17b-Estradiol Promotes CementoblastProliferation and Cementum Formationin Experimental PeriodontitisJavier Nunez,*† Sara Sanz-Blasco,* Fabio Vignoletti,† Fernando Munoz,‡ Raul G. Caffesse,†

Mariano Sanz,† Carlos Villalobos,* and Lucıa Nunez*

Background: Periodontitis is a chronic inflammatory condition thatleads to the destruction of the tooth-supporting tissues. Its treatment in-cludes the arrest of the inflammatory process and, in some circum-stances, the restoration of the lost anatomy and function, includingthe formation of new cementum, periodontal ligament (PDL), andbone. With this goal, we investigated the effects of low concentrationsof 17b-estradiol on human cementoblast proliferation and its possibleregenerative potential in vivo.

Methods: Human cementum-derived cells obtained from a healthyhuman premolar were isolated and characterized by immunocyto-chemistry. Cell proliferation assays were performed to test the effectsof 100 nM 17b-estradiol and enamel matrix derivative (EMD). Three-wall intrabony periodontal defects were created in beagle dogs. After1 month of plaque accumulation, 0.225 mg 17b-estradiol impregnatedin a collagen sponge was applied to randomly selected defects (testgroup), whereas a collagen sponge impregnated in a culture mediumwas applied to the control group. After 3 months, specimens wereobtained, and tissue regeneration was assessed by histometric analy-sis.

Results: Cells spreading out from human tooth-layer explants wereable to form cell colonies, produce a mineral matrix, and express osteo-calcin, indicating they were cementoblast-like cells. In contrast, PDLfibroblasts did not express osteocalcin. 17b-estradiol, but not EMD, in-creased the rate of human cementoblast cell proliferation in vitro by2.5-fold. Histometric results from the treated periodontal defectsrevealed that 17b-estradiol promoted the formation of 2.94 mm ofnew cementum, (45% of the defects) compared to 1.54 mm of new ce-mentum in the control group (28% of the defects). Furthermore, the testgroup showed an inhibition of epithelial downgrowth and a gain of newconnective tissue attachment.

Conclusion: 17b-estradiol promoted human cementoblast cell pro-liferation in vitro and periodontal regeneration in an experimental peri-odontitis model. J Periodontol 2010;81:1064-1074.

KEY WORDS

Cementoblasts; dental cementum; estradiol; periodontal disease;regeneration.

Periodontitis is a chronicinflammatorydiseasecau-sed by the bacteria resid-

ing in the subgingival biofilmand is characterized by the de-struction of tooth-supportingtissues, including the cementum,periodontal ligament (PDL), andalveolar bone. The main objec-tive of current periodontal ther-apy is to arrest this inflammatoryprocess by eliminating or re-ducing the subgingival plaquebacteria. Once this is accom-plished, another goal of treat-ment is the reconstruction ofthe periodontal tissues lost asa consequence of the disease.To achieve this, several surgicalregenerative approaches havebeen attempted with differentdegrees of success, but all ap-proaches have the formation ofnew cementum as a critical bi-ologic goal because its deposi-tion precedes the attachment ofPDL collagen fibers to the rootsurface.1,2

Cementum is a mineralizedtissue with a similar structure andcomposition to bone, althoughit does not contain vessels andunder normal conditions doesnot suffer remodeling. Threetypesofcementumhavebeende-scribed: 1) the acellular afibrillar

* Institute of Molecular Biology and Genetics, University of Valladolid, and Spanish Research Council,Valladolid, Spain.

† Faculty of Odontology, Complutense University of Madrid, Madrid, Spain.‡ Faculty of Veterinary, University of Santiago de Compostela, Santiago de Compostela, Spain.

doi: 10.1902/jop.2010.090678

Volume 81 • Number 7

1064

cementum (mineralized matrix without collagenfibers and cells), 2) the acellular extrinsic fiber ce-mentum (mineralized matrix containing collagen fi-bers), and 3) the cellular intrinsic fiber cementum(mineralized matrix, collagen fibers, and cells entrap-ped in the mineralized matrix).3 These latter cellsderive from cementoblasts, which are cells that areable to secrete mineralized matrix-containing non-collagenous proteins such as bone sialoprotein,osteopontin, osteocalcin, vitronectin, and fibronec-tin.4 Therefore, the presence of cementoblasts is crit-ical for the formation of reparative cementum andfor the regeneration of a functional periodontal appa-ratus.

Different therapeutic strategies aimed at regener-ating cementum have been proposed, although theuse of enamel matrix derivative (EMD) is the onlyone resulting in clinically significant outcomes.5,6

The ability of EMD in vitro to stimulate mesenchy-mal cells derived from the dental follicle and to dif-ferentiate them into cementoblasts,7 as well as toenhance the proliferation of human PDL fibroblasts,was demonstrated.8,9 Animal6 and human10 studiesdemonstrated that when EMD was applied to af-fected root surfaces, the key biologic elements werethe deposition of acellular cementum6 and theformation of new periodontal attachment with per-pendicularly inserted fibers to the newly formedcementum.10

The relationship between estrogens and periodon-tal tissues was studied mainly for its possible implica-tion in the inflammatory process, with a cleardemonstration that estrogens do not increase inflam-mation.11-14 However, there is a lack of informationregarding other possible direct effects of estrogenson periodontal tissues, despite the fact that mostcells in the periodontium (fibroblasts, endothelialcells, epithelial gingival cells, osteoclasts, and oste-oblasts) express estrogen receptors a and b.15-17 It iswell known that estrogens decrease bone resorptionand have a positive effect on bone formation.18 Anin vitro study19 with PDL fibroblasts showed en-hanced alkaline phosphatase activity and mineral-ized nodule formation when 17b-estradiol wasadded to the cell-culture medium. However, to ourknowledge, there is no information on the possibleeffect of 17b-estradiol on cementoblast activity. Inlight of this paucity of information on effects of17b-estradiol on periodontal tissues, and specificallyon cementum-forming cells, the aim of this investi-gation is to isolate human cementoblast-like cellsin vitro and to study the effects of 17b-estradiol ontheir proliferation. In addition, we investigated thepossible effect of 17b-estradiol on promoting peri-odontal regeneration in an experimental peri-odontitis model in beagle dogs.

