role of bone marrow-derived progenitor cells in the maintenance and regeneration of dental...

Role of Bone Marrow-DerivedProgenitor Cells in theMaintenance and Regenerationof Dental Mesenchymal TissuesJING ZHOU,1,2 SONGTAO SHI,3 YUANYUAN SHI,4 HAN XIE,2,5 LEI CHEN,2,6 YONG HE,1,2

WEIHUA GUO,2 LINGYING WEN,1* AND YAN JIN2*1Department of Pedodontics, School of Stomatology, Fourth Military Medical University, Xi’an, Shaanxi, People’s Republic of China2Department of Oral Histology and Pathology, School of Stomatology, Fourth Military Medical University, Xi’an, Shaanxi,

People’s Republic of China3Center for Craniofacial Molecular Biology, University of Southern California School of Dentistry, Los Angeles, California4Department of Ophthalmology, Eye Institute of Chinese PLA, Xijing Hospital, Fourth Military Medical University, Xi’an,

Shaanxi, People’s Republic of China5Department of Orthodontics, School of Stomatology, Fourth Military Medical University, Xi’an, Shaanxi, People’s Republic of China6Department of Periodontology, School of Stomatology, Tongji University, Shanghai, People’s Republic of China

While dental mesenchymal stem cells are well-studied, the origin of these cells is still unclear. Bone marrow-derived cells (BMDCs) havethe potential to engraft into several tissues after injury, but whether they can become dental tissue-specific progenitor cells under normalconditions and the relationship of these cells to the tissue-resident cells are unknown. Thus, we transplanted green fluorescent protein(GFP)-labeled BMDCs into irradiated wild-type mice. We found that the engraftment of BMDCs participated in the regeneration anddifferentiated into periodontal specific cells after injury. Under normal conditions, there were more BMDCs engrafting into the dentalmesenchymal tissue than other organs, in which the expression of stromal cell-derived factor-1 (SDF-1) was significantly higher than inother organs, and the engraftment of cells increased with time. A small fraction of GFPþ cells maintained the mesenchymal stem cellphenotype positive for CD105, CD106, and CD90, which were significantly less than the tissue-resident stem cells; meanwhile,GFPþ/CD45þ cells were rare. Isolation and characterization of the dental pulp cells showed that the number of GFPþ/Ki67þ cells weregreater than theGFP�/Ki67þ cells. In addition, someGFPþ cells differentiated into the dental-specific cells and expressed dental-specificproteins, and can be found in the odontoblast layer after implantation of the apical bud. In conclusion, these data suggest that bonemarrowprogenitor cells communicate with dental tissues and become tissue-specific mesenchymal progenitor cells to maintain tissuehomeostasis.J. Cell. Physiol. 226: 2081–2090, 2011. � 2010 Wiley-Liss, Inc.

Bone marrow-derived cells (BMDCs) are capable of homing tomany tissues in response to injury signals and differentiating intotissue-specific cells, including muscle, liver, brain, pancreasesislet, and even epithelial lineages, suggesting high developmentalplasticity (Lagasse et al., 2000; Krause et al., 2001; Badiavas et al.,2003; Hess et al., 2003; Mezey et al., 2003; Li et al., 2008).However, the potential of the BMDCs to contribute to thedental mesenchymal cells is still unknown. In addition, animaland clinical studies have suggested that BMDCs may not belimited to direct replacement of the damaged cells, they alsoguide the regeneration bymultiple effects on the tissue-residentcells, including enhancement of vascularization and productionof growth factors (Hess et al., 2003; Chamberlain et al., 2007).However, the relationships between the BMDCs and tissue-resident cells and their functional role in the physiologicalcondition have not been well-studied. Meanwhile, stromal-derived factor-1 (SDF-1) and its unique receptor CXCR-4 areknown to be involved in regulation of the BMDCs in theengraftment into the damaged tissues, and they are alsoinvolved in several physiological process during embryo/organogenesis, including hematopoiesis, vascular development,and tooth development (McGrath et al., 1999; Askari et al.,2003; Ratajczak et al., 2006; Son et al., 2006; Laird et al., 2008).Additionally, recent studies indicated that SDF-1 and CXCR-4are expressed in the inflamed dental mesenchymal tissue

(Havens et al., 2008; Jiang et al., 2008). However, whether theSDF-1/CXCR-4 pathway is involved in the engraftment ofBMDC in the dental mesenchymal tissue is unknown.

BMDCs also have the ability to engraft into healthy tissue(Chamberlain et al., 2007). Although such an event occurs atlow frequency, BMDCs are considered as an important sourceof procurable adult stem cells (Alvarez-Dolado et al., 2003;Borue et al., 2004). It is believed that the dental mesenchymalstem cells may originate from the cranial neural crest

Contract grant sponsor: Nature Science Foundation of China;Contract grant numbers: 30725042, 31030033.Contract grant sponsor: National Basic Research Program (973Program);Contract grant numbers: 2010CB944800, 2011CB964700.

*Correspondence to: LingyingWen and Yan Jin, 145West ChangleRoad, Xi’an, Shaanxi 710032, People’s Republic of China.E-mail: [email protected], [email protected]

Received 28 May 2010; Accepted 5 November 2010

Published online in Wiley Online Library(wileyonlinelibrary.com), 6 December 2010.DOI: 10.1002/jcp.22538

ORIGINAL RESEARCH ARTICLE 2081J o u r n a l o fJ o u r n a l o f

CellularPhysiologyCellularPhysiology

� 2 0 1 0 W I L E Y - L I S S , I N C .

ectomesenchyme (CNC) (Chai et al., 2000). However,whether non-CNC-derived cells contribute to dentalmesenchymal stem cells is unknown. Bone marrow-derivedmesenchymal stem cells (BMMSCs) and dental pulp stem cells(DPSC), or periodontal ligament stem cells (PDLSCs) sharesimilar gene expression profiles, differentiation abilities, andperivascular niche (Gronthos et al., 2000; Seo et al., 2004;Huang et al., 2008). Our hypothesis is that bone marrow-derived progenitor cells can engraft in dental mesenchymaltissue compartment and convert into dental mesenchymalstem/progenitor cells.

