Fully functional bioengineered tooth replacementas an organ replacement therapyEtsuko Ikedaa,b,1, Ritsuko Moritaa,c,1, Kazuhisa Nakaoa,c, Kentaro Ishidaa,c, Takashi Nakamuraa,c,Teruko Takano-Yamamotod, Miho Ogawab, Mitsumasa Mizunoa,c,d, Shohei Kasugaie, and Takashi Tsujia,b,c,2

aDepartment of Biological Science and Technology, Faculty of Industrial Science and Technology, and cResearch Institute for Science and Technology,Tokyo University of Science, Noda, Chiba 278-8510, Japan; bOrgan Technologies Inc., Tokyo 101-0048, Japan; dDivision of Orthodontics and DentofacialOrthopedics, Graduate School of Dentistry, Tohoku University, Sendai, Miyagi 980-8575, Japan; and eOral and Maxillofacial Surgery, Departmentof Oral Restitution, Division of Oral Health Sciences, Graduate School, Tokyo Medical and Dental University, Tokyo 113-8510, Japan

Edited by Robert Langer, Massachusetts Institute of Technology, Cambridge, MA, and approved June 30, 2009 (received for review March 17, 2009)

Current approaches to the development of regenerative therapieshave been influenced by our understanding of embryonic devel-opment, stem cell biology, and tissue engineering technology. Theultimate goal of regenerative therapy is to develop fully function-ing bioengineered organs which work in cooperation with sur-rounding tissues to replace organs that were lost or damaged as aresult of disease, injury, or aging. Here, we report a successful fullyfunctioning tooth replacement in an adult mouse achieved throughthe transplantation of bioengineered tooth germ into the alveolarbone in the lost tooth region. We propose this technology as amodel for future organ replacement therapies. The bioengineeredtooth, which was erupted and occluded, had the correct toothstructure, hardness of mineralized tissues for mastication, andresponse to noxious stimulations such as mechanical stress andpain in cooperation with other oral and maxillofacial tissues. Thisstudy represents a substantial advance and emphasizes the po-tential for bioengineered organ replacement in future regenera-tive therapies.

regenerative therapy � transplantation

The current approaches being used to develop future regen-erative therapies are influenced by our understanding of

embryonic development, stem cell biology, and tissue engineer-ing technology (1–4). One of the more attractive concepts underconsideration in regenerative therapy is stem cell transplantationof enriched or purified tissue-derived stem cells (5), or in vitromanipulated embryonic stem (ES) and induced pluripotent stem(iPS) cells (6, 7). This therapy has the potential to restore thepartial loss of organ function by replacing hematopoietic stemcells in hematopoietic malignancies (8), neural stem cells inParkinson’s disease (9), mesenchymal stem cells in myocardialinfarction (10), and hepatic stem cells in cases of hepaticinsufficiency (11).

The ultimate goal of regenerative therapy is to develop fullyfunctioning bioengineered organs that can replace lost or dam-aged organs following disease, injury, or aging (4, 12–14). Thefeasibility of this concept has essentially been demonstrated bysuccessful organ transplantations for various injuries and dis-eases (15). It is expected that bioengineering technology will bedeveloped for the reconstruction of fully functional organs invitro through the precise arrangement of several different cellspecies. However, these technologies have not yet achieved3-dimensional reconstructions of fully functioning organs. Toachieve the functional replacement of lost or damaged tissuesand organs, the development of 3-dimensional bioengineeredtissues comprising a single cell type is now being attempted usingbiodegradative materials (3), appropriate cell aggregation (16),or uniform cell sheets (17). These are now clinically applied forcorneal dysfunction (18), myocardial infarction (19), and hepaticinsufficiency (20) using oral mucosal epithelial cells, myocardialcells, and liver cells, respectively, with favorable clinical results.

A concept has also now been proposed to develop a bioengi-neered organ by reproducing the developmental processes dur-ing organogenesis (13, 21, 22). Almost all organs arise from theirrespective germs through reciprocal interactions between theepithelium and mesenchyme in the developing embryo (23–25).Therefore, it is predicted that a functional bioengineered organcould be produced by reconstituting organ germs betweenepithelial and mesenchymal cells in vitro, although the existenceof organ-inductive stem cells in the adult body has not been fullyelucidated yet with the exception of hair follicles (26) and themammary gland (27). Tooth replacement regenerative therapy,which is also induced by typical reciprocal epithelial, and mes-enchymal interactions (25, 28), is thought to be a feasible modelsystem to evaluate the future clinical application of bioengi-neered organ replacement (13, 21). The strategy to develop abioengineered third tooth after the loss of deciduous andpermanent teeth is to properly reproduce the processes whichoccur during embryonic development through the reconstitutionof a bioengineered tooth germ in vitro (21). We have recentlydeveloped a method for creating 3-dimensional bioengineeredorgan germ, which can be used as an ectodermal organ such asthe tooth or whisker follicle (29). Our analyses have provided aneffective method for reconstituting this organ germ and raisedthe possibility of tooth replacement with integrated blood vesselsand nerve fibers in an adult oral environment (29). However, itremains to be determined whether a bioengineered tooth canachieve full functionality, including sufficient masticatory per-formance, biomechanical cooperation with tissues in the oraland maxillofacial regions, and proper responsiveness via sensoryreceptors to noxious stimulations in the maxillofacial region.There are currently no published reports describing successfulreplacement with a fully functional bioengineered organ.

In our current study, we describe a fully functioning toothreplacement achieved by transplantation of a bioengineeredtooth germ into the alveolar bone of a lost tooth region in anadult mouse. We propose this as a model for future organreplacement therapy. The bioengineered tooth, which waserupted and reached occlusion in the oral environment, had thecorrect tooth structure, hardness of mineralized tissues formastication, and responsiveness to experimental orthodontictreatment and noxious stimulation in cooperation with tissues inthe oral and maxillofacial regions. Our results thus demonstrate

Author contributions: T.T. designed research; E.I., R.M., and K.N. performed research; E.I.,K.N., T.T.-Y., and S.K. contributed new reagents/analytic tools; E.I., R.M., K.N., K.I., T.N.,M.O., and M.M. analyzed data; and E.I., R.M., K.N., and T.T. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Freely available online through the PNAS open access option.

1E.I. and R.M. contributed equally to this work.

2To whom correspondence should be addressed. E-mail: t-tsuji@rs.noda.tus.ac.jp.

This article contains supporting information online at www.pnas.org/cgi/content/full/0902944106/DCSupplemental.

www.pnas.org�cgi�doi�10.1073�pnas.0902944106 PNAS � August 11, 2009 � vol. 106 � no. 32 � 13475–13480

MED

ICA

LSC

IEN

CES

the potential of bioengineered organ replacement for use infuture regenerative therapies.