MATERIALS AND METHODS

Isolation and Culture of Human PDL Fibroblastsand Human Cementum-Derived Cells (HCDCs)A healthy, human upper right premolar from an11-year-old male patient was extracted in May 2006in a private dental office in Valladolid, Spain, under or-thodontic prescription. Written consent was given afterthe patient and his parents were informed of the need forextraction for orthodontic reasons, and scientific use ofthe extracted teeth. The protocol of this investigationwas approved by the Institutional Review Board for Hu-man Subjects at Complutense University of Madrid,Faculty of Odontology, and was conducted in accor-dance with the Helsinki Declaration of 1975, as revisedin 2000. The tooth was kept in serum-free and phenolred-free Dulbecco’s modified Eagle’s medium (DMEM)containing 100 U/ml penicillin and 100 mg/ml strepto-mycinovernight at4�C.ThePDL fromthe tooth rootwasscraped with a Gracey curet.§ The tissue was digestedby incubation with minimal essential medium (MEM)containing collagenase P for 90 minutes at 37�C. Afterdigestion, the cell suspension was collected, centrifugedfor 7 minutes at 200 · g, and washed four times withfresh medium. Next, the cells were plated on Petri dishescontaining phenol red-free nutrient mixture (DMEM/F12) medium supplemented with 10% charcoal-stripped fetal bovine serum (FBS) in a 5% CO2/95%air incubator at 37�C. Under these conditions, fibro-blasts grew after 5 days of incubation without formingcolonies. When the cultures became semiconfluent,the cells were rinsed with trypsin-EDTA for subcul-ture. PDL fibroblasts of the fourth and fifth passageswere used for the experiments.

With the use of 7/8 Gracey curet,i the most externalroot layer containing remains of the PDL that includecollagen fibers and fibroblasts was removed. Then,the tooth was immersed in MEM containing collage-nase P for 90 minutes at 37�C. After digestion, the toothwas washed four times with fresh medium, and thinlayers were removed using a surgical scalpel. Theselayers were washed and cut in smaller pieces using mi-crosurgery scissors, and pieces were digested again inMEM containing collagenase P for 2 hours at 37�C. Thedigested tissue was washed and dispersed with a sili-cone Pasteur pipette. Small tissue pieces were platedin a Petri dish containing phenol red-free DMEM/F12medium supplemented with 10% charcoal-strippedFBS in a 5% CO2/95% air incubator at 37�C. After 10days of incubation and under these conditions, HCDCsgrew, forming discrete colonies. When the cultures be-came semiconfluent, the cells were rinsed with trypsin-EDTA medium and subcultured. HCDCs from thefourthandfifthpassages were used for the experiments.

§ Hu-Friedy, Chicago, IL.i Hu-Friedy.

J Periodontol • July 2010 Nunez, Sanz-Blasco, Vignoletti, et al.

1065

ImmunocytochemistryHCDCs and PDL cells were plated at 104 cells/ml onpoly-L-lysine (0.01 mg/ml)–coated coverslips, andcultured for 4 to 5 days. Then, these cells were fixedwith 4% formaldehyde and permeated with 0.1% (vol-ume/volume [v/v]) Triton ·100 in phosphate bufferedsaline (PBS). The blocking of non-specific bindingsites was carried out using 20% (v/v) goat serum for10 minutes. Primary antibodies dissolved in 10% goatserum were used against osteocalcin¶ and incubated1:100 overnight at 4�C. Samples were rinsed with PBSand further incubated with rabbit, anti-mouse fluores-cein isothiocyanate secondary antibodies# for 60minutes. Stained cells were imaged using a fluores-cence microscope coupled to a digital camera.**

Cell Proliferation AssayHCDCs were trypsinized and plated at a density of 104

cells/well in sterile six-well plates. To test the effectsof 17b-estradiol on cell proliferation, the obtainedHCDCs were cultured with phenol red-free DMEM/F12 medium containing 2% charcoal-stripped FBS,antibiotics, and vehicle (dimethyl sulfoxide [DMSO])or 17b-estradiol 100 nM dissolved in DMSO. As a con-trol, we used EMD†† in this cell proliferation assay byculturing HCDCs with phenol red-free DMEM/F12medium containing 2% charcoal-stripped FBS, anti-biotics, and acetic acid (5 mM) or 150 ml EMD dis-solved in 5 mM acetic acid. EMD is not soluble atphysiologic pH and body temperature because of itsprotein composition and requires acid pH or low tem-peratures to increase its solubility.8 Therefore, it wasdissolved in 5 mM acetic acid, which did not interferewith the experimental conditions.9 After 60 days invitro (DIV), the osteocalcin-positive (cementoblast-like) cells were detached from the culture and countedusing a counting chamber (hemocytometer). Thesecells were plated again in the presence of differenttreatments for 7 days. Replicate wells were countedat the beginning and after 7 days of treatment, andthree samples of each well were counted separatelyusing the hemocytometer.