To test the above hypothesis, we used the green fluorescentprotein (GFPþ) bone marrow chimeric mice to identify thecharacteristics of migrating cells in pulp and periodontium.Wealso determined the relationships and differences of themigrating and tissue-resident stem cells. In addition to the pulptissue analysis, DPSCs were also isolated and compared withBMMSCs. We found that BMDCs can migrate into normal aswell as damaged dental tissues and differentiate into the dentaltissue-specific cell type, while only a small fraction of migratingBMDCs maintain the stem cell phenotype.

Materials and MethodsBone marrow transplantation

Ten 4-week-old female GFPþ/þ transgenic mice (C57BL/6background, Institute of Neurosciences, The Fourth MilitaryMedical University) and 50 age-matched C57 wild-type mice(Institute of Neurosciences, The Fourth Military MedicalUniversity) were used as the donors and recipients, respectively,which were maintained in laminar flow cages in pathogen-freeconditions. All experiments were performed following theGuidelines of the Fourth Military Medical University IntramuralAnimal Use and Care Committee. The bone marrow cells wereobtained from the GFPþ/þ mice by flushing the tibia and femurbones with phosphate-buffered saline (PBS) containing 2% fetalbovine serum (FBS; Gibco, Grand Island, NY) and washed afterpassing through a nylon mesh. The erythrocytes were removed bytreatment with 2ml erythrocyte lysis buffer (eBioscience, SanDiego, CA). Then 2� 106 GFPþ bone marrow cells weresuspended into 100ml PBS and injected intravenously by tail veininto wild-type C57 mice within 24 h after 8.5 Gray of whole bodyirradiation. The successful creation of GFP chimera was confirmedby analysis of the GFPþ peripheral blood mononuclear cells(PBMCs) by flow cytometry (FACSCalibur; BD, San Jose, CA). Thepercentage of GFPþ PBMCs in recipients quickly increased andreached 85%by 30days (datawas not shown), and then the chimeramice were used in the study. At 1, 2, and 4 months post-bonemarrow transplantation (BMT), the recipient mice were perfusedtranscardially with 20ml 0.9% saline followed by 50ml 4%paraformaldehyde under euthanasia. The mandibles of therecipient mice were post-fixed with 4% paraformaldehyde for 5 h,and decalcified for 10 days with ethylene diamine tetraacetic acid(EDTA) in 48C. Then the 4mm paraffin or frozen sections wereprepared for further analysis.

Experimental alveolar bone defect

Experimental alveolar bone defect (EABD) was performed aspreviously described (Ohta et al., 2007). In brief, 10 adult femalechimeric mice, each weighing approximately 20 g, wereanesthetized using barbital sodium. The gingival and the palatemucosa in the mesial of the maxillary first molar were flapped toexpose the alveolar bone overlying themesial root of first molar. Adefect was made by a round dental burr cooled by PBS, and filledwith pieces of ceramic bovine bone (CBB) <1mm. Finally, thegingival flaps were repositioned (Fig. 1A). After operations, theexperimental mice were maintained in laminar flow cages inpathogen-free conditions. The animals were killed by 5, 15, and

30 days after the operation. And then the paraffin and frozensections were made as described above.

Cell culture

TheDPSCswere cultured as previously described (Yu et al., 2006).In brief, the dental pulps were extracted from the lower incisors ofthe chimeric mice and were physically separated from the enamelorgans and apical buds, then minced into <1mm3 pieces anddigested in a mixture of 3mg/ml collagenase type I (0.66mg/ml;Sigma, St. Louis, MO) and 4mg/ml dispase (Roche, Mannheim,Germany) for 40–60min at 378C. Subsequently, the digested tissuewas passed through a 70mm cell strainer (Falcon, BD Labware,Franklin Lakes, NJ) to obtain single cell suspensions. The cells wereseeded into 24-well plates (Costar, Cambridge, MA) at 2� 105/mlin modified Eagle’s medium (MEM, Gibco) supplemented with 15%FBS (Gibco), 2mM glutamine (Sigma–Aldrich, St. Louis, MO), 1%penicillin–streptomycin (Biofluids, Rockville, MD), and incubated in5% CO2 at 378C. BMMSCs from the chimeric mice were culturedas previously described (Wieczorek et al., 2003).

Multipotent differentiation of DPSCs

The DPSCs from chimeric mice were analyzed for theirmultipotential differentiation of the GFPþ cells in the dental pulp,such as adipogenesis and neurogenesis. DPSCs were seeded in the24-well plates at 2� 105 cells/well. Adipogenic differentiationmedium was DMEM supplemented with 10% FBS, 0.5mM isobutylmethylxanthine (Sigma), 2mM insulin (Sigma), and 10 nMdexamethasone (Sigma). The medium was changed every 3 daysuntil 4 weeks and assessed by the Oil Red O (Sigma) staining. Forneurogenic induction, theDPSCswere culture in the pre-inductionmedium, including DMEM, 15% FBS, and 1mmol b-mercaptoethanol (Sigma) for 24 h followed by the establishedprotocol (Woodbury et al., 2000) and analyzed by theimmunofluorescent expression of neural stem cell marker nestin.