ResultsEruption and Occlusion of a Bioengineered Tooth. We first investi-gated whether a bioengineered molar tooth germ, which wasreconstituted from embryonic day 14.5 (ED14.5) molar toothgerm-derived epithelial and mesenchymal cells by our previouslydeveloped organ germ method, could erupt and reach occlusionwith an opposing tooth in the mouse adult oral environment(Fig. 1A). After 5–7 days in an organ culture, a single bioengi-neered molar tooth germ, which had developed at the early bellstage of a natural tooth germ and was with a mean length of534.4 � 45.6 �m (Fig. 1B), was then transplanted with the correctorientation into a properly-sized bony hole in the upper firstmolar region of the alveolar bone in an 8-week-old adult murinelost tooth transplantation model. In this model, the upper firstmolar had been extracted, and the resulting wounds had beenallowed to heal for 3 weeks (Fig. 1 A and Fig. S1 A). The cusp tipof the bioengineered tooth was exposed into the oral cavity at36.7 � 5.5 days after transplantation at a frequency of 34/60(56.6%) (Fig. 1C Center and Fig. S1 B–D Center). In currenttransplantation model, the non-erupted explants also occurredat low frequency and were due to the microsurgery for thetransplantation, such as transplantation with the reverse direc-tion or the falling off the explants. The vertical dimension of thetooth crown continually increased and the bioengineered toothfinally reached the plane of occlusion with the opposing lowerfirst molar at 49.2 � 5.5 days after transplantation (Fig. 1C Right,and Fig. S1 B–D Right and E). During the course of eruption andocclusion, the alveolar bone at the bony hole gradually healed inthe areas around the bioengineered tooth and the regeneratedtooth had sufficient periodontal space between itself and thealveolar bone (Fig. 1D and Fig. S1D). The bioengineered tooth

also formed a correct structure comprising enamel, ameloblast,dentin, odontoblast, dental pulp, alveolar bone, and bloodvessels (Fig. 1D). It is known that mice have a considerableamount of cellular cementum that increases in thickness both onthe sides of the roots and in the interradicular area and formsaround the apex of the molar roots (30). The fully occludedbioengineered tooth was also observed to have a large amountof cellular cementum that was equivalent to a normal murinemolar tooth (Fig. 1D and Fig. S1 A). The root of the bioengi-neered tooth was also observed to be surrounded by sufficientperiodontal ligaments (PDL) (Fig. 1D). Observations of thebioengineered tooth morphology revealed that the crown hadplural cusp structure. The lengths and crown widths of theerupted bioengineered teeth were 1,474.4 � 115.1 and 690.7 �177.7 �m, respectively. However, the bioengineered tooth wassmaller than the other normal teeth, since at present we cannotregulate the crown width, cusp position, and tooth patterningincluding anterior/posterior and buccal/lingual structures usingin vitro cell manipulation techniques.

We also transplanted green fluorescence protein (GFP)-labeled bioengineered tooth germ, which was reconstituted bynormal epithelial cells and the mesenchymal cells from GFP-transgenic mice into non-transgenic mice as described above(29). A GFP-labeled bioengineered tooth was produced andcould be observed in the bony hole in the alveolar bone of adultmice (Fig. 1E and Fig. S1F). GFP-positive mesenchymal cellswere also detectable both in the odontoblasts and in the dentalpulp and PDL, which differentiate from the dental papilla anddental follicle cells, respectively (Fig. 1F). Green fluorescencewas also observed in the dentinal tubules of the GFP-positiveodontoblasts in the regenerated tooth (Fig. 1F Lower).

We next investigated the gene expression profiles of colony-stimulating factor 1 (Csf1) and parathyroid hormone receptor(Pthr1), which are thought to regulate osteoclastogenesis during

A B

C

D

E

F

G H

Fig. 1. Eruption and occlusion of a bioengineeredtooth. (A) Schematic representation of the transplan-tation technology used for the generation of reconsti-tuted tooth germ. (B) Phase contrast image of bioengi-neered tooth germ on day 5 of an organ culture. (Scalebar, 200 �m.) (C) Oral photographs of a bioengineeredtooth during eruption and occlusion processes, includ-ing before eruption (Left), immediately after eruption(Center), and full occlusion (Right). (Scale bar, 200 �m.)(D) Histological analysis of the bioengineered toothduring the eruption and occlusion processes, includingbefore eruption (Left), immediately after eruption(Center), and full occlusion (Right). (Scale bar, 100 �m.)(E) Oral photograph of a bioengineered tooth recon-stituted using a combination of epithelial cells fromnormal mice and mesenchymal cells from GFP-trans-genic mice (GFP bioengineered tooth). A merged im-age is shown. (Scale bar, 200 �m.) (F) A sectional imageof a GFP bioengineered tooth. Fluorescent and DICimages are merged. (Scale bar, 100 �m.) (G) Oral pho-tographs showing occlusion of normal (Upper) andbioengineered (Lower) teeth. (Scale bar, 200 �m.) (H)MicroCT images of the occlusion of normal (Left) andbioengineered (Right) teeth. External (Left) and crosssection (Right) images are shown. The bioengineeredtooth is indicated by the arrowhead.

13476 � www.pnas.org�cgi�doi�10.1073�pnas.0902944106 Ikeda et al.

tooth eruption (31). Those genes were detectable in the eruptionpathway and at the boundary surface between the dental follicleof the bioengineered tooth and osseous tissues, as is seen innormal teeth (Fig. S2). These observations suggest that theeruption of the bioengineered tooth germ faithfully reproducedthe molecular mechanisms involved in the normal tooth eruptionprocess.

We next analyzed the occlusion established between thebioengineered tooth and the opposing lower teeth. We oftenobserved that the bioengineered tooth moved physiologicallybefore achieving the occlusion during the transplantation ex-periments. The regenerated tooth achieved normal occlusion inharmony with other teeth in the recipient animal and hadopposing cuspal contacts that maintained the proper occlusalvertical dimensions between the opposing arches (Fig. 1 G andH, and Fig. S1 B–E). Following the achievement of occlusion at49.2 � 5.5 days after transplantation, there was no excessiveincrease in the tooth length or perforation of the maxillary sinusby the erupted bioengineered tooth at up to 120 days aftertransplantation. These results indicated that the bioengineeredtooth moved in response to mechanical stress and achievedfunctional occlusion with the opposing natural tooth.

Masticatory Potential of the Bioengineered Tooth. The masticatorypotential of a bioengineered tooth is essential for achievingproper tooth function (32). We thus performed a Knoop hard-ness test, which is a test for mechanical hardness and is used inparticular for very brittle materials or thin sheets. This was animportant parameter for evaluating masticatory functions in ourbioengineered tooth, including both the dentin and the enamelcomponents. The Knoop hardness of both the enamel and dentinof normal teeth in 3-week-old and 9-week-old mice significantlyincreases in according to the postnatal period (Fig. 2). Thesevalues for enamel and dentin in the normal teeth of 9-week-oldadult mice were measured at 447.7 � 88.9 and 88.4 � 10.2 Knoophardness number (KHN), respectively (Fig. 2). The same mea-surements in the bioengineered tooth were 461.1 � 83.2 and81.4 � 7.53 KHN, respectively (Fig. 2). These findings indicatedthat the hardness of the bioengineered tooth is in the normalrange.

Bioengineered Tooth Response to Mechanical Stress. It has beenpostulated that regeneration of a fully functional tooth could beachieved by fulfilling critical functions in an adult oral environ-ment such as the cooperation of the bioengineered tooth with theoral and maxillofacial regions through the PDL (31, 33). Histo-chemical analysis of the PDL of our bioengineered tooth (Fig.1D) showed a positive connection between this tooth and the

alveolar bone, and suggesting that this tooth may be responsiveto mechanical stress. It has been demonstrated previously thatalveolar bone remodeling is induced via the response of the PDLto mechanical stress such as the treatment of orthodonticmovements (31, 33). These same studies have further demon-strated that the localization of osteoclasts for bone resorptionand osteoblasts for bone formation can be observed in the areaof compression and on the tension side, respectively (31, 33).Thus, we analyzed the movement of our bioengineered tooth andalso the osteoclast and osteoblast localization for remodeling inthe alveolar bone by inducing orthodontic movements experi-mentally.

