDCs), representing28, 77 and 74% of the defect, respec-tively. New-bone formation was similarin all groups, being 2.63 ± 0.67 mm in

the control, 2.63 ± 0.44 mm in theCDCs and 3.08 ± 1.06 mm in thePDLDCs, representing 48, 54 and 57%of the defect, respectively (Fig. 6).Overall, there were no di!erences inany parameter evaluated between theCDCs and the PDLDCs, with similarresults being obtained for both groups.

Discussion

The aim of this study was to evaluatewhether a cell therapy based on theimplantation during periodontal sur-gery of CS impregnated with eithercanine CDCs or canine PDLDCs inexperimental intrabony periodontaldefects would result in periodontalregeneration in the beagle dog. Thehistological findings obtained showedthat this regenerative approach usingcell therapy was able to promote sig-nificantly new attachment with forma-tion of new cementum and new bone,and connective tissue attachment to apreviously exposed root surface, whencompared with the control groupwhere the same carrier, but lacking thecells, was implanted in the defects.

The first phase of this investigationwas to isolate canine CDCs and caninePDLDCs because there are few reportsin the literature where these cells havebeen isolated in dogs. In a previousinvestigation from our research group(32) we reported a protocol forthe isolation of human CDCs andPDLDCs, modifying the techniquepreviously reported by Seo et al. (2).We followed this protocol in the pres-ent investigation. Once isolated, thecells obtained were able to grow andproliferate, forming colonies withthe distinct capability of secretingmineralized matrix, as shown by

bright-field microscopy and Von Kossastaining. In order to further charac-terize the cells, we used antibodiesagainst CEMP-1, a protein specificallypresent in human cementoblasts (37).Colonies from explants after 60 d ofgrowth in vitro demonstrated positiveimmunostaining in a similar manner toprevious investigations characterizinghuman cementoblasts (33,37). In con-trast, PDLDCs were negative toa-CEMP-1, but positive to a-CD44and a-STRO-1, antigens previouslyreported to be expressed in MSCs fromhuman PDL (2,38). As there is con-troversy on the use of STRO-1 as asingle marker for stem cells (7), thecombination of several additionalmarkers was used to positively identifyPDLDCs. STRO-1 is also an earlymarker of pre-osteogenic cells, theexpression of STRO-1 being progres-sively lost after cell proliferation anddi!erentiation into mature osteoblasts(39). This pre-osteogenic populationwas further di!erentiated into anosteogenic cell population by growth inan osteogenic medium containingascorbic acid 2-phosphate (39). Weattempted to further characterizePDLDCs using flow cytometry. Usinga large series of antibodies raisedagainst a number of human stem cellmarkers, we found that caninePDLDCs stained positive for CD105but negative for the remaining mark-ers. Unfortunately, the lack of canine-specific antibodies does not allow us toconclude whether PDLDCs are MSCsor if they simply cannot be stained byhuman-specific antibodies. A furthercharacterization of the phenotypiccharacteristics of these cells usingcanine-specific markers is warranted.

One of the key factors in cell therapyis the use of an appropriate carrier toensure cell viability during delivery, ascells need to be maintained in a viablestate and released during the healingprocess. In the present study we used aCS with a resorption time of 72 h. Asimilar carrier has been used to applyrecombinant human bone morphoge-netic protein-2 in the regenerativetreatment of three-wall intrabonydefects in dogs (40) and in the treat-ment of one-wall intrabony periodon-tal defects with the application of

Table 2. Clinical measurements of periodontal defects

Parameter Control (mm) CDCs (mm) PDLDCs (mm)

ABC-CEJ 3.25 ± 0.32 2.62 ± 0.31 3.50 ± 0.50ABC-BD 7.00 ± 0.70 6.12 ± 0.47 6.50 ± 0.64BOC-LOC 3.12 ± 0.12 2.75 ± 0.14 2.75 ± 0.25

Data are shown as mean ± standard deviation; n = 4.Parameters include the distance between the alveolar bone crest and the cemento–enameljunction (ABC-CEJ), the distance between the alveolar bone crest and the base of the defect(ABC-BD) and the buccolingual width of the defect (BOC-LOC).CDCs, cementum-derived cells; PDLDCs, periodontal ligament-derived cells.