Immunohistochemistry

For immunohistochemical studies, paraffin sections were treatedwith heated antigen retrieval solution and blocked in 10% normalgoat serum in PBS for 30min at room temperature. Then thesections were incubated with primary rabbit anti-GFP antibody(1:100, AnaSpec, San Jose, CA) at 48C overnight, and HRP-conjugated secondary anti-rabbit antibody (DAKO, Glostrup,Denmark) followed by visualized using 3,3-diaminobenzidinetetrahydrochloride (DAB, DAKO). For immunofluorescentanalysis, frozen sections and cells slides ofDPSCs andBMMCswereincubated in 0.4% triton X-100 for 15min and then blocked in 10%normal goat serum in PBS for 30min at room temperature, thenincubated with primary antibodies, including goat anti-mousealkaline phosphatase (ALP; 1:100, Santa Cruz Biotechnology, SantaCruz, CA), goat anti-mouse CD31 (1:100, Santa CruzBiotechnology), rat anti-mouse CD90 (1:100, eBioscience), rabbitanti-mouse CD105 (1:100, Santa Cruz Biotechnology), rat anti-mouse CD106 (1:100, Chemicon Inc.,Temecula, CA), rat anti-mouse CD45 (1:100, Abcam, Cambridge, MA), rat anti-mouseCD34 (1:100, Abcam), goat anti-mouse CD14 (1:100, Santa CruzBiotechnology), rat anti-mouse Nestin (1:200, Abcam), goat anti-mouse dentin matrix protein-1 (DMP-1; 1:50, Santa CruzBiotechnology), rabbit anti-mouse dentin sialoprotein (DSP; 1:100,Santa Cruz Biotechnology), rabbit anti-mouse Fibronectin (1:100,Thermo Inc., Fremont, CA), rabbit anti-mouse Ki67 (1:100,Abcam), mouse anti-mouse Osteopontin (OPN; 1:100, Santa CruzBiotechnology), rabbit anti-mouse SDF-1 (1:100, Santa CruzBiotechnology), rabbit anti-mousea-smoothmuscle actin (a-SMA;1:100, Santa Cruz Biotechnology) at 48C overnight. PBS instead ofprimary antibodies was used as negative control. The fluorescentsecondary antibodies including rabbit anti-mouse, goat anti-rabbit,rabbit anti-rat, and rabbit anti-mouse conjugated to Rhodamine(1:200, KPL Inc., Gaithersburg, MD) for 40min at room

JOURNAL OF CELLULAR PHYSIOLOGY

2082 Z H O U E T A L .

Fig. 1. BMDCs participate in the regeneration of periodontium. The diagrammatic illustration and image of the EABD. Immunohistochemicalstaining ofGFPat 5 days (B,C), 15 days (D), and 30 days (E) after operation.A:Thediagrammatic illustration and the imageof theEABD.B: In theEABDarea, thecavitywasfilledwithfibrinandCBB.C:Ahighmagnificationofbox inBshowedtheedgeof thealveolarbonedefects, andthearrowindicated theGFPR cells. D: At 15 days, the newly forming alveolar bones were filled with the GFPR osteoblasts. E: Some osteoblasts in the newalveolarboneandafewcells inthebonemarrowcavitywereGFPR,andtherewereafewGFPRcells inthegingiva.F:Thenormalalveolarbone.TheexpressionofGFPwasmore inEABDthannormal periodontal area (G,I,K).Meanwhile, theexpressionofSDF-1 in theEABDwas stronger than innormalperiodontal tissue(H,J,K).Theco-expressionoftheGFPandperiodontalassociatemarker,suchasOPN(L),Ki67(M),Fibronectin(N),ALP(O), andCD106 (P),a-SMA(Q)were in theEABDareaat 30days after operation. MP<0.05.AB, alveolar bone;CBB, ceramicbovinebone;R, root;PDL, periodontal ligament; V, vessel. Scale bar represents 50mm.


B M D C s B E C O M E D E N T A L P R O G E N I T O R C E L L S 2083

temperature. Then the sections were counterstained either withHoechst 33342 (5mg/ml; Sigma) or hematoxylin to identify allnuclei. Then image collection and superimposition were processedby DP controller (Olympuls, Tokyo, Japan) and DP manager.

In vivo transplantation

To investigate the in vivo regeneration capacity of the BMDCs,renal capsule transplantation of apical bud was carried out. Incisorgerms were dissected from the mandibles of chimeric mice4 months post-transplantation. For transplantation of the incisorgerms, the apical end regions were mechanically separated fromthe incisor. The regions were transplanted underneath the renalcapsules of wild-type female mice (C57, 8w), allowed to incubatefor 4 weeks and removed. The tissues were fixed in 4%paraformaldehyde, decalcified in 10% EDTA in PBS, and frozensections were made for further analysis.

Quantification of immunofluorescent staining

Fluorescent Rhodamine staining of SDF-1 and GFP expressions inliver, lung, adipose, and dental tissue were captured on the DPmanager and quantified by using the Photoshop software (AdobeSystems, San Jose, CA). After capturing the images at the samemagnification, the threshold was set and maintained for eachsection in the experiment, and the ratios of the areas of positivecells to the areas of the nuclear regions were calculated by use ofthe color selection function within the software. The values of theratios from the three tissues were given as the mean� SD, and theCD (105, 106, and 90) þ cell numbers were calculated for each ofthe five images in dental pulp or periodontium. Results areexpressed as the mean score� SD. Statistical significance wasdetermined by Student’s t-test. P values of<0.05were judged to bestatistically significant.

ResultsBMDCs participate in periodontal regeneration

Five days after the operation, the cavity was filledwith fibrin andCBB. The GFPþ cells were mainly observed around the CBBand alveolar bone (Fig. 1B). At high magnification, the defects atthe edge of the alveolar bone were observed, where the cellssurrounding the vessels or theCBBwere positively stainedwiththe anti-GFP antibody, the arrow indicated the GFPþ cells(Fig. 1C). Within 15 days of the operation, the newly formedalveolar bone was filled with the GFPþ osteoblasts. The newlyformed bone was mostly woven bone with primary osteocyte(Fig. 1D). At 30 days after the operation, the new alveolar boneand the cells in the bone marrow cavity were positively stainedwith anti-GFP, and gingiva were also positive (Fig. 1E). Whilethere was fewer GFPþ cells in the normal alveolar bone thannew forming bone (Fig. 1F). Meanwhile, we found theexpression of SDF-1 was elevated in the EABD area (Fig. 1H)with the more migration of the GFPþ BMDCs (Fig. 1J)compared with the normal periodontal tissues (Fig. 1G,I,K;P< 0.05). Further identification of the function of BMDCs in theperiodontal regeneration, we analyzed the mineral-associatedmarkers and cell surfacemarkers in the EABDarea 30 days afterthe operation. OPNþ cells were mainly observed around thenewly forming bone and periodontal ligament (PDL), and mostGFPþ cells were also OPN positive (Fig. 1L). The Ki67þ cellswould be found in the periodontium, alveolar bone, and gingival,and theKi67þ/GFPþ cells weremainly around the root surface(Fig. 1M). Fibronectinþ/GFPþ cells almost arranged on theroot surface (Fig. 1N). Therewere several ALPþ/GFP� cells onthe surface of the newly forming bone. The ALPþ/GFPþ cellswere found in the periodontal region and around the bone(Fig. 1O). There also some CD106þ/GFPþ and a-SMAþ/GFPþ in the periodontal area. The CD106þ/GFP� and a-

SMAþ/GFP� can also be observed in the EABD area(Fig. 1P,Q).