When the bioengineered tooth was moved buccally for 17 dayswith a mechanical force in an experimental tooth movementmodel, it performed as well as a normal tooth (Fig. 3A and Fig.S3). Histochemical analysis additionally revealed morphologicalchanges in the PDL in both the sides containing lingual tensionand buccal compression following 6 days of treatment (Fig. 3Aand Fig. S3). Osteoblast-like cells, which have a cuboidalshape and rounded nuclei, and osteoclast-like cells, which aremultinucleated giant cells, were observed on the surface of thealveolar bone within the tension and compression sides, respec-tively (Fig. 3A and Fig. S3). During experimental tooth move-ment, tartrate-resistant acid phosphatase (TRAP)-positive

Fig. 2. Assessment of the hardness of the bioengineered tooth. Knoopmicrohardness values of the enamel (Left) and dentin (Right) of the bioengi-neered tooth at 11-weeks post transplantation were compared with those ofnormal teeth from 3- and 9-week-old mice. Error bars show the standarddeviation (n � 3). P � 0.001 (*) and �0.0001 (**) was regarded as statisticallysignificant (t test).

A

B

C

Fig. 3. Experimental tooth movement. (A) Horizontal sections of the root ofa normal tooth (Upper) and a bioengineered tooth (Lower) were analyzed byhematoxylin-eosin staining (HE) at days 0 (Left), 6 (Center), and 17 (Right) ofexperimental orthodontic treatment. (Scale bar, 100 �m.) (B) Sections of anormal and bioengineered tooth were analyzed by TRAP staining and in situhybridization of Ocn at day 6 of the orthodontic treatment. TRAP-positive cells(arrow) and Ocn mRNA-positive cells (arrowhead) are indicated. (Scale bar,100 �m.) (C) The root of the bioengineered tooth was analyzed for boneformation. The image in the box in Left is shown at higher magnification inRight. Tetracycline (arrowhead) and calcein (arrow) labeling was detectableon the tension side. (Scale bar, 50 �m.)

Ikeda et al. PNAS � August 11, 2009 � vol. 106 � no. 32 � 13477

MED

ICA

LSC

IEN

CES

staining was observed in the multinucleated cells, indicating thatosteoclast-like giant cells were dominant on the compression side(Fig. 3B). In contrast, the localization of osteocalcin (Ocn)mRNA-positive cells was observed in the cells on the tensionside, indicating that osteoblast-like cells were dominant (Fig.3B). A fluorescent double-labeling experiment using calcein andtetracycline further showed that incorporation of these reagentsinto the alveolar bone on the tension side, but not the compres-sion side, was clearly observable in the double-labeled line within10 days of the orthodontic treatment (Fig. 3C). These findingssuggest that the PDL of the bioengineered tooth successfullymediates bone remodeling via the proper localization of oste-oclasts and osteoblasts in response to mechanical stress.

Perceptive Potential of Neurons Entering the Tissue of the Bioengi-neered Tooth. The perception of noxious stimulations such asmechanical stress and pain, are important for the protection andproper functions of teeth (34). Neurons in the trigeminal gan-glion, which innervate the pulp and PDL, can detect these stressevents and transduce the corresponding perceptions to thecentral nervous system (34). We have previously reported thatnerve fibers are detectable in the pulp of a developing bioengi-neered tooth in the oral cavity (29). In our current experiments,we evaluated the responsiveness of nerve fibers in the pulp andPDL of the bioengineered tooth to induced noxious stimulations.

Anti-neurofilament (NF)-immunoreactive nerve fibers weredetected in the pulp, dentinal tubules, and PDL of the bioengi-neered tooth as in a normal tooth (Fig. 4A and Fig. S4).Neuropeptide Y (NPY), which is synthesized in sympatheticnerves (34), was also detected in the pulp and PDL neurons (Fig.4A and Fig. S4 C and D). Calcitonin gene-related peptide(CGRP), which is synthesized in sensory nerves and is involvedin sensing tooth pain (34) was also observed in both pulp andPDL neurons (Fig. 4A and Fig. S4 E and F). NPY and CGRPwere detected in both the anti-NF positive and negative-immunoreactive neurons (Fig. 4A and Fig. S4 C–F).

We next evaluated the perceptive potential of these neuronsin the bioengineered tooth against noxious stimulations such asorthodontic treatment and pulp stimulation. The expression ofgalanin, which is a neuropeptide involved in pain transmission(35), increased in response to persistent painful stimulation ofthe nerve terminals within the PDL of the bioengineered toothto the same extent as in a normal tooth (Fig. 4B). Thus PDLnerve fibers in the bioengineered tooth appear to respond tonociceptive stimulation caused by our experimental tooth move-ments. Previous studies have reported that neurons expressingthe proto-oncogene c-Fos protein are detectable in the super-ficial layers of the medullary dorsal horn following noxiousstimulations such as electrical, mechanical and chemical stimu-lation of intraoral receptive fields involving the tooth pulp, PDL,and peripheral nerves innervating the intraoral structures (34,35). We found in our current analyses that the c-Fos-immunoreactive neurons present in both the normal tooth andthe bioengineered tooth drastically increased at 2 h after exper-imental tooth movement, and then gradually decreased within48 h (Fig. 4C). Following pulp stimulation, positive neurons inboth normal and bioengineered teeth also increased at 2 h afterstimulation, but could not be detected at 48 h (Fig. 4D). Thesedata indicate that the nerve fibers innervating both the pulp andPDL of the bioengineered tooth have perceptive potential fornociceptive stimulations and can transduce these events to thecentral nervous system (the medullary dorsal horn).

DiscussionWe successfully demonstrate herein that our bioengineered toothgerm develops into a fully functioning tooth with sufficient hardnessfor mastication and a functional responsiveness to mechanical stressin the maxillofacial region. We also show that the neural fibers that

have re-entered the pulp and PDL tissues of the bioengineeredtooth have proper perceptive potential in response to noxiousstimulations such as orthodontic treatment and pulp stimulation.

A

B

C

D

Fig. 4. Pain response to mechanical stress. (A) Nerve fibers in the pulp and PDL inthe normal (Upper) and bioengineered (Lower) tooth were analyzed immunohisto-chemicallyusingspecificantibodiesforthecombination(left2columns)ofNF(green)andNPY(red)andthecombination(right2columns)ofNFandCGRP(red).(Scalebar,25 �m.) (B) Analysis of galanin immunoreactivity in the PDL of a normal (Upper) andbioengineered (Lower) tooth for the assessment of orthodontic force. No galaninexpressionwasevidentintheuntreatedtooth(Left).Galaninexpression(arrowhead)was detected in the PDL of a normal and bioengineered tooth after 48 h of orth-odontic treatment (Right). (Scale bar, 25 �m.) (C) Analysis of c-Fos-immunoreactivityinthemedullarydorsalhornofmicewithanormaltooth(Upper)orabioengineeredtooth(Lower)after0h(Left),2h(Center),and48h(Right)oforthodontictreatment.c-Fos expression (arrowhead) was also detected. (Scale bar, 50 �m.) (D) Analysis ofc-Fos immunoreactivity in the medullary dorsal horn of mice with a normal tooth(Upper) or a bioengineered tooth (Lower) after 0 h (Left), 2 h (Center), and 48 h(Right)of stimulationbypulpexposure. c-Fosexpression (arrowhead)wasevident inthe medullary dorsal horn after 2 and 48 h of pulp exposure. (Scale bar, 50 �m.) D,dentin; P, pulp; B, bone; PDL, periodontal ligament; T, spinal trigeminal tract.