Regeneration by cellular implantation 39

growth/di!erentiation factor-5 (41).Hydroxyapatite/b-tricalcium phosphatehas also been used as a carrier for celltherapy in artificially created peri-odontal defects in mandibular molars

in rats (2) and in peri-implant defectsin dogs (11) but the viability of cells inthis sca!old is uncertain.

We have shown that canine CDCsor PDLDCs applied within a CS pro-

moted periodontal regeneration in anexperimental three-wall intrabonyperiodontal defect. Significantly highernew-cementum formation and newperiodontal tissue attachment was

A B

C D

E F

Fig. 4. Cellular therapy promotes connective tissue attachment. After 3 mo of healing in the control group, micrograph A shows a parak-

eratinized epithelium covering the gingiva (a), the root notch (b) and alveolar bone (c). (B) Higher magnification of the framed area shows the

long junctional epithelium (a), connective tissue (b), cementum (c) and dentin (d) [Original magnification · 40; samples were stained with

Toluidine Blue (TB)]. In the cementum-derived cell (CDC) group, micrograph C also depicts parakeratinized epithelium covering the gingiva

(a), the root notch (b) and alveolar bone (c). (D) Higher magnification of the framed area shows the presence of new connective tissue

attachment in the CDC group. The connective tissue fibers (a) coronal to the osseous crest (b) are inserted perpendicularly to the newly

formed cementum (c) and junctional epithelium (d) (original magnification · 40; TB). In micrograph E, the periodontal ligament-derived cell

(PDLDC) group shows parakeratinized epithelium covering the gingiva (a), the root notch (b) and alveolar bone (c). (F) Higher magnification

of the framed area shows the presence of new connective tissue attachment. The connective tissue fibers (a) coronal to the osseous crest (b) are

inserted perpendicularly to the newly formed cementum (c) (original magnification · 40; TB).

40 Nunez et al.

observed when compared with theapplication of CS without the cells.The PDLDC group also regeneratednew cementum resembling a healthyperiodontium, thus suggesting regen-eration ad integrum of the perio-dontal tissues. Also, both CDCs andPDLDCs yielded similar results rela-tive to periodontal regeneration. Inother words, CDCs promoted as muchperiodontal regeneration as PDLDCs.This result is not unexpected becauseCDCs are able to produce specificgrowth factors and form new cemen-tum, factors which are consideredcritical in the regeneration of peri-odontal tissue. Our results provide aproof of principle about the potentialuse of this mode of cell therapy for thetreatment of periodontal intraosseouslesions. In a previous study using celltherapy with a seeding of fibroblast-like cells in artificial fenestrationdefects in dogs, the amount of newcementum formed was only 9–12% inmost specimens (42). Better resultswere reported in similar defects withthe use of hyaluronic acid as a carrierto the fibroblast-like cells, attaining54% of new cementum in the experi-mental group vs. 40% in the controlgroup (43). In this study, the amountof new cementum achieved was 77%

(3.98 mm) for the CDC group and74% (4.07 mm) for the PDLDCgroup, both being significantly greater(p < 0.05) than the control group,which achieved 28% (1.56 mm) of newcementum. These values of new-cementum formation are similar tothose reported in other studies whenEMDs were used to treat buccal-dehiscence defects in monkeys (44) ortwo-wall intrabony defects in dogs(45). In contrast to the results obtainedregarding new cementum, similar levelsof new-bone formation occurred in theexperimental and control groups, rep-resenting 54% in the CDC group, 57%in the PDLDC group and 48% in thecontrol group. These results onnew-bone formation may be explainedby the experimental model used in thisstudy, as the three-wall periodontaldefect has a self-contained anatomysurrounded by bony walls that allowsgood stabilization of the blood clotand the promotion of new-bone for-mation. In spite of this, the PDLDCgroup showed a larger amount of new-bone formation, although the between-group di!erences were not statisticallydi!erent.