Distribution of GFPR BMDC in diverse organs

In order to identify the capacity of BMDCs to engraft intomultiple organs under normal conditions, the frozen sections ofseveral organs from the chimeric mouse were evaluated byfluorescence microscopy. The results indicated that theBMDCs have the ability to migrate into multiple organsincluding dental pulp, periodontium, liver, lung, heart, pancreas,skin, and adipose at 4 months post-BMT (Fig. 2A–H). Moreimportantly, we found that the GFPþ BMDCs homed to thedental tissues to a greater degree than in liver, lung, and adipose(P< 0.05; Fig. 2I). To further dissect the mechanism and toidentify the signal that regulate the stem cell recruitment, wedetected SDF-1 expression in the dental tissues, adipose, lung,and liver (Fig. 2J–M), which were measured by the area of SDF-1þ/Hoechst staining. The SDF-1 expression was significantlyhigher in dental than in other tissues (P< 0.05; Fig. 2N).Mandible tissues from the GFPþ chimeric mouse were stainedfor CD31 to test the assumption that the delivery of BMDCsinto the dental tissue may be through the circulation.We foundthat the majority of GFPþ cells located near and surroundingthe endothelial cells, which were also positive for CD31(Fig. 2O). At high magnification, we observed that the CD31signals were mainly on the membrane of the endothelial cells,and the GFPþ BMDCs were between the endothelial cells(Fig. 2P). Interestingly, we found that the GFPþ cells in the apassage which connects the bonemarrow cavity and the PDL inthe mandible (Fig. 2Q).

Dynamic distribution of GFPR BMDCs in dentalmesenchymal tissues

In order to understand the interaction between the bonemarrow and dental tissue, we serially analyzed the engraftmentof the BMDCs in the dental tissues after BMT. We found thatthe migration of the GFPþ BMDCs into dental pulp (Fig. 3A)and periodontium of the secondary molar (Fig. 3B) wasgradually increasedwith time after BMT.Onemonth after BMT,the GFPþ BMDCs were mainly located in the cell-rich layer ofpulp. This was a rare event that the odontoblast cells stainedpositively with the GFP antibody (Fig. 3C). The cells thengradually appeared in both the cell-free layer and the cell-richlayer at 4 months post-BMT. Moreover, the GFPþ cells couldbe observed, although rarely, in the odontoblast layer adjacentto the dentin, while the GFP signal was not detected in theodontoblast process (Fig. 3D). On the other hand, 1 monthafter BMT, there were a fewGFPþ cells in the PDL and alveolarbone (Fig. 3E), and at 4 months post-BMT, the GFPþ cells weremore widely distributed in the tissues, including the PDL andalveolar bone (Fig. 3F).

Immunohistochemical analysis of the surface marker

To characterize the phenotype of the engrafting cells in thedental tissues in vitro, the mandible sections were stained forCD105, CD90, CD106, and CD45 (Fig. 4A). The red arrowsindicate the GFPþ/CD105þ, GFPþ/CD90þ, and GFPþ/CD106þ cells (migrating mesenchymal progenitor cells).Actually, most of the GFPþ cells were not positive for themesenchymal stem cells’ surface markers in the pulp (the leftpart) and in the periodontium (the middle part). The datasuggest that the BMDCs could engraft as the mesenchymalprogenitor cells in the dental tissues. The yellow arrowsindicate the GFP�/CD105þ, CD90þ, 106þ cells (tissue-resident mesenchymal progenitor cells). There were fewGFPþ/CD45þ cells, indicating the non-hematopoietic origin, inthe pulp or periodontium. The right parts show the positivecontrol in the mandible bone marrow cavities. Of note, there


2084 Z H O U E T A L .

were significantly more GFP�/CD105þ, CD90þ, 106þ cellsthan GFPþ/CD105þ, CD90þ, 106þ cells (P< 0.05; Fig. 4B).

Function of the BMDCs in the dental tissues

Secondary molars from GFPþ chimeric mouse 4 months afterBMT were analyzed for markers such as the odontoblastmarkers (DSP and DMP-1) and osteoblast/cementoblastmarkers (OPN, ALP, and fibronectin). The DSP-positive cellswere found in the odontoblast layer and dentin, suggesting thedonor GFPþ cells could differentiate into the functional dentalpulp cells (Fig. 5A). The expression ofDMP-1was similar to thatof DSP, and the DMP-1þ/GPPþ cells would be observed in theodontoblast layer (Fig. 5B). Therewas rare regenerative dentin,indicating the impact of radiation on the odontoblasts was notsevere.OPNþ cells were observed in the pulp, gingiva, and PDL(Fig. 5C). There were a small number of OPNþ/GFPþ cells inthe pulp, gingival, and PDL. In the periodontal area, ALPþ/

GFPþ cells were observed along the alveolar bone and the rootsurface, indicating the BMDCs near the alveolar bone may haveconverted into the PDL cells or osteoblasts (Fig. 5D). Thefibronectinþ/GFPþ cells were only observed in this PDL(Fig. 5E). Most of the Ki67 positive cells were also GFPþ,suggesting the BMDCs have the ability to proliferate in theperiodontal tissues (Fig. 5F). We also transplanted the apicalbud from the lower incisor of the chimeric mouse into the renalcapsule of the wild-type mice, in which the GFPþ cells werescattered in the pulp (Fig. 5G). The implanted apical budsdeveloped into osteo-dentin-like tissue, as indicated by GFPþcells lining the inside of the newly formed dentin-like structures(Fig. 5H).