13478 � www.pnas.org�cgi�doi�10.1073�pnas.0902944106 Ikeda et al.

These findings indicate that bioengineered tooth generation tech-niques can contribute to the rebuilding of a fully functional tooth.

Critical issues in tooth regenerative therapy are whether the bioengi-neered tooth can reconstitute functions such as mastication (32) andresponsive potential to mechanical stress (31, 33) and noxious stimu-lations (34), including cooperation of the regenerated tooth with boththe oral and maxillofacial regions. Eruption and occlusion are essentialfirst steps toward dental organ replacement therapy and successfulincorporation into the oral and maxillofacial region (21, 36). Ourlaboratory has demonstrated previously that a bioengineered toothgerm can develop into a tooth with the correct structure in an adultmouse(29). Ithasalsobeenreportedpreviously thatnormal toothgermisolated from murine embryos and a bioengineered tooth constructedfrom cultured tooth bud cells can develop and erupt in a toothless oralsoft tissue region (diastema) of adult mice and in the tooth extractionsockets of an adult rat (37–42). In our current study, we provideevidence that a bioengineered tooth with the same hardness as an adultnatural toothcaneruptwithnormalgeneexpression, includingCsf1andPthr1, which are thought to regulate osteoclastogenesis, and achievefunctional occlusion with the opposing natural teeth. Previous reportshave suggested that the eruption of tooth germ is generally induced atthe site of tooth development and by the gubernacular cord, which isderived from the epithelium of the dental lamina (43). Hence, ourfindings provide significant insights into tooth eruption mechanismsand strongly suggest that masticatory potential can be successfullyrestored by the transplantation of bioengineered tooth germ.

To establish cooperation between the bioengineered tooth andthe maxillofacial region, 1 critical issue to address is whether afunctional PDL is achieved and thereby the restoration of interac-tions between the bioengineered tooth and the alveolar bone (31,33). The PDL has essential roles in tooth support, homeostasis, andrepair, and is involved in the regulation of periodontal cellularactivities such as cell proliferation, apoptosis, the secretion ofextracellular matrices, the resorption and repair of the root cemen-tum, and remodeling of the alveolar bone proper (31, 33). Althoughimplant therapy has been established and is effective for replace-ment of a missing tooth, this therapy involves osseointegration intothe alveolar bone that does not reconstitute the PDL (44). Theregeneration of PDL has been studied previously using cell sheets(17) and stem cells (22), but has not yet been fully successful. It isthought that orthodontic tooth movement, a process involvingpathogenic and physiologic responses to extreme forces applied toa tooth through bone remodeling controlled by osteogenesis andosteoclastgenesis (31), is a good assay model for the evaluation ofPDL functions. In our present study, the PDL associated with thebioengineered tooth performed in complete cooperation with theoral and maxillofacial regions and bone remodeling successfullyoccurred following the application of orthodontic mechanical force.These findings indicate that it is possible to restore and re-establishcooperation between the bioengineered tooth and maxillofacialregions and thus regenerate critical dental functions.

The peripheral nervous system plays important roles in theregulation of organ functions and the perception of externalstimuli such as pain and mechanical stress (45). During devel-opment of the peripheral nervous system, growing axons navi-gate and establish connections to their developing target organs(46). The recovery of the nervous system, which is associatedwith the reentry of nerve fibers, is critical for organ replacement(47). Although the functions of several internal organs, includingthe liver, kidney, and pancreas, are also mediated by specifichumoral factors such as hormones and cytokines via bloodcirculation (45), perceptions of external stimuli are also essentialto the functions of several organs, such as the eye, limbs, andteeth (45). The tooth is well recognized as a peripheral target

organ for sensory trigeminal nerves, which are required for thefunction and protection of the teeth (46). It is known also thatthe perception of mechanical forces during mastication is limitedin implant patients (48). Thus, the restoration of nerve functionsis also critical for tooth regenerative therapy and future organreplacement therapy (13, 45). In our current study, we demon-strate that several species of nerve fibers, including NF, NPY,CGRP, and galanin-immunoreactive neurons, successfully re-entered both the pulp and/or PDL region of the bioengineeredtooth. These nerves could thereby transduce the signals fromnoxious stimulations such as mechanical stress by orthodontictreatment and the exposure of pulp. Previous studies have alsorevealed that trigeminal nerve fibers navigate and establish theiraxonal projections into the pulp and PDL during early toothdevelopment in a spatiotemporally controlled manner throughexpression of regulatory factors such as nerve growth factor, glialcell line-derived neurotrophic factor, and semaphorin 3a (46).Our present results suggest the possibility that the transplanta-tion of regenerated tooth germ can induce trigeminal axoninnervation and establishment in an adult jaw through thereplication of trigeminal axon pathfinding and nerve fiber pat-terning during early tooth development (46).

In conclusion, this study provides evidence of a successfulreplacement of an entire and fully functioning organ in an adultbody through the transplantation of bioengineered organ germ,reconstituted by single cell manipulation in vitro. Our studytherefore makes a substantial contribution to the developmentof bioengineering technology for future organ replacementtherapy. Further studies on the identification of available adulttissue stem cells for the reconstitution of a bioengineered toothgerm and the regulation of stem cell differentiation into odon-togenic cell lineage will help to achieve the realization of toothregenerative therapy for missing teeth.

MethodsTransplantation. The upper first molars of 5-week-old C57BL/6 (SLC) mice wereextracted under deep anesthesia. Mice were maintained for 3 weeks to allowfor natural repair of the tooth cavity and oral epithelium. Before transplan-tation, we confirmed using microCT analysis that the remaining tooth rootcomponents and/or the tooth that had developed from them could not beobserved in the bony holes (SI Methods). Following repair, an incision ofapproximately 1.5 mm in length was made through the oral mucosa at theextraction site with fine scissors to access the alveolar bone. A fine pin vice(Tamiya) was used to create a bony hole of about 0.5–1.0 mm in diameter inthe exposed alveolar bone surface. Just before transplantation, we removedthe collagen gel from the bioengineered tooth germ in the in vitro organculture and marked the top of the dental epithelium with vital staining dye,such as methylene blue, to ensure the correct direction of the explants. Theexplants were then transplanted into the bony hole according to the dye. Theincised oral mucosa was next sutured with 8–0 nylon (8–0 black nylon 4 mm1/2R, Bear Medic Corp.) and the surgical site was cleaned. The mice containingthe transplants were fed a powdered diet (Oriental Yeast) and skim milk untilthe regenerated tooth had erupted.

ACKNOWLEDGMENTS. We thank Dr. Masaru Okabe (Osaka University) forkindly providing the C57BL/6-TgN (act-EGFP) OsbC14-Y01-FM131 mice. We arealso grateful to Dr. Toru Deguchi and Dr. Masahiro Seiryu (Tohoku University)for their analysis of the tooth perceptive potential; Dr. Nobuo Takeshita andDr. Yuichi Sakai (Tohoku University) for analysis of experimental tooth move-ment; Dr. Masahiro Saito for critical reading of this manuscript and valuablediscussions; and Mr K. Koga (Carl Zeiss) for providing technical support for ourmicroscopic observations. This work was partially supported by Health andLabour Sciences Research Grants from the Ministry of Health, Labour, andWelfare (No. 21040101) to S.K. and T.T., a Grant-in Aid for Scientific Researchin Priority Areas (No. 50339131) to T.T., a Grant-in-Aid for Scientific Research(A) and by an ‘‘Academic Frontier’’ Project for Private Universities to T.T.(2003–2007) from Ministry of Education, Culture, Sports and Technology,Japan.