Experimental Animal StudyFour periodontally healthy, male beagle dogs‡‡ (age:1.5 years) weighing 10 to 12 kg were used in thisstudy. Once the animal study protocol was approvedby the Ethical Research Committee of the Gomez-UllaCentral Military Hospital, Madrid, Spain, the animalsunderwent the prescribed quarantine period. Withthe use of ultrasonic scalers and manual curets, su-pra- and subgingival calculus deposits were removedto reduce gingival inflammation, and the first and thirdmandibular premolars were extracted under generalanesthesia, which was achieved by sedation with 80mg/kg medetomidine,§§ 20 mg/kg utorfanol,ii and100 mg/kg atropine sulfate¶¶ and then by anesthesia

with a mixture of sevofluorane 2% supplemented by2% lidocaine (1:50,000 epinephrine) as a local anes-thetic and oxygen using a mechanical respirator. Onemonth after extractions and after the same anestheticprotocol, full-thickness flaps were elevated using in-trasulcular incisions from the mesial aspect of the sec-ond mandibular premolars to the mandibular firstmolars on both sides, buccally and lingually. Withthe use of a handpiece and fissure burs at low speedunder copious irrigation, 3-wall periodontal defectsmeasuring 3 to 4 mm in bucco-lingual, mesio-distal,and apico-coronal directions were surgically createdat the mesial aspect of fourth premolars and the me-sial and distal aspects of second mandibular pre-molars. The periodontal intrabony defects werecarefully debrided, and the roots were planed withcurets until the contaminated cementum was fully re-moved. Orthodontic wire ligatures were placed aroundthe teeth and inserted into the defects, and the muco-periosteal flaps were repositioned and sutured with bi-oabsorbable interrupted sutures.## During healing,the animals were fed with a soft diet to accumulateplaque and induce periodontal inflammation.

After 1 month ofplaqueaccumulation,animals weresedated with 80 mg/kg medetomidine, 20 mg/kg butor-fanol, and 100 mg/kg atropine sulfate, and ligatureswere removed together with the accumulated plaqueandcalculususingultrasonicscalers.Clinicalmeasure-ments were recorded including the probing depth (PD),clinical attachment level (CAL), and width of the kerati-nized gingiva (KG) at the mid-buccal aspect of the sec-ondandfourthmandibularpremolars.Sevendayslater,the surgical procedure, which aimed to treat the peri-odontal defects, was performed using the same anes-thetic protocol described previously. Buccal andlingual mucoperiosteal flaps were elevated includingboth mandibular premolars. The periodontal intrabonydefects were carefully debrided, and the roots wereplaned with curets. The following intrabony measure-ments were recorded with a periodontal probe: the dis-tance between the alveolar bone crest (ABC) and thecemento-enamel junction (CEJ), the distance betweenABC to the base of the defect (BD), and the bucco-lingual width of the defect. With the use of a round bur,a reference notch was made at the BD. The periodontalbone defects, which were randomized by the toss ofa coin, were treated with a collagen sponge***

¶ Alexis Biochemicals, Lausen, Switzerland.# Invitrogen, Barcelona, Spain.** Nikon Eclipse 30i fluorescence microscope coupled to a DM 1200c

digital camera, Amstelveen, The Netherlands.†† Straumann, Basel, Switzerland.‡‡ Isoquimen, Barcelona, Spain.§§ Pfizer, Madrid, Spain.ii Fort Dodge, Gerona, Spain.¶¶ Pharmaceutical Institute FAS, Burgos, Spain.## Vicryl, Ethicon, Somerville, NJ.*** Xemax Surgical Products, Napa, CA.

Estradiol Promotes Cementum Formation Volume 81 • Number 7

1066

Figure 1.Culture and characterization of HCDCs and PDL fibroblasts. Layers froma human tooth were removed and digested using collagenase. A) Cellsfrom cultured tissue explants started to spread out after 10 DIV. Cellsstarted to form colonies, and mineral formation was apparent. Arepresentative example at 60 DIV is shown. B) Suspended cultured cellsfrom the first layer started to grow after 5 DIV and formed confluentmonolayers with no colony or mineral formation at 10 DIV. C) HCDCs

forming colonies were subjected to immunocytochemistry (ICC), stainingwitha-osteocalcin antibodyafter 4and10weeks, respectively, andpositivecells were revealed using a fluorescein-conjugated second antibody. D)PDL fibroblast cultures were incubated with the same antibody, but noimmunofluorescence was observed. E) Control ICC: first antibody wasomitted in HCDCs, and no immunofluorescence was observed. Left =immunofluorescence images; right = bright-field images (Figures 1Athrough E). Bars = 10 mm. Left and right panels at same magnification.

Figure 2.Effects of 17b-estradiol and EMD on HCDC proliferation. A) HCDCsforming colonies were detached from the Petri dish, counted usinga hemocytometer, and plated in the presence of either vehicle (DMSO) or17b-estradiol (100 nM). Cells were allowed to grow for 7 days, detachedagain, and counted. Pictures show bright-field images of HCDCs after 7DIV and cultured in vehicle (control) or 17b-estradiol. The number of cells(mean – SEM) before and after 7 DIV in both conditions is shown. Thegrowth rate in control and experimental conditions is also shown. B)HCDCs forming colonies were detached, counted, and plated in thepresence of either vehicle (acetic acid) or EMD (150 mg/ml). Cells wereallowed to grow for 7 days and counted. Pictures show representativebright-field images of HCDCs after 7 DIV and cultured in vehicle (aceticacid) or EMD. The number of cells (mean – SEM) before and after 7 DIVin both conditions is shown. The growth rate (cells per hour) in control andexperimental conditions is also shown. EMD did not affect HCDCproliferation. *P <0.05 versus control (Student t test). Bar = 10 mm.

J Periodontol • July 2010 Nunez, Sanz-Blasco, Vignoletti, et al.