Isolation and characterization of the BMDCs in pulp

To further identify the characterizations of the dental pulp cells,we isolated and cultured DPSCs at low seeding density. At 4

Fig. 2. Distribution of the BMDCs in multiple organs. Distribution of bone marrow-derived GFPR cells in the liver (A), lung (B), heart (C),pancreas (D), skin (E), adipose (F) pulp (G), andperiodontium(H) 4months followingBMT. I: ThereweremoremigrationofBMDCs in thedentaltissuesthanliver, lung,andadipose(MP<0.05).SDF-1(red)expressedinthedental tissues(J),adipose(K), lung(L),and liver(M).(N)Theexpressionof SDF-1 in dental tissueswas significanthigher than in adipose, lung, and liver (MP<0.05).O:TheGFPRBMDCs locatedaround the vessels,wheretheendothelialcellswerepositivefortheCD31(red).P:HighmagnificationoftheboxintheO.Q:TheGFPRcells’distributionattherootfurcation.Therewas a passage (arrow) connecting thebonemarrowcavity andPDL.AB, alveolar bone;C, bonemarrowcavity; PDL, periodontal ligament;V, vessel. Scale bar represents 50mm.



days, DPSCs formed small colonies, and at 10 days the coloniesgrew large, and 15 days the cells showed typical spindle orpolygonal shaped cells (Fig. 6A–C). The colonies with moreGFPþ cells expressed stronger Ki67 signals than the colonieswithout GFP signals (Fig. 6D). Further investigation of themesenchymal and hematogenic phenotype of themigrating cellsin dental pulp and comparisonwith BMMSCs showed that someof the GFPþ BMDCs did not stain positively for CD105,CD106, and CD90 (Fig. 6E); however, most of the GFPþBMMSCs were positive for those markers. While both GFPþBMDCs and BMMSCs were rarely positive for the CD45,CD34, and CD14.

The multipotential capacity of the GFPþ BMDCs in thedental pulp was determined. Under adipogenic induction for 4

weeks, The GFPþ DPSCs could differentiate into theadipocytes producing lipid droplets [Fig. 6F(b,c)], while theuntreated cells could not differentiation [Fig. 6F(a)]. To assessthe neurogenic potential, the DPSCs were cultured in theinduction medium. Morphologic changes in the cells as well asthe expression of the nestin were observed [Fig. 6F(e,f)], andthe GFPþ/nestinþ cells were also be detected [Fig. 6F(d)].

Discussion

Our results indicate that the BMDCs could engraft as theperiodontal relevant cells and are involved in regenerationof periodontium. After the EABD operation, the BMDCsengrafted into periodontal tissues resulted in significantly

Fig. 3. Distribution of BMDCs in the dental tissues. Distribution of GFPR BMDCs in the pulp (A) and periodontium (B) at 1, 2, and 4 monthsfollowing BMT. From1 to 4months after BMT, the bonemarrow-derivedGFPR cellsmigrating into the dental pulp and periodontal tissues bothincreased.C:OnemonthafterBMT, theGFPRcellsweremainlyobserved in thecell-rich layerof thepulp.This isa rareevent that theodontoblastcellswerepositive forGFP.D:TheGFPsignalswere found inallpulpcell layers,even intheodontoblastat4monthsafterBMT.ThearrowindicatestheGFPRcells intheodontoblast layer.E:At1monthafterBMT,GFPwasmainlyexpressed inthefibroblast.F:ThereweremoreGFPRcellsof theperiodontal tissue at 4months.AB, alveolar bone;C, cementum;D, dentin; P, pulp;O, odontoblast; PDL, periodontal ligament; R, root. Scale barrepresents 50mm.


2086 Z H O U E T A L .

higher uptake in the EABD area than in the normal periodontaltissues. The increased engraftment levels were related with theup-regulation of SDF-1 in the EABD, suggesting that theengraftment in the periodontal tissue may be regulated by theSDF-1 signal. Furthermore, the BMDCs mainly differentiatedinto the tissue relevant cells, including PDL fibroblasts andosteoblasts, with a small fraction of the BMDCs converting intotissue-specific mesenchymal progenitor cells, suggesting thedonor and recipient progenitor cells may both regulated theregeneration of the periodontium. We hypothesize that in theEABD environment, the migrating of BMDCs may be regulated

by the SDF-1 signaling. Once in the area, the precursor cells areexposed to the periodontium microenvironment, whichnormally promotes their proliferation, and differentiation intotissue-specific cells. Additionally, both the donor and recipientsprogenitor/stem cells may play crucial roles in the initiation ofthe regenerative process.

Dental mesenchymal stem/progenitor cells play a significantrole inmaintaining homeostasis of dental tissues.We found thatBMDCs could migrate into dental pulp and periodontal tissues,and there was more engraftment in the dental tissue than inother organs, such as lung, liver, and adipose. The phenomenonthat higher engraftment of BMDC in dental tissuemay be due tothree reasons. First, the continuous remolding of theperiodontal tissue throughout life needs supplementarysupport from systemic progenitor/stem cells. Second, we foundthat therewas a connective pathway between the bonemarrowcavity and periodontium in the mandible, through which theGFPþ BMDCsmay directly pass without migrating through theendothelial cells. It was previous shown that abundant bloodflow may contribute to the engraftment of BMDC in themaxillofacial region (Uehara et al., 2001; Echlin and McKeag,2004). Additionally, with systemic administration of avasodilator, the homing of labeled mesenchymal stem cells tothe marrow of long bones was significantly increased, furtherindicating that the blood flow may influence the engraftment(Gao et al., 2001). Third, although the mechanisms that governthe recruitment and homing of BMDCs to the injury site is stillunknown, recent work showed that BMDCs are capable ofmigrating into many tissues in response to the injury signals andeven in the healthy animal by the SDF-1/CXCR4 pathway(Ceradini et al., 2004).We found that the expressions of SDF-1protein were significantly higher in dental mesenchymal tissuethan in other organs, suggesting the high attraction of BMDCsto dental tissues, which may be attributed to the higherengraftment levels. On all accounts, the special environment ofthe dental tissue makes high attraction for the BMDCs.