1. Brockes JP, Kumar A (2005) Appendage regeneration in adult vertebrates and impli-cations for regenerative medicine. Science 310:1919–1923.

2. Watt FM, Hogan BL (2000) Out of Eden: Stem cells and their niches. Science 287:1427–1430.

Ikeda et al. PNAS � August 11, 2009 � vol. 106 � no. 32 � 13479

MED

ICA

LSC

IEN

CES

3. Langer RS, Vacanti JP (1999) Tissue engineering: The challenges ahead. Sci Am 280:86–89.

4. Atala A (2005) Tissue engineering, stem cells and cloning: Current concepts andchanging trends. Expert Opin Biol Ther 5:879–892.

5. Korbling M, Estrov Z (2003) Adult stem cells for tissue repair–a new therapeuticconcept? N Engl J Med 349:570–582.

6. Chien KR, Moretti A, Laugwitz KL (2004) Development. ES cells to the rescue. Science306:239–240.

7. Nishikawa S, Goldstein RA, Nierras CR (2008) The promise of human induced pluripo-tent stem cells for research and therapy. Nat Rev Mol Cell Biol 9:725–729.

8. Copelan EA (2006) Hematopoietic stem-cell transplantation. N Engl J Med 354:1813–1826.

9. Lindvall O, Kokaia Z (2006) Stem cells for the treatment of neurological disorders.Nature 441:1094–1096.

10. Segers VF, Lee RT (2008) Stem-cell therapy for cardiac disease. Nature 451:937–942.11. Wang X, et al. (2003) The origin and liver repopulating capacity of murine oval cells.

Proc Natl Acad Sci USA 100:11881–11888.12. Griffith LG, Naughton G (2002) Tissue engineering–current challenges and expanding

opportunities. Science 295:1009–1014.13. Ikeda E, Tsuji T (2008) Growing bioengineered teeth from single cells: Potential for

dental regenerative medicine. Expert Opin Biol Ther 8:735–744.14. Purnell B (2008) New release: The complete guide to organ repair. Introduction.

Science 322:1489.15. Lechler RI, Sykes M, Thomson AW, Turka LA (2005) Organ transplantation–how much

of the promise has been realized? Nat Med 11:605–613.16. Layer PG, Robitzki A, Rothermel A, Willbold E (2002) Of layers and spheres: The

reaggregate approach in tissue engineering. Trends Neurosci 25:131–134.17. Yang J, et al. (2007) Reconstruction of functional tissues with cell sheet engineering.

Biomaterials 28:5033–5043.18. Nishida K, et al. (2004) Corneal reconstruction with tissue-engineered cell sheets

composed of autologous oral mucosal epithelium. N Engl J Med 351:1187–1196.19. Miyahara Y, et al. (2006) Monolayered mesenchymal stem cells repair scarred myocar-

dium after myocardial infarction. Nat Med 12:459–465.20. Ohashi K, et al. (2007) Engineering functional two- and three-dimensional liver systems

in vivo using hepatic tissue sheets. Nat Med 13:880–885.21. Sharpe PT, Young CS (2005) Test-tube teeth. Sci Am 293:34–41.22. Duailibi SE, Duailibi MT, Vacanti JP, Yelick PC (2006) Prospects for tooth regeneration.

Periodontol 2000 41:177–187.23. Kuure S, Vuolteenaho R, Vainio S (2000) Kidney morphogenesis: Cellular and molecular

regulation. Mech Dev 92:31–45.24. Hogan BL, Yingling JM (1998) Epithelial/mesenchymal interactions and branching

morphogenesis of the lung. Curr Opin Genet Dev 8:481–486.25. Pispa J, Thesleff I (2003) Mechanisms of ectodermal organogenesis. Dev Biol 262:195–205.26. Claudinot S, Nicolas M, Oshima H, Rochat A, Barrandon Y (2005) Long-term renewal of

hair follicles from clonogenic multipotent stem cells. Proc Natl Acad Sci USA102:14677–14682.

27. Shackleton M, et al. (2006) Generation of a functional mammary gland from a singlestem cell. Nature 439:84–88.

28. Tucker A, Sharpe P (2004) The cutting-edge of mammalian development; how theembryo makes teeth. Nat Rev Genet 5:499–508.

29. Nakao K, et al. (2007) The development of a bioengineered organ germ method. NatMethods 4:227–230.

30. Klingsberg J, Butcher EO (1960) Comparative histology of age changes in oral tissuesof rat, hamster, and monkey. J Dent Res 39:158–169.

31. Wise GE, King GJ (2008) Mechanisms of tooth eruption and orthodontic tooth move-ment. J Dent Res 87:414–434.

32. Manly RS, Braley LC (1950) Masticatory performance and efficiency. J Dent Res 29:448–462.

33. Shimono M, et al. (2003) Regulatory mechanisms of periodontal regeneration. MicroscRes Tech 60:491–502.

34. Byers MR, Narhi MV (1999) Dental injury models: Experimental tools for understandingneuroinflammatory interactions and polymodal nociceptor functions. Crit Rev OralBiol Med 10:4–39.

35. Deguchi T, Takeshita N, Balam TA, Fujiyoshi Y, Takano-Yamamoto T (2003) Galanin-immunoreactive nerve fibers in the periodontal ligament during experimental toothmovement. J Dent Res 82:677–681.

36. Yen AH, Sharpe PT (2006) Regeneration of teeth using stem cell-based tissue engi-neering. Expert Opin Biol Ther 6:9–16.

37. Ohazama A, Modino SA, Miletich I, Sharpe PT (2004) Stem-cell-based tissue engineer-ing of murine teeth. J Dent Res 83:518–522.

38. Song Y, Yan M, Muneoka K, Chen Y (2008) Mouse embryonic diastema region is anideal site for the development of ectopically transplanted tooth germ. Dev Dyn237:411–416.

39. Duailibi SE, et al. (2008) Bioengineered dental tissues grown in the rat jaw. J Dent Res87:745–750.

40. Zhang W, et al. (2009) Tissue engineered hybrid tooth-bone constructs. Methods47:122–128.

41. Abukawa H, et al. (2009) Reconstructing mandibular defects using autologous tissue-engineered hybrid tooth and bone constructs. J Oral Maxillofac Surg 67:335–347.

42. Duailibi MT, et al. (2004) Bioengineered teeth from cultured rat tooth bud cells. J DentRes 83:523–528.

43. Carollo DA, Hoffman RL, Brodie AG (1971) Histology and function of the dentalgubernacular cord. Angle Orthod 41:300–307.

44. Voruganti K (2008) Clinical periodontology and implant dentistry, 5th edition. Br DentJ 205:216.

45. Guyton AC (1967) Textbook of Medical Physiology. Am J Med Sci 253:126.46. Luukko K, Kvinnsland IH, Kettunen P (2005) Tissue interactions in the regulation of

axon pathfinding during tooth morphogenesis. Dev Dyn 234:482–488.47. Bengel FM, et al. (2001) Effect of sympathetic reinnervation on cardiac performance

after heart transplantation. N Engl J Med 345:731–738.48. Hammerle CH, et al. (1995) Threshold of tactile sensitivity perceived with dental

endosseous implants and natural teeth. Clin Oral Implants Res 6:83–90.