1067

impregnated with either 50 ml culture medium (controlsites)or0.225mg17b-estradiol(testsites).Thegingivalflaps were repositioned and sutured with bioabsorbableinterruptedsutures.Post-surgically,all animals receivedantibiotics (amoxicillin, 15 mg/kg, intramuscularly, ev-ery 48 hours for 7 days) and anti-inflammatory medica-tion (meloxicam, 0.2 mg/kg, subcutaneously, every 24hours for3days)andwere fedwithasoft diet for2weeksto reduce potential mechanical trauma to the surgicalsites. Every third day, a 0.12% chlorhexidine gluconatespray was topically applied for 3 months.

Histologic MethodsThree months postoperatively, the animals were sac-rificed with an overdose of sodium pentobarbital andfixedbyvascularperfusion through thecarotidarterieswith a Karnovsky fixative solution.20 The lower jawswere removed and immersed in Karnovsky solutionfor 1 week. Both experimental premolars were hemi-sected, and tissue blocks containing the treated rootand its surrounding soft and hard tissues were dis-sected and processed for ground sectioning accordingto the method described by Donath and Breuner.21

From each tissue block, one mesio-distal sectionwas prepared and reduced to a thickness of ;20 mmusingamicrotome.††† Thesectionswere stainedusingthe Levai Laczko technique.22 All specimens wereanalyzed histologically and histometrically undera light microscope‡‡‡ equipped with a computerizedimage-analysis system.§§§ Awell-trained,maskedex-aminer (JN) carried out the histologic evaluation andhistometric analyses measuring all parameters in du-plicate. The following measurements were carried outin each section: lengthof the defect = distance from theCEJ to the notch; length of the epithelium = distancebetween the gingival margin (GM) and the apicalend of the junctional epithelium (JE) (GM–JE); lengthof the connective tissue adhesion = distance betweenthe most coronal cementum (CC) and the most apicalextent of the JE (CC–JE); new cementum formation =length of the newly formed cementum coronal to thenotch, apical cementum at notch (AC) to coronal ce-mentum (CC) (AC–CC); new bone = length of thenewly formed bone coronal to the notch; length ofthe CEJ to the osseous crest (OC) (CEJ–OC); andthe length between the CEJ to the GM.

Data AnalysesDifferences between groups in the proliferation assaywere tested by means of the Student t test. In the an-imal experimental study, the beagle dog was the unitof analysis (N = 4). Means and standard deviationswere calculated for each clinical and histologic pa-rameter in both experimental and control groups,and differences between groups were analyzed bymeans of the Student t test.

RESULTS

Proliferation Assay With PDL Fibroblastsand HCDCsOnce the tissue explants were plated, after 10 DIV, thecells started to spread out from the explants. After 60DIV, HCDCs started to form colonies where mineralformation was apparent (Fig. 1A) and later confirmedwith von Kossa staining (not shown). In contrast, thegrowth of PDL fibroblasts was apparent after a few DIVand formed a confluent cell culture after 10 DIV. Thesecells were not able to form cell colonies and failed togenerate any mineral matrix (Fig. 1B). The phenotypeof cementoblast-like cells was further tested by immu-nofluorescence against osteocalcin, a protein pro-duced and released by hard-tissue forming cells butnot by PDL fibroblasts. Figure 1C shows the strong im-munofluorescence staining of HCDC explants after 4and 10 weeks, respectively. Cell colonies spreadingout from the tissue explants were able to produceosteocalcin, whereas PDL fibroblasts were not stainedby the a-osteocalcin antibody (Fig. 1D). HCDCs that

Table 1.

Clinical Measurements (mm; mean – SEM;N = 4) Recorded at Baseline (BL) and 1Month After Placing Ligatures

Control 17b-Estradiol

Parameter BL 1 Month BL 1 Month

PD 1.62 – 0.12 4.75 – 0.14* 1.75 – 0.28 4.25 – 0.25*

CAL 0.25 – 0.25 5.25 – 0.32* 0.25 – 0.25 4.25 – 0.25*

KG 3.50 – 0.28 3.25 – 0.47 3.75 – 0.25 3.75 – 0.25

* P <0.05 (Student t test).

Table 2.

Clinical Measurements (mm; mean – SEM;N = 4) of Periodontal Defects

Parameter Control 17b-Estradiol

ABC–CEJ 3.25 – 0.32 3.00 – 0.57

ABC–BD 7.00 – 0.70 6.50 – 0.95

BOC–LOC 3.12 – 0.12 3.25 – 0.25

ABC–CEJ = distance between the alveolar bone crest (ABC) to the CEJ;ABC–BD = distance between the ABC to the base of the defect (BD); BOC–LOC = distance between the bucco-lingual width of the defect.

††† Exakt Apparatebau, Norderstedt, Germany.‡‡‡ Eclipse E800, Nikon, Tokyo, Japan.§§§ NIS Elements BR, Nikon DS-Ri1, Amstelveen, The Netherlands.

Estradiol Promotes Cementum Formation Volume 81 • Number 7

1068

were not incubated with the a-osteocalcin antibodywere also negative (Fig. 1E).

Effects of 17b-Estradiol and EMD on HCDCProliferationOnce HCDCs were successfully isolated and cul-tured, the effects of low concentrations of 17b-estra-

diol and EMD on cell proliferation weretested. Vehicle-treated (control) cellsshowed limited proliferation (Fig.2A). In contrast, cells treated with 17b-estradiol showed a 2.5-fold increase inthe rate of cell proliferation above con-trol cells (Fig. 2A). However, the groupof cells treated with EMD showed a sim-ilar proliferation rate compared to con-trol cells not treated with EMD, asshown in Figure 2B.