One aspect di"cult to avoid duringhealing after periodontal surgical pro-cedures is the apical location of the

gingival margin relative to the CEJ(gingival recession). To avoid thisoutcome, minimally invasive surgicalapproaches have been recentlydescribed for the treatment of peri-odontal defects with the use of bio-logicals (EMD), and minimal gingivalrecession has been reported (46,47). Inthis study, both test groups failed toshow any gingival recession, in con-trast to the control group in which agingival recession measurement of0.72 mm was obtained. These resultsrepresent a gain of coronal level ofhistological attachment of 3.50 mm inthe CDC group, 3.73 mm in thePLCDC group and 2.39 mm in thecontrol group, the results in both testgroups being significantly di!erent(p < 0.05) from the result in the con-trol group.

One of the shortcomings of thisinvestigation was the lack of proof thatthe cells remained active during heal-ing, because the transplanted cells werenot labelled. It was therefore impossi-ble to ascertain how these cells mayhave contributed to the tissue regener-ation achieved. We demonstrated,however, that the cells were alive whenthey were implanted in the caninedefects, and both groups using the celltherapy achieved significantly higher

A B C

Fig. 5. Cell therapy with canine cementum-derived cells (CDCs) and canine periodontal ligament-derived cells (PDLDCs) promotes new bone

formation, PDL and new cellular cementum. (A) Control group: connective tissue fibers (a) coronal to the osseous crest (b), the root notch (c),

cementum (d) and the long junctional epithelium (e) [original magnification · 100; Toluidine Blue (TB) stain]. (B) CDC group: new alveolar

bone formation (a), old alveolar bone (b), root notch (c), new cementum formation (d) and old cementum (e) (original magnification · 100;

TB). (C) PDLDC group: new alveolar bone formation (a), old alveolar bone (b), root notch (c), new cementum formation (d) and old

cementum (e) (original magnification · 100; TB).

Regeneration by cellular implantation 41

histological regenerative outcomeswhen compared with the control usingthe same carrier but lacking the cells.We may speculate that even if the cellsdied shortly after transplantation, theymay have contributed positively to thehealing process. Furthermore, if thecells remained viable, they could fur-ther di!erentiate into those cellsresponsible for the periodontal-regen-eration process.

Although the number of dogs usedin this investigation was relativelysmall, mainly for ethical reasons, theresults obtained clearly suggest that theimplantation of these cells has thepotential to promote regeneration,thus paving the way for future cell-therapy studies in the treatment ofperiodontal disease. In conclusion, thecell therapy developed and tested in thepresent experimental study suggests

that the implantation of CDCs andPDLDCs in a CS in experimentallycreated periodontal defects in a caninemodel promotes the regeneration ofperiodontal tissues through the for-mation of new cellular cementum, newPDL and new bone. The results fromthis experimental investigation cannot,however, be extrapolated to the treat-ment of human patients. For this, first,a reliable and predictable noninvasive

Fig. 6. Histometric results 3 mo after surgery. (A) Histometric landmarks used: gingival margin (GM), cemento–enamel junction (CEJ),

apical extent of the junctional epithelium (JE), coronal level of newly formed cementum (CC) and coronal level of newly formed alveolar bone

(OC). (B) Histometric distances (mm) between the notch and the CEJ, between the GM and the apical end of the junctional epithelium (GM-

JE) and from the CEJ to the gingival margin (CEJ-GM). Length of the coronal level of histological attachment: the distance between the CEJ

and the junctional epithelium (CEJ-JE). Length of the connective tissue adhesion: distance between the most apical extent of the junctional

epithelium and the most coronal cementum (JE-CC), length of the newly formed cementum coronal to the notch (CC-notch) and length of the

newly formed bone coronal to the notch (New bone). The length of the histological attachment level and the amount of new-cementum

formation showed significant di!erences 3 mo after treatment. The unit of analysis is the number of dogs (n = 4), *p < 0.05. (C) Bars show

the percentage distribution of the experimental values of the di!erent parameters shown above (data are given as mean ± standard error of

the mean). *p < 0.05.