13480 � www.pnas.org�cgi�doi�10.1073�pnas.0902944106 Ikeda et al.

Supporting InformationIkeda et al. 10.1073/pnas.0902944106SI MethodsReconstitution of a Bioengineered Tooth Germ from Single Cells.Molar tooth germs were dissected from the mandibles of ED14.5mice. The isolation of tissues and each single cell preparationfrom epithelium and mesenchyme has been described previously(1). Dissociated epithelial and mesenchymal cells were precip-itated by centrifugation in a siliconized microtube and thesupernatant was completely removed. The cell density of theprecipitated epithelial and mesenchymal cells after the removalof supernatants reached a concentration of 5 � 108 cells/mL asdescribed previously (1). Bioengineered molar tooth germ wasreconstituted using our previously described 3-dimensional cellmanipulation method, the ‘organ germ method’ (1). In ourprevious conditions to reconstitute a bioengineered tooth germusing 5 � 104 and 1 � 105 cells of epithelial and mesenchymalcells, respectively, the bioengineered tooth germ generatedplural tooth. In current study, we successfully established acondition to generate a single bioengineered molar from abioengineered primordium by controlling the number of epithe-lial and mesenchymal cells. We used each 5 � 104 cells ofepithelial and mesenchymal cells to generate single tooth struc-tures. The bioengineered tooth germs were incubated for 10 minat 37 °C, placed on a cell culture insert (0.4-�m pore diameter;BD), and then further incubated at 37 °C for 5 days in an in vitroorgan culture as described previously (1).

Animals. C57BL/6 mice and C57BL/6-TgN (act-EGFP) mice werepurchased from SLC Inc. C57BL/6-TgN (act-EGFP) OsbC14-Y01-FM131 mice were obtained from the RIKEN BioresourceCenter (Chiba, Japan). All mouse care and handling compliedwith the NIH guidelines for animal research. All experimentalprotocols were approved by the Tokyo University of ScienceAnimal Care and Use Committee.

Histochemical Analysis and Immunohistochemistry. Histochemicaltissue analyses were performed as described previously (1).Tissue sections (5–10 �m) were stained with hematoxylin-eosinand observed using Axioimager A1 (Carl Zeiss) with AxioCAMMRc5 (Carl Zeiss) microscopes.

For tissue preparation for immunohistochemistry, mice underdeep anesthesia were perfused transcardially with 4% parafor-maldehyde in PBS (-). Tissues were then removed and furtherpostfixed for 2–16 h at 4 °C. After fixation, the tissues wereprepared as described previously (1). For fluorescent immuno-histochemistry, the sections (35 �m) were incubated with thefollowing primary antibodies: neurofilament SMI312 (mouseanti-NF Ab; 1:1,000, Abcam); calcitonin gene-related peptide(goat anti-CGRP Ab; 1:250, AbD Serotec); and Neuropeptide Y(rabbit anti-NPY Ab; 1:250, Peninsula Laboratories). Primaryantibodies were detected using the following secondary antibod-ies: f luorescein isothiocyanate (FITC)-conjugated donkey anti-mouse IgG (1:400, Jackson Immunoresearch Laboratories);Alexa Fluor594-conjugated donkey anti-goat IgG (1:200, Mo-lecular Probes); and Alexa Fluor594-conjugated goat anti-rabbitIgG (1:200, Molecular Probes). All f luorescence microscopyimages were captured under a confocal microscope (LSM 510;Carl Zeiss). Enzyme immunohistochemistry was performed asdescribed (2, 3). Sections of the frontal jaw (50 �m) wereincubated with anti-galanin Ab (1:10,000, Peninsula Laborato-ries). Antibody binding was intensified using the avidin-biotinperoxidase (ABC) method (Vectastain ABC kit, Vector Lab.Inc.). Sections (50 �m) of the brainstem were incubated with

anti-c-Fos Ab (1:10,000, Santa Cruz Biotechnology). The sec-tions were then immunostained with peroxidase-labeled goatanti-rabbit IgG (1:300, Cappel Laboratories) and PAP immunecomplex (1:3,000, Cappel).

In Situ Hybridization. In situ hybridizations were performed using5 �m frozen sections as described previously (1). Digoxygenin-labeled probes for specific transcripts were prepared by PCRwith primers designed using published sequences (GenBankAccession Numbers; Csf1: NM�007778, Pthr1: NM�001083936,Ocn: NM�007541).

Hardness Measurements. Polished enamel and dentin samples of abioengineered tooth at 11 weeks after transplantation and alsoa normal tooth (3, 6, and 9 weeks postnatal) were embedded inacrylic resin (n � 3 for each group). The Knoop hardness (4) testwas then performed using a Miniload Hardness Tester, (HM-102, Mitutoyo) equipped with a Knoop diamond tip (19BAA061,Mitutoyo). Three indentations were made on each specimenwith a 10-g load for 10 s.

Microcomputed Tomography (MicroCT) Measurement. The heads ofmice implanted with bioengineered tooth germ and control micewere fixed in the centric occlusal position and radiographicimaging was then performed by x-ray using an inspeXio SMX-90CT device (Shimadzu) with exposure at 90 kV and 0.1 mA, andwith a source-to-sample distance of 34 mm. Microcomputedtomography was performed using Imaris (Carl Zeiss).

Experimental Orthodontic Treatment. A nickel-titanium wire witha diameter of 0.012 inches (VIM-NT, Oralcare Co., Ltd.) wasbonded to the upper incisors of mice under deep anesthesia witha light curing resin (5). One end of the orthodontic wire was bentin a round shape and was placed around the cervical area of theincisor. The other end of the wire was placed on the buccal sideof the cervical area in the regenerated tooth, bonded with resinmaterial, and then applied to the lingual side of the cervical areaof the tooth for orthodontic treatment in the buccal direction.The aforementioned procedures were used with reference to themethods used by Sakai et al. (6). Orthodontic force was appliedin the lingual direction in the study by Sakai et al., but in thebuccal direction in our present experiment. Experimental toothmovement consisted of a horizontal orthodontic force of about10–15 g applied continuously to the bioengineered tooth of theanimals in the experimental group in a buccal direction using adial tension gauge (Mitutoyo) for 17 days. In the control group,orthodontic force was applied in the buccal direction in theupper first molar of 7-week-old normal C57BL/6 mice in thesame manner as that in the experimental group.

At experimental orthodontic treatment days 0 (before treat-ment), 6 (during treatment), and 17 (completion of the exper-iment), mice were deeply anesthetized and the upper jaws wereexcised and immersed rapidly in 4% paraformaldehyde in PBS(-). After fixation, the tissues were prepared as described pre-viously (1). The tissues were cut into 4-�m thick sectionshorizontally. The horizontal sections of the cervical region of thebioengineered tooth were prepared at approximately one-thirdof the distance from the alveolar crest and were stained withhematoxylin and eosin (HE) at experimental orthodontic treat-ment days 0, 6, and 17. The serial sections at day 6 were alsoanalyzed by TRAP staining and by in situ hybridizations forosteocalcin (Ocn) mRNA. In the control group, the horizontal

Ikeda et al. www.pnas.org/cgi/content/short/0902944106 1 of 6

sections of distal buccal root were prepared and analyzed in asimilar manner to those in the experimental group. For TRAPstaining, small amounts of FAST GARNET GBC (Sigma-Aldrich) and 500 �g/mL of naphthol AS phosphate (Sigma-Aldrich) were dissolved in TRAP buffer containing 0.1 Msodium acetate and 26 mM sodium tartrate (Nakarai Tesque) toobtain a solution at pH 5.2. The 4-�m frozen sections werestained with this solution for 5 min. After staining, the sectionswere washed with distilled water, air-dried, and observed usingan Axioimager A1 (Carl Zeiss,) with AxioCAM MRc5 (CarlZeiss).