Effects of 17b-Estradiol on PeriodontalDefects in an Experimental DogModelOnce the effects of estradiol in vitrowere characterized, we aimed at inves-tigating the effects of this compoundin a dog model of experimental peri-odontitis. Table 1 depicts the clinicalmeasurements of PD, CAL, and KG atbaseline and after placing the ligatures.Differences between parameters (inmillimeters) at baseline and after re-moving the ligature were statisticallysignificant (P <0.05) for PD and CALin both groups, whereas KG did notchange. Table 2 shows the clinical mea-surements of the created periodontaldefects including the distances (inmillimeters) between the ABC and theCEJ, the ABC to the BD, and thebucco-lingual width of the defect. Simi-lar parameters were obtained, thereforerepresenting comparable conditions forthe periodontal bone defects in bothgroups.

To assess the in vivo effects of 17b-es-tradiol, experimentally created periodon-tal defects in beagle dogs were treatedwith a collagen sponge bathed with eitherestradiolor theculturemediaasacontrol.Four weeks after the regenerative sur-gery, healing occurred uneventfully, andno visible adverse reactions, such as in-fection or suppuration, were observed.

Histologic evaluations were carriedout in all defects from the four animals.No signs of acute inflammation were ob-

served. The histologic findings differed between con-trol and test groups. The control group showedhealing of the defect by repair. This was characterizedby a reduced amount of new cementum formation onthe previously denuded and contaminated dentin sur-face, which did not stop the apical migration of the JEand,hence, resulted in a significantly less dimension of

Figure 3.A) Schematic of the histometric measurements including GM, JE (apical border of junctionepithelium), CC (coronal level of newly formed cementum), OC (coronal level of newly formedalveolar bone), and AC (apical border of newly formed cementum). B) Histometric distancesbetween the notch to the CEJ (Notch–CEJ), GM–JE, AC–CC, CEJ–OC, CC–JE, and length of thenewly formed bone coronal to the notch (New Bone). The length of the epithelial migration andthe amount of new cementum formation showed significant differences 3 months aftertreatment. C) Bars show percentages of experimental values of the different parameters(mean – SEM values). N = 4 *P <0.05 (Student t test).

J Periodontol • July 2010 Nunez, Sanz-Blasco, Vignoletti, et al.

1069

connective tissue attachment compared to the testgroup. In contrast, the healing of the defect in the testgroup showed characteristics of regeneration, with anincreased amount of new cementum formation anda significantly higher dimension of connective tissueattachment. In both groups, a parakeratinized epithe-lium covered the oral GM, and a small inflammatoryinfiltrate was identified confined within the connectivetissues adjacent to the sulcular and junctional epithe-lia. The amount of new bone formation coronal to thethick alveolar crest was similar in both groups. In thePDL space between the newly formed bone and ce-mentum, rich capillary vessels were observed morefrequently in the test group. Coronal to the bone crestthe connective tissue fibers ran parallel to the root sur-face in the control group, whereas they ran perpendic-ular and inserted into the newly formed cementum inthe test group. No signs of root resorption or ankylosiswere observed in either group.

The histometric results from theexperimental and control groupspecimens are shown in Figure 3.Significant differences between thecontrol group (2.45 – 0.20 mm)and experimental group (1.77 –0.24 mm) (P <0.05) were found in re-gards to the dimension of the epithe-lium (GM–JE), representing 44% and27.2% of the defect height, respec-tively (Fig. 4). The dimension of theconnective tissue attachment to theroot surface between the new cemen-tum and the JE (CC–JE) was 0.80 –0.30 mm for the control group and0.84 – 0.47 mm for the experimentalgroup. In the experimental group, thedisposition of supracrestal connec-tive tissue fibers was perpendicularto the root surface and insertedinto the newly formed cementum,whereas in the control group, the fi-bers ran parallel to the root surface(Fig. 4). Newly formed cementum(AC–CC) on the previously denudedand contaminated dentin surface wasobserved in both groups, but a signif-icantly higher dimension was foundin the experimental group (2.94 –1.02 mm) when compared to the di-mension in control group (1.54 –0.38 mm), representing 45% and28% of the defects for the experimen-tal and control groups, respectively(Fig. 5). In contrast, new bone forma-tion was similar in both groups andamounted to 2.63 – 0.67 mm for

the control group and 3.23 – 0.45 mm for the exper-imental group, representing 48% and 49%, respec-tively (Fig. 5).

DISCUSSION

The aim of the present study is to develop a reliableprocess for assessing the effect of 17b-estradiol onthe proliferation of HCDCs and PDL-derived fibro-blasts and, further, to assess whether the use of estra-diol would promote periodontal regeneration in anexperimental periodontitis model. The method for iso-lating and growing HCDCs was based on a protocolpreviously described by Grzesik et al.23 These prolif-erated cells were able to form colonies and secretea mineralized matrix. Moreover, these cells werestrongly stained with antibodies against osteocalcin,a protein specifically present in hard tissue–formingcells (cementoblasts or osteoblasts). In contrast,PDL fibroblasts were readily visible after only several

Figure 4.Histomorphometric analysis of periodontal defects and effects of estradiol. After 3 months ofhealing, a parakeratinized epithelium covered the GM in both groups. A) Control-group specimenwith parakeratinized epithelium covering the gingiva (a), root notch (b), and alveolar bone (c). B)Higher magnification of the framed area with long JE (a), connective tissue (b), and dentin (c). C)Control group demonstrating that connective tissue fibers (a) in the area coronal to the OC (b) ranparallel to the newly formed cementum (c). Long JE (d) and the root notch (e) are shown. D)Experimental group specimen with parakeratinized epithelium covering the gingiva (a), root notch(b), and alveolar bone (c). E) Higher magnification of the framed area with a shorter JE (a),connective tissue (b), and dentin (c). F) Estradiol-treated specimen demonstrating connective tissuefibers (a) in the area coronal to the OC (b) inserted perpendicular to the newly formed cementum(c). (Toluidine blue stain)

Estradiol Promotes Cementum Formation Volume 81 • Number 7

1070

days of incubation and did not form a mineralized ma-trix. PDL fibroblasts were also not stained with osteocal-cin-specific antibodies. These findings are consistentwith a previous study4 where osteocalcin-specificstaining was used to differentiate cementoblasts fromPDL fibroblasts.