42 Nunez et al.

technique for CDC and PDLDCextraction must be developed and tes-ted. Second, these cells must be testedin the laboratory for consistent growthand di!erentiation. Third, their viabi-lity and release must be assessed incontact with di!erent carriers. Finally,safety studies using the appropriatestandard cell therapy protocols inhumans must be carried out before thistherapeutic approach can be imple-mented in human patients. Furtherresearch is warranted to clarify thespecific biological activity in the tissue-regeneration process of cementoblastsand stem cells from PDL.

Acknowledgements

We are grateful to Dr Songtao Shi(Dental School, University of SouthernCalifornia, Los Angeles, CA) forallowing Dr Nunez as a visitingresearcher in his laboratory to learntechniques for isolating human MSCfrom the PDL. We also thank MsCarmen Roman (Institute of MolecularBiology and Genetics-IBGM, Vallado-lid, Spain) for technical assistance.Financial support for these investiga-tions was obtained through grantsfrom the Federacion de Cajas deAhorros de Castilla y Leon, Junta deCastilla y Leon, Spain (VA022A05),Instituto de Salud Carlos III, Madrid,Spain, and the Spanish Ministry ofScience and Innovation, Madrid,Spain. This study was supported bygrants from Federacion de Cajas deAhorro de Castilla y Leon and theSpanish Institute of Health Carlos III(PI07/0766). The authors declare thatthey do not have any conflict ofinterest.

References

1. Bartold PM, Shi S, Gronthos S. Stem cells

and periodontal regeneration. Periodontol

2000 2006;40:164–172.

2. Seo BM, Miura M, Gronthos S et al.

Investigation of multipotent postnatal

stem cells from human periodontal liga-

ment. Lancet 2004;364:149–155.

3. Stamm C, Wesphal B, Kleine HD et al.

Autologous bone-marrow stem-cell trans-

plantation for myocardial regeneration.

Lancet 2003;361:45–56.

4. Derubeis AR, Cancedda R. Bone marrow

stromal cells (BMSCs) in bone engineer-

ing: limitations and recent advances. Ann

Biomed Eng 2004;32:160–165.

5. Ivanovski S, Gronthos S, Shi S, Bartold

PM. Stem cells in the periodontal liga-

ment. Oral Dis 2006;12:358–363.

6. Zhou Y, Hutmacher DW, Sae-Lim V,

Zhou Z, Woodruff M, Lim TM. Osteo-

genic and adipogenic induction potential

of human periodontal cells. J Periodontol

2008;79:525–534.

7. Coura GS, Garcez RC, Mendes de Aguiar

CBN, Alvarez-Silva M, Magini RS,

Trentin AG. Human periodontal liga-

ment: a niche of neural crest stem cells.

J Periodontal Res 2008;43:531–536.

8. Huang GT-J, Gronthos S, Shi S. Mesen-

chymal stem cells derived from dental tis-

sues vs. those from other sources: their

biology and role in regenerative medicine.

J Dent Res 2009;88:792–806.

9. Majumdar MK, Keane-Moore M, Buya-

ner D. Characteterization and functional-

ity of cell surface molecules on human

mesenchymal stem cell. J Biomed Sci

2003;10:228–241.

10. Dominici M, Le Blanc K, Mueller I et al.

Minimal criteria for defining multipotent

mesenchymal stromal cells. The Interna-

tional Society for Cellular Therapy position

statement. Cytotherapy 2006;8:315–317.

11. Kim SH, Kim KH, Seo BM et al. Alveolar

bone regeneration by transplantation of

periodontal ligament stem cells and bone

marrow stem cells in a canine peri-implant

defect model: a pilot study. J Periodontol

2009;80:1815–1823.

12. Gronthos S, Mankani M, Brahim J,

Robey PG, Shi S. Postnatal human dental

pulp stem cells (DPSCs) in vitro and

in vivo. Proc Natl Acad Sci USA 2000;97:

13625–13630.