Fluorescent Double-Labeling. For fluorescent double-labeling toanalyze bone formation (7), mice received s.c. injections of 16mg/kg calcein (Wako Pure Chemical Industries Ltd.) and 30mg/kg of tetracycline hydrochloride (Sigma-Aldrich) at days 6

and 10 of the orthodontic treatment, respectively. At day 11 ofthe orthodontic treatment, mice were killed with deep etheranesthesia, and each maxilla was dissected into 2 halves and fixedin 70% ethanol. After dehydration, the calcified specimens wereembedded in 4% CMC (carboxyl methyl cellulose, Finetech).Horizontal slices of the specimens were cut into sections (4-�mthickness). Fluorescence microscopy images were captured un-der a confocal microscope (LSM-510 META; Carl Zeiss).

Pulp Exposure. A minimal pinpoint mechanical exposure of thepulp (8) was made in the bioengineered tooth or control normalfirst molar of mice under deep anesthesia using a dental engine(Shofu Inc.) supplied with dental diamond burs (Shofu Inc.). Thediamond point was revolved and kept in contact with the toothuntil the pulp was exposed. For stimulation with cold water, icewas applied to the cavity of the tooth after pulp exposure.

1. Nakao K, et al. (2007) The development of a bioengineered organ germ method. NatMethods 4:227–230.

2. Deguchi T, Takeshita N, Balam TA, Fujiyoshi Y, Takano-Yamamoto T (2003) Galanin-immunoreactive nerve fibers in the periodontal ligament during experimental toothmovement. J Dent Res 82:677–681.

3. Fujiyoshi Y, Yamashiro T, Deguchi T, Sugimoto T, Takano-Yamamoto T (2000) Thedifference in temporal distribution of c-Fos immunoreactive neurons between themedullary dorsal horn and the trigeminal subnucleus oralis in the rat followingexperimental tooth movement. Neurosci Lett 283:205–208.

4. Bartlett JD, Beniash E, Lee DH, Smith CE (2004) Decreased mineral content in MMP-20null mouse enamel is prominent during the maturation stage. J Dent Res 83:909–913.

5. Shimpo S, et al. (2003) Compensatory bone formation in young and old rats duringtooth movement. Eur J Orthod 25:1–7.

6. Sakai Y, et al. (2009) CTGF and apoptosis in mouse osteocytes induced by toothmovement. J Dent Res 88:345–350.

7. Kawakami M, Takano-Yamamoto T (2004) Local injection of 1,25-dihydroxyvitamin D3enhanced bone formation for tooth stabilization after experimental tooth movementin rats. J Bone Miner Metab 22:541–546.

8. Byers MR, Narhi MV (1999) Dental injury models: Experimental tools for understandingneuroinflammatory interactions and polymodal nociceptor functions. Crit Rev OralBiol Med 10:4–39.

Ikeda et al. www.pnas.org/cgi/content/short/0902944106 2 of 6

C

D

E

A

B

F

Fig. S1. Eruption and occlusion of a bioengineered tooth. (A) A normal first molar that reached occlusion (Left) and the upper jaw with healing wound afterfirst molar extraction (Right) was analyzed histologically by hematoxylin-eosin staining (HE). (Scale bar, 200 �m.) (B) Oral photographs of the bioengineered toothduring the processes of eruption and occlusion, including before eruption (Left), immediately after eruption (Center), full occlusion (Right) at a 45-degree view(Upper) and occlusal view (Lower). (Scale bar, 200 �m.) (C) Oral photographs of a bioengineered tooth during the processes of eruption and occlusion, includingbefore eruption (Left), immediately after eruption (Center), and full occlusion (Right). (Scale bar, 200 �m.) (D) MicroCT images of a bioengineered tooth duringthe eruption and occlusion processes in the external surface area (Upper) and cross sections (Lower). Images were captured before eruption (Left), immediatelyafter eruption (Center), and after full occlusion (Right). (E) MicroCT images of the maxillofacial region of mice with a bioengineered tooth. (F) Oral photographsdepict the bioengineered tooth that was reconstituted using a combination of epithelial cells from normal mice and mesenchymal cells from GFP-transgenic mice.Upper Left shows a bright field image and Lower Left shows a fluorescent image. Right shows a GFP-labeled bioengineered tooth erupted in the oral environmentof adult mice. The bioengineered tooth is indicated by the arrowhead. (Scale bar, 200 �m.)

Ikeda et al. www.pnas.org/cgi/content/short/0902944106 3 of 6

Fig. S2. Gene expression in a bioengineered tooth during eruption. Normal tooth at postnatal 9 days (Upper) and bioengineered tooth after 14 daystransplantation (Lower) were analyzed histologically during eruption by hematoxylin-eosin staining (HE, Left). In situ hybridisation analyses of gene expressionin the normal and bioengineered tooth germ that had been induced for eruption (Right). Parathyroid hormone receptor 1 (Pthr1, arrowhead) andcolony-stimulating factor 1 (Csf1, arrow) mRNA expression were analyzed in sequential sections of the erupting tooth. AM, ameloblast; EO, enamel organ; PDL,periodontal ligament; B, bone. (Scale bar, 200 �m.)

Ikeda et al. www.pnas.org/cgi/content/short/0902944106 4 of 6

Fig. S3. Experimental tooth movement. (A) Horizontal sections of the normal tooth root were analyzed by hematoxylin-eosin staining (HE) at days 0 (Left),6 (Center), and 17 (Right) of the experimental orthodontic treatment. Higher magnifications of tension side (T) and compression side (C) in PDL are shown inCenter and Lower. (B) The horizontal sections of the root in the bioengineered tooth were analyzed by hematoxylin-eosin staining (HE) at days 0 (Left), 6 (Center),and 17 (Right) of the experimental orthodontic treatment. Higher magnifications of tension side (T) and compression side (C) in PDL are shown in Center andLower. Arrow, osteoblast; arrowhead, osteoclast. (Scale bar, 50 �m.)

Ikeda et al. www.pnas.org/cgi/content/short/0902944106 5 of 6

NPYNF

Pul

pP

DL

DIC

Pul

pP

DL

NF NPY

CGRPNFDIC NF CGRP

NF CIDCID NFA

C

E

Pul

pP

DL

Pul

pP

DL

NF CIDCID NF

NPYNFDIC NF NPY

CGRPNFDIC NF CGRP

B

D

F

Fig. S4. Response to the mechanical stresses of orthodontic force and pulp exposure. (A) Nerve fibers in the normal tooth were immunohistochemicallyanalyzed for neurofilament (NF). DIC (Left), fluorescence (Center), and merged (Right) images are shown. (Scale bar, 200 �m.) (B) Nerve fibers in thebioengineered tooth were analyzed immunohistochemically for NF expression. DIC (Left), fluorescence (Center), and merged (Right) images are shown. (Scalebar, 200 �m.) (C) Nerve fibers in pulp and periodontal ligament (PDL) in the normal tooth were analyzed immunohistochemically by specific antibodies for NFand neuropeptide Y (NPY). DIC (first columns from the left), NF image (second columns), NPY image (third columns), and merged images (fourth columns) areshown. (Scale bar, 25 �m) (D) Nerve fibers in pulp and PDL in the bioengineered tooth were analyzed immunohistochemically by specific antibodies for NF andNPY. DIC (first columns from the left), NF image (second columns), NPY image (third columns), and merged images (fourth columns) are shown. (Scale bar, 25�m.) (E) Nerve fibers in pulp and PDL in the normal tooth were analyzed immunohistochemically by specific antibodies for NF and calcitonin gene related peptide(CGRP). DIC (first columns from the left), NF image (second columns), CGRP image (third columns), and merged images (fourth columns) are shown. (Scale bar,25 �m.) (F) Nerve fibers in pulp and PDL in the bioengineered tooth were analyzed immunohistochemically using specific antibodies for NF and CGRP. DIC (firstcolumns from the left), NF image (second columns), CGRP image (third columns), and merged images (fourth columns) are shown. (Scale bar, 25 �m.)