Once the HCDCs were characterized, the effects oflow concentrations of 17b-estradiol were tested in anin vitro proliferation assay. We found that 17b-estra-diol, but not EMD, significantly increased the rate ofhuman cementoblast-like cell proliferation. Similarly,other investigators18 shown a similar proliferationwhen estradiol was tested with human osteoblasts,thus highlighting the close relationship between oste-

oblasts and cementoblasts as mineralized-tis-sue forming cells. This effect of estradiol isconsistent with a previous study24 that reportedthat estrogen deficiency may cause a progres-sion of periodontal disease and alveolar boneloss.

However, our observation that EMD failed toincrease the growth of HCDCs is in contrast witha previous study25 where EMD given at 100 mg/ml for 7 days promoted mouse cementoblastcell proliferation. A possible explanation for thisdiscrepancy might be related to the well-knowndifferences in tooth metabolism between mam-mals and rodents.26 Furthermore, EMD wasshown to promote the proliferation of cells in-volved in the development of the periodontium,including human PDL fibroblasts8 and osteo-blasts.27-29 The results shown in the present in-vestigation are obtained from cells from a singletooth sample from a young human male. Theseresults should be confirmed in different cell linesfrom additional samples.

The use of 17b-estradiol in the treatment ofexperimental 3-wall intrabony periodontal de-fects demonstrated a clear promotion of peri-odontal regeneration as demonstrated by theformation of new cementum and the perpendic-ular arrangement of connective tissue fibers tothe root surface. Incontrast, in thecontrol group,periodontal regeneration was limited, thus con-firming the validity of the experimental lesionused.30 Similar regeneration outcomes were at-tained with other therapeutic approaches in ex-perimentally created defects, including guidedtissue regeneration (GTR),31-33 EMD,34-36 andtissue engineering.37,38

Another goal of periodontal regenerative pro-cedures is to prevent epithelial downgrowth, thusallowing cells from the PDL and bone to repopu-late the root surface and to form a new PDL.39

The local application of 17b-estradiol signifi-cantly prevented this downgrowth when com-

pared to the control (1.77 mm versus 2.45 mm) withresults quantitatively comparable to those reportedin experimental studies using guided tissue regenera-tion with expanded polytetrafluoroethylene mem-branes,31 collagen membranes,32 or with the use ofenamel matrix proteins.40 In terms of new cementumformation, the group treated with 17b-estradiol dem-onstrated a significantly higher dimension of new ce-mentum compared to the control group (2.94 mmversus 1.54 mm), representing 45% and 28% of the de-fects, respectively. Caffesse et al.31 obtained similarpercentages of new cementum (45%) when usingGTR. Similarly, studies using EMD to treat buccal de-hiscence defects in monkeys reported new cementum

Figure 5.Estradiol promotes new bone formation, PDL, and new cellular cementum. A)Experimental group specimen demonstrating new bone formation (a), PDL (b), andnew cellular cementum (c) coronal to the root notch. B) Higher magnification of theframed area, alveolar bone (a), PDL (b), and cellular cementum (c). C) Lowermagnification of experimental group specimen demonstrating new bone formation(a). Arrows point to the limits between new bone and residual alveolar bone (b)and the root notch (c). (Toluidine blue stain; original magnification.).

J Periodontol • July 2010 Nunez, Sanz-Blasco, Vignoletti, et al.

1071

formation amounting to 60% of the defects,34 as well aswhen EMD was applied along with a bone graft to treat2-wall intrabony periodontal defects in beagle dogs.40

In a study41 that used recombinant human bone mor-phogenetic protein-2 (rhBMP-2) to treat intrabonyperiodontal defects in dogs, similar amounts of new ce-mentum formation (3.00 – 1.40 mm) were also re-ported.

In contrast to the results with new cementum andconnective tissueattachment, bone formationoccurredsimilarly in the experimental and control groups, rep-resenting 49% and 48% of the defects, respectively.The experimental model used in this study, the 3-wallperiodontal defect, previously showed its own abilityto stabilize blood clot formation and, hence, to formnew bone. These results are similar to those reportedwhen rhBMP-2 was compared to a buffer controlin the treatment of similar intrabony periodontaldefects.41

In this investigation, we used 17b-estradiol as a bi-ologic mediator to promote periodontal regenerationof intrabony defects once we demonstrated its abilityin vitro to promote human cementoblast cell prolif-eration. Other strategies were reported in the treat-ment of intrabony periodontal defects. For example,Yamamoto et al.40 reported the histologic results ofapplying a combination of a bovine-derived xenograftand EMD in the treatment of 2-wall intrabony defectsin dogs. The amount of new cementum was 4.5 mmand of new bone was 3.6 mm, representing 89.8%and 52.2% of the defects, respectively. These resultsare not directly comparable to ours because of the dif-ferences in the experimental defect and the use ofa bone graft as a carrier of the biologic mediator, aswe used a collagen sponge in the present study. Ex-periments in adult baboon species (Papio ursinus)38

showed the capacity of a tissue-engineered approachby combining the application of 100 and 500 mg of re-combinant human osteogenic protein-1 and xeno-genic bovine collagenous bone matrix to regeneratead integrum the periodontal tissues in mandibularClass II furcation defects. In addition, the combinationof recombinant human platelet-derived growth factor-BB with a bone graft as a carrier demonstrated peri-odontal regenerative outcomes.42 Further researchis needed with the use of 17b-estradiol either aloneor in combination with appropriate scaffold therapiesin more challenging experimental defects.