Another phenomenon is that the increase of the BMDCs indental tissue post-BMT from 1 to 4 months. In the traditionalthought, the dental pulp is a relatively static structure and is onlyactive in response to the dental matrix component orinflammatory signals (Shi andGronthos, 2003). In this study, weidentified that the persistent migration of BMDCs into the pulpwith the time increase, which imply that there may be an activecontact between dental pulp and bone marrow. Thus therelationship between them makes the dental pulp connectedthe whole body by the bone marrow. However, whether thiscontact is bidirectional or unidirectional is still unknown andneeds further researches. Besides, periodontal remodeling is acommon phenomenon which requires supplemental stem/progenitor cells (Grachtchouk et al., 2006). A persistentsupplemental cell source from the bone marrow may meet theneeds of the dental periodontal remodeling process.

We then determined the phenotype of the engraftingBMDCs by analyzing the surface markers in the dental tissue,and found that some BMDCs were positive for themesenchymal stem cell markers and rarely positive for thehematogenic stem cell markers. Although, a number of studieshave suggested that the whole BMDCs or hematopoietic stemcells (HSC) may also transdifferentiate into a variety of non-hematopoietic lineages under BMT (Brazelton et al., 2000;Jackson et al., 2001; Hashimoto et al., 2004;Ott et al., 2005; Huiet al., 2006), recent researches have concluded that purifiedHSCs do not have the potency to transdifferentiate into themesenchymal lineage (Koide et al., 2007). Our results indicatedthat the CD45þ, CD34, and CD14 cells were rare in the pulpand periodontal tissues, which confirmed that thehematopoietic cells have no potency to differentiate into dentalmesenchymal cells. Thus, we concluded that the dental tissuemicroenvironments may educate the BMDCs to convert into

Fig. 4. Phenotype of the BMDCs in the dental tissue. Dental tissuesof the chimeric mouse were stained for CD105, CD90, CD106, andCD45 (red) and counterstained for Hoechst (blue). The left partshows the pulp, the middle part is the periodontium, and the rightpart is the positive control. A: The red arrows show the GFPR/CD105R, GFPR/CD90R, and GFPR/CD106R cells. The yellowarrows show theGFP�/CD (105, 90, and 106)R cells. There were fewGFPR/CD45R cells in this area. The cell numbers of the GFP�/CDsRwas significantly higher than theGFPR/CDsR (B). MP<0.05. Scale barrepresents 50mm.



dental-specific progenitor cell type. Furthermore, we alsofound that the majority of cells positive for the mesenchymalsurface markers in dental pulp and periodontium are GFP�,which indicate that the resident stem cells play a predominantrole in the dental tissue. However, further isolation andcharacterization of the dental pulp cells in vitro showed that theexpression of Ki67 in migrating BMDCs was more than in thetissue-resident cells, which indicated that BMDCs could beamplified and have greater proliferative potential than residentcells. These evidences imply that the individual DPSC andPDLSC colonies with remarkable proliferative rates in vitromay have different origins (Shi andGronthos, 2003). In contrast,another study demonstrated that the majority of theproliferating cells in the ductal and islet regions are recipientcells of pancreatic origin under streptozotocin-inducedpancreatic damage (Hess et al., 2003). This may be due to thedifferent proliferative abilities of the different tissues. Inaddition, compared with iliac BMMCSs, maxilla BMMSCs canproliferate more rapidly, which may influence the proliferativeability of the migrating BMDCs in the dental tissues (Akintoyeet al., 2006). Based on these observations, we hypothesized that

the microenvironment of the dental pulp and periodontium cansupport viability, differentiation, and proliferation of BMDCs.

In order to test the functional role of BMDCs on the dentalmesenchymal cells, we co-stained them with antibodies againstdental tissue-specific markers and GFP, and assessed theregeneration ability of these cells. BMDCswould be detected inthe dental pulp even in the odontoblast layer with expression ofDSP and DMP-1. To further confirm that the BMDCs couldtransform into the dental mesenchymal stem cells, wedetermined the multipotential differentiation of these cells, andfound that BMDCs could undergo the adipogenesis andneurogenesis, which may imply that the BMDCs may convertinto the DPSCs and served as the supplementary pool for theregeneration of the dental pulp. However, therewas no signal inthe odontoblast process, a special structure embedded in thedentin tubular. Damaged odontoblasts can be replaced bythe newly formed odontoblasts derived from the stem cellsin the pulp. For example, irradiation can cause retraction ofthe odontoblast process at the dentinoenamal junction andreduction of the number of the vital odontoblasts in the pulp(Goracci et al., 1999; Al-Nawas et al., 2004). Therefore, the

Fig. 5. FunctionoftheBMDCsinthedental tissues.Secondarymolars fromGFPRchimericmouse4monthsafterBMTwerestained forDSP(A),DMP-1 (B), OPN (C) ALP (D), fibronectin (E), Ki67 (F) (red) and counterstained for Hoechst (blue). A: The DSP-positive cells were found in theodontoblast layer and dentin. There were a few DSPR/GFPR cells. B: The expression of DMP-1 was similar to that of DSP. C: There were a fewGFPR/OPNR cells in the pulp, periodontium, and gingival. D: ALPR/GFPR cells were observed along the alveolar bone and the root surface.E:TherewereafewfibronectinR/GFPRcells inthisfield.F:MostofKi67positivecellswerealsoGFPpositive.G:TheGFPRcellswere located inthelower incisorof theGFPRchimericmouse.H:The implantedapicalbuddeveloped intoosteo-dentin-liketissues.TherewereGFPRcells insidethenewly formingdentin-like layer.A, alveolar bone;AB, apical bud;D,dentin;G, gingiva;O,odontoblast;OD,osteodentin;P, pulp;PDL,periodontalligament; R, root. Scale bar represents 50mm.


2088 Z H O U E T A L .

irradiation procedure before the BMT may influence theodontoblasts, while the migrating cells can replace them. Inaddition, it was reported that BMDCs could fuse spontaneouslywith neural, hepatocyte, and cardiac muscle cells, resulting information of multinucleated cells (Alvarez-Dolado et al., 2003).Thus, such cell fusion events may not be avoided in theodontoblast. Meanwhile, after implantation, the apical budformed osteo-dentin structures with the GFPþ cells in theodontoblast layers. This indicates that BMDCs may participatein regeneration of the dentin, served as the source of stem cellsfor the replacement of odontoblasts.