13. Melcher AH. On the repair potential of

periodontal tissues. J Periodontol 1976;47:

256–260.

14. Miura M, Gronthos S, Zhao M et al.

SHED: stem cells from human exfoliated

deciduous teeth. Proc Natl Acad Sci USA

2003;100:5807–5812.

15. Sonoyama W, Liu Y, Fang D et al. Mes-

enchymal stem cell mediated functional

tooth regeneration in swine. PLoS ONE

2006;1:e79.

16. Zaky SH, Cancedda R. Engineering cra-

niofacial structures: facing the challenge.

J Dent Res 2009;88:1077–1091.

17. Weng Y, Wang M, Liu W et al. Repair of

experimental alveolar bone defects by tis-

sue-engineered bone. Tissue Eng 2006;12:

1503–1513.

18. De Kok IJ, Drapeau SJ, Young R,

Cooper LF. Evaluation of mesenchymal

stem cells following implantation in alve-

olar sockets: a canine safety study. Int J

Oral Maxillofac Implants 2005;20:511–

518.

19. Yamada Y, Ueda M, Hibi H, Nagasaka

T. Translational research for injectable

tissue-engineered bone regeneration using

mesenchymal stem cells and platelet-rich

plasma: from basic research to clinical

case study. Cell Transplant 2004;4:343–

355.

20. Shayesteh YS, Khojasteh A, Soleimani M,

Alikhasi M, Khoshzaban A, Ahmadbeigi

N. Sinus augmentation using human

mesenchymal stem cells loaded into a

beta-tricalcium phosphate/hydroxyapatite

scaffold. Oral Surg Oral Med Oral Pathol

Oral Radiol Endod 2008;106:203–209.

21. Moreau JL, Caccamese JF, Coletti DP,

Sauk JJ, Fisher JP. Tissue engineering

solutions for cleft palates. J Oral Max-

illofac Surg 2007;65:2503–2511.

22. Hibi H, Yamada Y, Ueda M, Endo Y.

Alveolar cleft osteoplasty using tissue-

engineered blood osteogenic material. Int

J Oral Maxillofac Surg 2006;35:551–555.

23. Nyman S, Gottlow J, Karring T, Lindhe J.

The regenerative potential of the peri-

odontal ligament. An experimental study

in the monkey. J Clin Periodontol 1982;9:

257–265.

24. Lang H, Schuler N, Arnhold S, Nolden R,

Mertens T. Formation of differentiated

tisues in vivo by periodontal cell popula-

tions cultured in vitro. J Dent Res

1995;74:1219–1225.

25. Brunette DM, Melcher AH, Moe HK.

Culture and epithelium-like and fibro-

blast-like cells from porcine periodontal

ligament explants and cell suspensions.

Arch Oral Biol 1976;21:393–400.

26. Somerman MJ, Archer SY, Imm GR,

Foster RA. A comparative study of

human periodontal ligament cells and

gingival fibroblast in vitro. J Dent Res

1988;67:66–70.

27. Piche JE, Carnes DL Jr, Graves DT. Ini-

cial characterization of cells derived from

human periodontia. J Dent Res 1989;68:

761–767.

28. Grzesik WJ, Kuzentsov SA, Uzawa K,

Mankani M, Gehron Robey P, Yamauchi

M. Normal human cementum-derived

cells: isolation, clonal expansion, and in

vitro and in vivo characterization. J Bone

Miner Res 1998;13:1547–1554.

29. Preisig E, Schroeder HE. Long-term cul-

ture of human periodontal ligament cells

with autologous root discs. J Periodontal

Res 1988;23:211–216.

30. Boyko GA, Melcher AH, Brunette DM.

Formation of new periodontal ligament

by periodontal ligament cells implanted in

vivo after culture in vitro. J Periodontal

Res 1981;16:73–88.

31. Zhao M, Jin Q, Berry JE, Nociti FH Jr,

Giannobile W, Somerman MJ. Cemento-

Regeneration by cellular implantation 43

blast delivery for periodontal tissue engi-

neering. J Periodontol 2004;75:154–161.