CONCLUSIONS

The method developed for culturing human cemento-blast-like cells enabled us to demonstrate that lowconcentrations of 17b-estradiol promoted in vitro hu-man cementoblast cell proliferation. Furthermore,the in vivo investigation demonstrated that the localapplication of 17b-estradiol in experimentally cre-

ated periodontal defects in a dog model promotednew cementum formation and the attachment offunctionally oriented connective tissue fibers. Al-though the number of dogs used in this study waslow for ethical reasons, the results clearly suggest thatlocally applied estradiol may promote periodontal re-generation. This consideration has to be taken withcaution because the application of estradiol might in-duce secondary effects on reproductive tissues, suchas the possibility of inducing or maintaining breast oruterine cancer. However, the dosages of estradiolused in this experiment were very low. In fact, in a pre-vious pilot study in three beagle dogs, we observedthat plasma 17b-estradiol concentrations measuredbefore and 5 and 20 days after the application of0.450 mg 17b-estradiol were similar (12.40 pg/mlversus 12.70 pg/ml after 5 days and 12.40 pg/ml af-ter 35 days (unpublished data). Nevertheless, the useof estrogen as a regenerative agent in the treatment ofperiodontal intrabony defects is still premature, andfurther investigations are required before consideringits use in humans.

ACKNOWLEDGMENTS

The authors thank Carmen Roman, a technician at theInstitute of Molecular Biology and Genetics, for tech-nical assistance. This study was supported by grantsfrom the Confederation of Savings Banks, Castilla yLeon, Spain, the Regional Government of Castilla yLeon, Valladolid, Spain (CSI12A08), the Health Insti-tute Carlos III and European Regional DevelopmentFund (FEDER), Madrid, Spain (PI07/0766), and theSpanish Ministry of Science and Innovation, Madrid,Spain (BFU2009-08967). The authors report no con-flicts of interest related to this study.

REFERENCES1. Loe H, Waerhaug J. Experimental replantation of teeth

in dogs and monkeys. Arch Oral Biol 1961;3:176-184.2. Lindskog S, Blomlof L. Mineralized tissue-formation in

periodontal wound healing. J Clin Periodontol 1992;19:741-748.

3. Grzesik WJ, Narayanan AS. Cementum and periodon-tal wound healing and regeneration. Crit Rev Oral BiolMed 2002;13:474-484.

4. D’Errico JA, Ouyang H, Berry JE, et al. Immortalizedcementoblasts and periodontal ligaments cells inculture. Bone 1999;25:39-47.

5. Davidson D, McCulloch CA. Proliferative behavior ofperiodontal ligament cell populations. J PeriodontalRes 1986;21:414-428.

6. Hammarstrom L, Heijl L, Gestrelius S. Periodontalregeneration in buccal dehiscence model in monkeysafter application of enamel matrix proteins. J ClinPeriodontol 1997;24:669-677.

7. Slavkin HC, Mino W, Bringas P Jr. The biosynthesis andsecretion of precursor enamel protein by ameloblasts

Estradiol Promotes Cementum Formation Volume 81 • Number 7

1072

as visualized by autoradiography after tryptophanadministration. Anat Rec 1976;185:289-312.

8. Gestrelius S, Andersson C, Johansson AC, et al.Formulation of enamel matrix derivative for surfacecoating. Kinetics and cell colonization. J Clin Peri-odontol 1997;24:678-684.

9. Cattaneo V, Rota C, Silvestri M, et al. Effect of enamelmatrix derivate on human periodontal fibroblast:Proliferation, morphology and root surface coloniza-tion. An in vitro study. J Periodontal Res 2003;38:568-574.

10. Heijl L. Periodontal regeneration with enamel matrixderivate in one human experimental defect. A casereport. J Clin Periodontol 1997;24:693-696.

11. Lindhe J, Sonesson B. Effect of sex hormones oninflammation. J Periodontal Res 1967;2:7-12.

12. Lindhe J, Attstrom R, Bjorn AL. Influence of sexhormones on gingival exudation in dogs with chronicgingivitis. J Periodontal Res 1968;3:279-283.

13. Norderyd OM, Grossi SG, Machtei EE, et al. Periodon-tal status of women taking postmenopausal estrogensupplementation. J Periodontol 1993;64:957-962.

14. Provenzano A, Carr MP, Horton JE. The effect of themenstrual cycle on BOP in females. J Dent Res 2002;81(Spec. Issue A):2962.

15. Vittek J, Hernandez MR, Wenk EJ, Rappaport SC,Southren AL. Specific estrogen receptors in humangingival. J Clin Endocrinol Metab 1982;54:608-662.

16. Komm BS, Terpening CM, Benz DJ, et al. Estrogenbinding, receptor mRNA, and biologic response inosteoblast-like osteosarcoma cells. Science 1988;241:81-84.

17. Oursler MJ, Osdoby P, Pyfferoen J, Riggs BL, SpelsbergTC. Avian osteoclasts as estrogen target cells. Proc NatlAcad Sci USA 1991;88:6613-6617.

18. DiSilvio L, Jameson J, Gamie Z, Giannoudis PV,Tsiridis E. In vitro evaluation of the direct effect ofestradiol on human osteoblasts (HBO) and humanmesenchymal stem cells (h-MSCs). Injury 2006;37(Suppl. 3):S33-S42 (erratum 2007; 38:1224).

19. Morishita M, Yamamura T, Shimazu A, Bachch AH,Iwamoto Y. Estradiol enhances the production ofmineralized nodules by human periodontal ligamentcells. J Clin Periodontol 1999;26:748-751.