Conclusions

In this work, we first reported that the BMDCs preferentiallymigrate into the pulp and periodontium over other organs.Most of the migrating cells differentiated in both healthy andinjury conditions with a small fraction of the BMDCs becomingthe dental tissue-specific mesenchymal progenitor/stem cellsinvolved in the maintenance and regeneration of the dentalmesenchymal tissue. Furthermore, although the migrating stemcells were fewer in number compared with the tissue-residentstem cells, they have greater proliferative potential. All these

results indicate that the BMDCs serve as the pool of dentalmesenchymal progenitor/stem cells.We also observed that thebone marrow and dental tissues have a persistent connectivepathway for cellular communications. Because purifying theBMMSC in mice is technically challenging, we will need tofurther developmethods for investigating the role of BMMSC inthe dental tissues.

Acknowledgments

This work was supported by grants from the Nature ScienceFoundation of China (grant nos: 30725042 and 31030033) andthe National Basic Research Program (973 Program; grant nos:2010CB944800 and 2011CB964700).

Disclosures

The authors declare no potential conflicts of interests.

Literature Cited

Akintoye SO, Lam T, Shi S, Brahim J, Collins MT, Robey PG. 2006. Skeletal site-specificcharacterization of orofacial and iliac crest human bone marrow stromal cells in sameindividuals. Bone 38:758–768.

Fig. 6. Characterization of BMDCs in the pulp. A: At 4 days, in primary culture, DPSCs formed small colonies. B: At 10 days, the colonies grewlarge, (C)at15days, theDPSCsshowspindleorpolygonal shapedcells.D:ThecolonieswithmoreGFPRcells expressedstrongerKi67signals thanthe colonieswithout theGFP signals. E: TheGFPpositive cellswere located in theDPSCculture, and theCD105;CD106;CD90expressions in theDPSCwere detected. There was only a small fraction of GFPR/CD105R, CD106R, and CD90R cells in the culture. However, most of the GFPRBMMSCswerealsoCDs (CD105,CD106, andCD90)positive.TherewererareCD45,CD34, andCD14positivecells in theDPSCsandBMMSCs.F:Multipotential capacity of the GFPR BMDCs in the dental pulp was analyzed by the adipogenic and neurogenic differentiation. (a–c) Oil Red Ostaining of the DPSCs under untreated culture (a) and adipogenic induction (b,c). The red arrow showed the GFPR BMDCs differentiated intoadipocytes. Neurogenic differentiation was evidenced by themorphology (e), expression of nestin (f), while the untreated cells did not change inmorphology (d). The yellow arrow showed the GFPR BMDCs differentiated into the neuron expressing the nestin. Scale bar represents 50mm.



Al-Nawas B, Duschner H, Grotz KA. 2004. Early cellular alterations in bone after radiationtherapy and its relation to osteoradionecrosis. J Oral Maxillofac Surg 62:1045.

Alvarez-DoladoM, Pardal R, Garcia-Verdugo JM, Fike JR, LeeHO, Pfeffer K, Lois C,MorrisonSJ, Alvarez-Buylla A. 2003. Fusion of bone-marrow-derived cells with Purkinje neurons,cardiomyocytes and hepatocytes. Nature 425:968–9973.

Askari AT, Unzek S, Popovic ZB, Goldman CK, Forudi F, Kiedrowski M, Rovner A, Ellis SG,Thomas JD, DiCorleto PE, Topol EJ, Penn MS. 2003. Effect of stromal-cell-derived factor 1on stem-cell homing and tissue regeneration in ischaemic cardiomyopathy. Lancet362:697–703.

Badiavas EV, Abedi M, Butmarc J, Falanga V, Quesenberry P. 2003. Participation of bonemarrow derived cells in cutaneous wound healing. J Cell Physiol 196:245–250.

BorueX, Lee S, Grove J, Herzog EL, Harris R, DifloT, Glusac E, HymanK, TheiseND, KrauseDS. 2004. Bone marrow-derived cells contribute to epithelial engraftment during woundhealing. Am J Pathol 165:1767–1772.

Brazelton TR, Rossi FM, Keshet GI, Blau HM. 2000. From marrow to brain: Expression ofneuronal phenotypes in adult mice. Science 290:1775–1779.

Ceradini DJ, Kulkarni AR, Callaghan MJ, Tepper OM, Bastidas N, Kleinman ME, Capla JM,Galiano RD, Levine JP, Gurtner GC. 2004. Progenitor cell trafficking is regulated byhypoxic gradients through HIF-1 induction of SDF-1. Nat Med 10:858–864.

Chai Y, Jiang X, Ito Y, Bringas P, Jr., Han J, Rowitch DH, Soriano P, McMahon AP, Sucov HM.2000. Fate of the mammalian cranial neural crest during tooth and mandibularmorphogenesis. Development 127:1671–1679.

Chamberlain G, Fox J, Ashton B, Middleton J. 2007. Concise review:Mesenchymal stem cells:Their phenotype, differentiation capacity, immunological features, and potential forhoming. Stem Cells 25:2739–2749.

Echlin P, McKeag DB. 2004. Maxillofacial injuries in sport. Curr Sports Med Rep 3:25–32.Gao J, Dennis JE, Muzic RF, Lundberg M, Caplan AI. 2001. The dynamic in vivo distribution ofbone marrow-derived mesenchymal stem cells after infusion. Cells Tissues Organs169:12–20.

Goracci G, Mori G, Baldi M. 1999. Terminal end of the human odontoblast process: A studyusing SEM and confocal microscopy. Clin Oral Investig 3:126–132.

Grachtchouk M, Liu J, Wang A, Wei L, Bichakjian CK, Garlick J, Paulino AF, Giordano T,Dlugosz AA. 2006. Odontogenic keratocysts arise from quiescent epithelial rests and areassociated with deregulated hedgehog signaling in mice and humans. Am J Pathol 169:806–814.

Gronthos S, Mankani M, Brahim J, Robey PG, Shi S. 2000. Postnatal human dental pulp stemcells (DPSCs) in vitro and in vivo. Proc Natl Acad Sci USA 97:13625–13630.