Ikeda et al. www.pnas.org/cgi/content/short/0902944106 6 of 6

13213 Transgenic corn attracts insect-killing nematodes

13236 Single molecule analysis illuminates DNA replication

13469 Eye drops for glaucoma nerve damage

13475 Growing new organs in place

AGRICULTURAL SCIENCES

Transgenic corn attracts insect-killingnematodesMany plants, when under attack by herbivorous insects, emit vola-tiles that attract natural enemies of the assailants and help wardoff the assault. For example, larvae of the western corn rootworm,

a beetle that is rapidly spread-ing and devastating maize cropsworldwide, induce the roots ofsome maize varieties to emit(E)-�-caryophyllene (E�C),which attracts nematodes thatinfect and kill the voraciousroot pest. Many North Ameri-can maize varieties no longeremit E�C because they havelost E�C synthase activity. Jorg

Degenhardt et al. used transgenic techniques to restore this signalto a nonemitting maize variety by inserting an E�C synthase genefrom oregano. The authors sowed the transgenic maize in experi-mental plots alongside maize lines unable to emit E�C and in-fested the crops with rootworm larvae. When the authors releasednematodes into the bug-ridden plots, the transgenic maize plantssuffered significantly less root damage and had 60% fewer adultbeetles emerge than the nonemitting maize varieties. Manipulatinga plant’s volatile emissions can enhance its natural defenses andmay help researchers find ecologically sound strategies to fight avariety of insect pests, the authors suggest. — B.A.

‘‘Restoring a maize root signal that attracts insect-killing nematodes tocontrol a major pest’’ by Jorg Degenhardt, Ivan Hiltpold, Tobias G.Kollner, Monika Frey, Alfons Gierl, Jonathan Gershenzon, Bruce E.Hibbard, Mark R. Ellersieck, and Ted C. J. Turlings (see pages13213–13218)

BIOCHEMISTRY

Single molecule analysis illuminatesDNA replicationDuring DNA replication, a phalanx of binding proteins and en-zymes synthesize new strands of DNA complementary to theparental strands. This complex of proteins, called a replisome,

tethers together the DNA polymerases that add nucleotides tothe leading and lagging strands, which are synthesized concur-rently. Nina Yao et al. analyzed the movement of the Esche-richia coli replisome duringreplication in real-time, single-molecule studies, which al-lowed for a detailed study ofthe rate of DNA replication aswell as the processivity of thepolymerases. The methodmeasured the number of nu-cleotides added to the repli-cating strand before theenzymes fall off the DNA. The research confirmed that DNAmust loop for both strands to be replicated at the same timeby a single enzyme complex. Although lagging strand synthe-sis increased the processivity of the replisome by 61%, it de-creased the overall rate of replication fork progression by23%. This research sheds light on the mechanism of DNAreplication and illustrates the utility of single-molecule studiesin the understanding of fundamental cellular processes,according to the authors. — F.A.

‘‘Single-molecule analysis reveals that the lagging strand increases repli-some processivity but slows replication fork progression’’ by Nina Y.Yao, Roxana E. Georgescu, Jeff Finkelstein, and Michael E.O’Donnell (see pages 13236–13241)

MEDICAL SCIENCES

Eye drops for glaucoma nerve damageThe buildup of intraocular pressure in glaucoma, a majorcause of irreversible blindness, can sometimes be prevented,but no effective treatment exists to restore nerve function inretinal cells. A study in rats and human patients reveals thatadministering nerve growth factor (NGF) can potentially res-cue retinal ganglion cells from the high intraocular pressureof glaucoma. Alessandro Lambiase et al. used a model ofglaucoma in rats to determine whether treatment with NGFcould prevent the retinal cells from dying. The authors foundthat rats with NGF applied topically to the eye showed lesscell death than controls. The researchers also tested the treat-ment in three human glaucoma patients whose intraocularpressure had been reigned in but who still had progressive

A Diabrotica larva on a maize root.

Leading strand replication of im-

mobilized mini-rolling circle by E.

coli Pol III*.

www.pnas.org�cgi�doi�10.1073�iti3209106 PNAS � August 11, 2009 � vol. 106 � no. 32 � 13145–13146

PNASProceedings of the National Academy of Sciences of the United States of America www.pnas.org

In This Issue

August 11, 2009 � vol. 106 � no. 32 � 13145–13636

vision deterioration. After3 months of treatment, thevision of two of the patientshad improved, while that ofthe third patient stabilized.The improvement remainedat 3 months, and up to18 months, after treatment.NGF is able to cross into theeye and past the blood–brainbarrier, making it a poten-tial treatment for glau-coma, according to theauthors. — T.H.D.

‘‘Experimental and clinical evidence of neuroprotection by nervegrowth factor eye drops: Implications for glaucoma’’ by AlessandroLambiase, Luigi Aloe, Marco Centofanti, Vincenzo Parisi, FlavioMantelli, Valeria Colafrancesco, Gian Luca Manni, Massimo Gil-berto Bucci, Stefano Bonini, and Rita Levi-Montalcini (see pages13469–13474)

MEDICAL SCIENCES

Growing new organs in placeThe ability to grow a fully functional bioengineered organ in-side the body from stem cells or other germ cells could ensurea ready supply of custom replacements. Investigating strategies

for regenerating organs in mice, Etsuko Ikeda et al. engineeredthe growth of replacement teeth in laboratory animals. Tech-nology exists to develop limited tissues in the lab that can betransplanted, and the authors ex-plored ways to grow a 3D organin place, starting with teeth. Theauthors developed a bioengi-neered tooth germ, which is aseed-like tissue containing thecells and instructions necessary toform a tooth, and transplanted thegerm into the jawbones of mice.Tracking gene expression in thetransplanted germ with a fluores-cent protein, the authors showedthat genes normally activated intooth development are active dur-ing the bioengineered replacement’s growth. The engineeredtooth’s hardness was in line with natural teeth, and nerve fibersgrew throughout and responded to pain stimulation. The resultsrepresent an advance in the investigation of bioengineered organreplacement, according to the authors. — T.H.D.

‘‘Fully functional bioengineered tooth replacement as an organreplacement therapy’’ by Etsuko Ikeda, Ritsuko Morita, KazuhisaNakao, Kentaro Ishida, Takashi Nakamura, Teruko Takano-Yamamoto, Miho Ogawa, Mitsumasa Mizuno, Shohei Kasugai, andTakashi Tsuji (see pages 13475–13480)

Episcleral vein (arrow) was used to

induce glaucoma in rats.

GFP-labeled bioengineered

molar tooth.

13146 � www.pnas.org�cgi�doi�10.1073�iti3209106