20. Karnovsky MJ. A formaldehyde-glutaraldehyde fixa-tive of high osmolarity for use in electrom microscopy.J Cell Biol 1965;27:137A-138A.

21. Donath K, Breuner G. A method for the study ofundecalcified bones and teeth with attached softtissue. The Sage-Schliff (sawing and grinding) tech-niques. J Oral Pathol 1982;11:318-326.

22. Jeno L, Geza L. A simple differential staining methodfor semi-thin sections of ossifying cartilage and bonetissues embedded in epoxy resin. Mikroskopie 1975;31:1-4.

23. Grzesik WJ, Kuzentsov SA, Uzawa K, Mankani M,Gehron-Robey P, Yamauchi M. Normal human ce-mentum-derived cells: Isolation, clonal expansion, andin vitro and in vivo characterization. J Bone Miner Res1998;13:1547-1554.

24. Payne JB, Zach NR, Reinhardt RA, Nummikoski PV,Patil K. The association between estrogen status andalveolar bone density changes in postmenopausalwomen with a history of periodontitis. J Periodontol1997;68:24-31.

25. Tokiyasu Y, Takata T, Saygin E, Somerman M.Enamel factors regulate expression of genes associ-

ated with cementoblasts. J Periodontol 2000;71:1829-1839.

26. Jordan HV. Rodent model systems in periodontaldisease research. J Dent Res 1971;50:236-242.

27. Schwartz Z, Carnes DL Jr., Pulliam R, et al. Porcinefetal enamel matrix derivate stimulates proliferationbut not differentation of pre-osteoblastic 2T9 cells,inhibits proliferation and stimulates differentation ofosteoblast-like MG63 cells, and increases proliferationand differentation of normal human osteoblast NHOstcells. J Periodontol 2000;71:1287-1296.

28. Galli C, Macaluso GM, Guizzardis S, Vescovini K,Passeri M, Passeri G. Osteoprotegerin and receptoractivator of nuclear factor-kappa b ligand modulationby enamel matrix derivated in human alveolar osteo-blast. J Periodontol 2006;77:1223-1228.

29. Bosshardt DD. Biological mediators and periodontalregeneration: A review of enamel matrix proteins atthe cellular and molecular levels. J Clin Periodontol2008;35:87-105.

30. Kim CS, Choi SH, Chai JK, et al. Periodontal repair insurgically created intrabony defects in dogs: Influenceof the number of bone walls on healing response. JPeriodontol 2004;75:229-235.

31. Caffesse RG, Smith BA, Castelli WA, Nasjleti CE. Newattachment achieved by guided tissue regeneration inbeagle dogs. J Periodontol 1988;59:589-594.

32. Pitaru S, Tal H, Soldinger M, Noff M. Collagenmembranes prevent apical migration of epitheliumand support new connective tissue attachment duringperiodontal wound healing in dogs. J Periodontal Res1989;24:247-253.

33. Gottlow J, Karring T, Nyman S. Guided tissue re-generation following treatment of recession-typedefects in the monkey. J Periodontol 1990;61:680-685.

34. Heijl L, Heden G, Swardstrom G, Ostgrem A. Enamelmatrix derivative (EMDOGAIN) in the treatment ofintrabony periodontal defects. J Clin Periodontol 1997;24:705-714.

35. Pontoriero R, Wennstrom J, Lindhe L. The use ofbarrier membrane and enamel matrix proteins in thetreatment of angular defects. A prospective con-trolled clinical study. J Clin Periodontol 1999;26:833-840.

36. Sculean A, Windish P, Chiantella GC, Donos N, BrecxM, Reich E. Treatment of intrabony defects withenamel matrix proteins and guide tissue regeneration.A prospective controlled clinical study. J Clin Peri-odontol 2001;28:397-403.

37. Nevins M, Camelo M, Nevins ML, Schenk RK, LynchSE. Periodontal regeneration in human using recombi-nant human platelet-derived growth factor-BB (rhPDGF-BB) and allogenic bone. J Periodontol 2003;74:1282-1292.

38. Ripamonti U, Petit JC, Teare J. Cementogenesis andthe induction of periodontal tissue regeneration bythe osteogenic proteins of the transforming growthfactor-b superfamily. J Periodontal Res 2009;44:141-152.

39. Cortellini P, Tonetti MS. Focus on intrabony defects:Guided tissue regeneration. Periodontol 2000 2000;22:104-132.

40. Yamamoto S, Masuda H, Shibukawa Y, YamadaS. Combination of bovine-derived xenografts andenamel matrix derivative in the treatment of intrabony

J Periodontol • July 2010 Nunez, Sanz-Blasco, Vignoletti, et al.

1073

periodontal defects in dogs. Int J Periodontics Re-storative Dent 2007;27:471-479.

41. Choi SH, Kim CK, Cho KS, et al. Effect of recombinanthuman bone morphogenetic protein-2/absorbablecollagen sponge (rhBMP-2/ACS) on healing in 3-wallintrabony defects in dogs. J Periodontol 2002;73:63-72.

42. Ridgway HK, Melloning JT, Cochran DL. Humanhistologic and clinical evaluation of recombinat humanplatelet-derived growth factor and beta-tricalciumphosphate for treatment of periodontal intraosseous

defects. Int J Periodontics Restorative Dent 2008;28:171-179.

Correspondence: Dr. Javier Nunez, Institute of MolecularBiology and Genetics (IBGM), University of Valladolid andSpanish Research Council (CSIC), c/ Sanz y Fores 3,47003 Valladolid, Spain. Fax: 34-983-184800; e-mail:javilds@hotmail.com.

Submitted December 3, 2009; accepted for publicationFebruary 16, 2010.

Estradiol Promotes Cementum Formation Volume 81 • Number 7

1074