Hashimoto N, Jin H, Liu T, Chensue SW, Phan SH. 2004. Bone marrow-derived progenitorcells in pulmonary fibrosis. J Clin Invest 113:243–252.

Havens AM, Chiu E, Taba M, Wang J, Shiozawa Y, Jung Y, Taichman LS, D’Silva NJ,Gopalakrishnan R, Wang C, Giannobile WV, Taichman RS. 2008. Stromal-derived factor-1alpha (CXCL12) levels increase in periodontal disease. J Periodontol 79:845–853.

Hess D, Li L, Martin M, Sakano S, Hill D, Strutt B, Thyssen S, Gray DA, Bhatia M. 2003.Bonemarrow-derived stem cells initiate pancreatic regeneration. Nat Biotechnol 21:763–770.

Huang AH, Snyder BR, Cheng PH, Chan AW. 2008. Putative dental pulp-derived stem/stromal cells promote proliferation and differentiation of endogenous neural cells in thehippocampus of mice. Stem Cells 26:2654–2663.

Jackson KA, Majka SM,Wang H, Pocius J, Hartley CJ, Majesky MW, Entman ML, Michael LH,Hirschi KK, Goodell MA. 2001. Regeneration of ischemic cardiac muscle and vascularendothelium by adult stem cells. J Clin Invest 107:1395–1402.

Jiang HW, Ling JQ, Gong QM. 2008. The expression of stromal cell-derived factor 1 (SDF-1)in inflamed human dental pulp. J Endod 34:1351–1354.

Jin H, Aiyer A, Su J, Borgstrom P, Stupack D, Friedlander M, Varner J. 2006. A homingmechanism for bone marrow-derived progenitor cell recruitment to the neovasculature. JClin Invest 116:652–662.

KoideY,Morikawa S,Mabuchi Y,MugurumaY,Hiratsu E, HasegawaK, KobayashiM, AndoK,Kinjo K, Okano H, Matsuzaki Y. 2007. Two distinct stem cell lineages in murine bonemarrow. Stem cells 25:1213–1221.

Krause DS, Theise ND, Collector MI, Henegariu O, Hwang S, Gardner R, Neutzel S, SharkisSJ. 2001.Multi-organ,multi-lineage engraftment by a single bonemarrow-derived stem cell.Cell 105:369–377.

Lagasse E, Connors H, Al-DhalimyM, ReitsmaM, DohseM,Osborne L,Wang X, FinegoldM,Weissman IL, Grompe M. 2000. Purified hematopoietic stem cells can differentiate intohepatocytes in vivo. Nat Med 6:1229–1234.

Laird DJ, von Andrian UH, Wagers AJ. 2008. Stem cell trafficking in tissue development,growth, and disease. Cell 132:612–630.

Li S, Tu Q, Zhang J, Stein G, Lian J, Yang PS, Chen J. 2008. Systemically transplanted bonemarrow stromal cells contributing to bone tissue regeneration. J Cell Physiol 215:204–209.

McGrath KE, Koniski AD, Maltby KM, McGann JK, Palis J. 1999. Embryonic expression andfunction of the chemokine SDF-1 and its receptor, CXCR4. Dev Biol 213:442–456.

Mezey E, Key S, Vogelsang G, Szalayova I, Lange GD, Crain B. 2003. Transplanted bonemarrow generates new neurons in human brains. ProcNatl Acad Sci USA 100:1364–1369.

Ohta S, Yamada S, Matuzaka K, Inoue T. 2007. The behavior of stem cells and progenitor cellsin the periodontal ligament during wound healing as observed using immunohistochemicalmethods. J Periodontal Res 43:595–603.

Ott I, Keller U, Knoedler M, Gotze KS, Doss K, Fischer P, Urlbauer K, Debus G, von BubnoffN, Rudelius M, Schomig A, Peschel C, Oostendorp RA. 2005. Endothelial-like cellsexpanded from CD34þ blood cells improve left ventricular function after experimentalmyocardial infarction. FASEB J 19:992–994.

Ratajczak MZ, Zuba-Surma E, Kucia M, Reca R, Wojakowski W, Ratajczak J. 2006. Thepleiotropic effects of the SDF-1-CXCR4 axis in organogenesis, regeneration andtumorigenesis. Leukemia 20:1915–1924 .

SeoBM,MiuraM,Gronthos S, Bartold PM,Batouli S, Brahim J, YoungM, Robey PG,WangCY,Shi S. 2004. Investigation of multipotent postnatal stem cells from human periodontalligament. Lancet 364:149–155.

Shi S, Gronthos S. 2003. Perivascular niche of postnatal mesenchymal stem cells in humanbone marrow and dental pulp. J Bone Miner Res 18:696–704.

Son BR,Marquez-Curtis LA, Kucia M,Wysoczynski M, Turner AR, Ratajczak J, RatajczakMZ,Janowska-Wieczorek A. 2006. Migration of bone marrow and cord blood mesenchymalstem cells in vitro is regulated by stromal-derived factor-1-CXCR4 and hepatocyte growthfactor-c-met axes and involves matrix metalloproteinases. Stem Cells 24:1254–1264.

Uehara M, Helman JI, Lillie JH, Brooks SL. 2001. Blood supply to the platysmamuscle flap: Ananatomic study with clinical correlation. J Oral Maxillofac Surg 59:642–646.

Wieczorek G, Steinhoff C, Schulz R, Scheller M, Vingron M, Ropers HH, Nuber UA. 2003.Gene expression profile of mouse bone marrow stromal cells determined by cDNAmicroarray analysis. Cell Tissue Res 311:227–237.

Woodbury D, Schwarz EJ, Prockop DJ, Black IB. 2000. Adult rat and human bone marrowstromal cells differentiate into neurons. J Neurosci Res 61:364–370.

Yu J, Deng Z, Shi J, Zhai H, Nie X, Zhuang H, Li Y, Jin Y. 2006. Differentiation of dental pulpstem cells into regular-shaped dentin-pulp complex induced by tooth germ cellconditioned medium. Tissue Eng 12:3097–3105.


2090 Z H O U E T A L .

role of bone marrow-derived progenitor cells in the maintenance and regeneration of dental...

Documents