32. Nunez J, Sanz-Blasco S, Vignoletti F et al.

17ß-estradiol promotes cementoblast pro-

liferation and cementum formation in

experimental periodontitis. J Periodontol

2010;81:1064–1074.

33. Nunez J, Sanz M, Hoz-Rodrıguez L,

Zeichner-David M, Arzate H. Human

cementoblasts express enamel-associated

molecules in vitro and in vivo. J Periodon-

tal Res 2010;45:809–814.

34. Karnosky MJ. A formaldehyde-glutaral-

dehyde fixative of high osmolarity for use

in electrom microscopy. J Cell Biol

1965;27:137A–138A.

35. Donath K, Breuner GA. Method for the

study of undecalcified bones and teeth

with attached soft tissue. The Sage-Schliff

(sawing and grinding) techniques. J Oral

Pathol 1982;11:318–326.

36. Jeno L, Geza L. A simple differential

staining method for semi-thin sections of

ossifying cartilage and bone tisues embed-

ded in epoxy resin. Mikroskopie 1975;31:

1–4.

37. Alvarez-Perez MA, Narayanan S, Zeich-

ner-David M, Rodrıguez Carmona B,

Arzate H. Molecular cloning, expression

and immunolocalization of a novel human

cementum-derived protein (CP-23). Bone

2006;38:409–419.

38. Trubiani O, Di primio R, Traini T et al.

Morphological and cytofluorimetric anal-

ysis of adult mesenchymal stem cells

expanded ex vivo from periodontal liga-

ment. Int J Immunopathol Pharmacol 2005;

18(2):213–221.

39. Ohta S, Yamada S, Matuzaka K, Inoue T.

The behaviour of stem cells and progeni-

tor cells in the periodontal ligament during

wound healing as observed using immu-

nohistochemical methods. J Periodontal

Res 2008;43:595–603.

40. Choi SH, Kim CK, Cho KS et al. Effect

of recombinat human bone morphoge-

netic protein-2/absorbable collagen sponge

(rhBMP-2/ACS) on healing in 3-wall

intrabony defects in dogs. J Periodontol

2002;73:63–72.

41. Kim TG, Wikesjo UM, Cho KS et al.

Periodontal wound healing/regeneration

following implantation of recombinant

human growth/differentiation factor-5

(rhGDF-5) in an absorbable collagen

sponge carrier into one-wall intrabony

defects in dogs: a dose-range study. J Clin

Periodontol 2009;36:589–597.

42. Dogan A, Ozdemir A, Kubar A, Oygur T.

Healing of artificial fenestration defects by

seeding of fibroblast-like cells derived

from regenerated periodontal ligament in

a dog: a preliminary study. Tissue Eng

2003;9:1189–1196.

43. Akizuki T, Oda S, Komaki M et al.

Application of periodontal ligament cell

sheet for periodontal regeneration: a pilot

study in beagle dogs. J Periodontal Res

2005;40:245–251.

44. Hammarstrom L, Heijl L, Gestrelius S.

Periodontal regeneration in a buccal

dehiscence model in monkeys after appli-

cation of enamel matrix proteins. J Clin

Periodontol 1997;2:669–677.

45. Yamamoto S, Masuda H, Shibukawa Y,

Yamada S. Combination of bovine-de-

rived xenografts and enamel matrix

derivative in the treatment of intrabony

periodontal defects in dogs. Int J Peri-

odontics Restorative Dent 2007;27:471–

479.

46. Cortellini P, Nieri M, Prato GP, Tonetti

MS. Single minimally invasive surgical

technique with an enamel matrix deriva-

tive to treat multiple adjacent intra-bony

defects: clinical outcomes and patient

morbidity. J Clin Periodontol 2008;35:

605–613.

47. Fickl S, Thalmair T, Kebschull M, Bohm

S, Wachtel H. Microsurgical access flap in

conjunction with enamel matrix derivative

for the treatment of intra-bony defects: a

controlled clinical trial. J Clin Periodontol

2009;36:784–790.

44 Nunez et al.