Insulin-like growth factor 1 modulates bioengineered tooth morphogenesis

Regenerative therapy to replace missing teeth is a critical area of research. Functional bioengineered teeth have been produced by the organ germ method using mouse tooth germ cells. However, these bioengineered teeth are significantly smaller in size and exhibit an abnormal crown shape when compared with natural teeth. The proper sizes and shapes of teeth contribute to their normal function. Therefore, a method is needed to control the morphology of bioengineered teeth. Here, we investigated whether insulin-like growth factor 1 (IGF1) can regulate the sizes and shapes of bioengineered teeth, and assessed underlying mechanisms of such regulation. IGF1 treatment significantly increased the size of bioengineered tooth germs, while preserving normal tooth histology. IGF1-treated bioengineered teeth, which were developed from bioengineered tooth germs in subrenal capsules and jawbones, showed increased sizes and cusp numbers. IGF1 increased the number of fibroblast growth factor (Fgf4)-expressing enamel knots in bioengineered tooth germs and enhanced the proliferation and differentiation of dental epithelial and mesenchymal cells. This study is the first to reveal that IGF1 increases the sizes and cusp numbers of bioengineered teeth via the induction of enamel knot formation, as well as the proliferation and differentiation of dental epithelial and mesenchymal cells.

Regenerative therapy to replace missing teeth is a critical area of research. Functional bioengineered teeth have been produced by the organ germ method using mouse tooth germ cells. However, these bioengineered teeth are significantly smaller in size and exhibit an abnormal crown shape when compared with natural teeth. The proper sizes and shapes of teeth contribute to their normal function. Therefore, a method is needed to control the morphology of bioengineered teeth. Here, we investigated whether insulin-like growth factor 1 (IGF1) can regulate the sizes and shapes of bioengineered teeth, and assessed underlying mechanisms of such regulation. IGF1 treatment significantly increased the size of bioengineered tooth germs, while preserving normal tooth histology. IGF1-treated bioengineered teeth, which were developed from bioengineered tooth germs in subrenal capsules and jawbones, showed increased sizes and cusp numbers. IGF1 increased the number of fibroblast growth factor (Fgf4)-expressing enamel knots in bioengineered tooth germs and enhanced the proliferation and differentiation of dental epithelial and mesenchymal cells. This study is the first to reveal that IGF1 increases the sizes and cusp numbers of bioengineered teeth via the induction of enamel knot formation, as well as the proliferation and differentiation of dental epithelial and mesenchymal cells.
In the field of tooth regeneration, the ultimate goal is the regeneration of fully functional teeth in living bodies 1 . Tsuji's group developed the organ germ method, which is a three-dimensional cell manipulation technique for the generation of bioengineered tooth germs 2 . Upon transplantation into the jawbones of mice, bioengineered tooth germs developed into bioengineered teeth and erupted into the oral cavity; these bioengineered teeth exhibited similar hardness to that of natural teeth, and were responsive to noxious stimuli 3 . These reports suggest that functional bioengineered teeth can be regenerated via transplantation of in vitro-generated bioengineered tooth germs into an edentulous area of the jaw.
The proper sizes and shapes of organs contribute to their normal function. There are site-specific types of teeth in the dentition, such as incisors, canines and molars. These have different sizes and shapes, and play unique roles in stomatognathic function and aesthetics. Tooth size reportedly affects the rate of food breakdown 4 ; furthermore, tooth shape correlates with masticatory efficiency 5 . From a clinical perspective, each patient exhibits individual sizes and shapes of teeth that are genetically and environmentally determined. Accordingly, the goal of dental regenerative therapy is to generate bioengineered teeth that erupt into the oral cavity, with sizes and shapes similar to those of natural teeth, in order to rehabilitate the aesthetics and stomatognathic function of individual patients. However, the bioengineered mouse teeth generated by Tsuji's group were remarkably small in size and exhibited an abnormal crown shape, compared with that of natural teeth 3 . Therefore, a method is needed to control the morphology of bioengineered teeth. Tooth development progresses via interactions between dental epithelial and mesenchymal tissues 1,6 . Various molecules participate in spatio-temporal regulation of cellular processes, including proliferation and differentiation, during tooth development 1,6 . Notably, molecules that affect tooth size and shape during tooth development could be useful for the morphological control of bioengineered teeth. Insulin-like growth factor 1 (IGF1) regulates the normal development of various tissues, such as bone, muscle, and nerve, by regulating cell proliferation, differentiation, apoptosis, and migration [7][8][9] . Leprechaunism is a congenital syndrome associated with insulin resistance; it is characterized by retarded intrauterine and postnatal growth 10 . We previously reported a leprechaunism patient who had undergone long-term treatment with recombinant IGF1 and had developed extremely large teeth 11 . Conversely, Sarnat et al. 12 reported that patients with Laron-type dwarfism, which is characterized by the loss of growth hormone receptors, exhibit a deficiency in IGF1 secretion, such that they develop teeth with reduced mesiodistal width. Moreover, IGF1 and its receptor are expressed in both dental epithelial and mesenchymal tissues in tooth germs 13,14 ; IGF1 signalling promotes cell proliferation, differentiation, and matrix secretion in mouse tooth germs 15,16 . Based on this previous evidence, we hypothesized that IGF1 regulates tooth morphogenesis, and can thus be used to control the size and shape of bioengineered teeth.
The aim of this study was to clarify whether IGF1 has the potential to control the size and shape of bioengineered teeth. We transplanted bioengineered tooth germs, which were treated with IGF1 during organ culture, into the jawbones of living mice, in accordance with the method of our previous report 3 . The sizes and shapes of the developed bioengineered teeth were then examined. We further analysed the effects of IGF1 on enamel knot formation, as well as on the proliferation and differentiation of dental epithelial and mesenchymal cells in vitro, in order to elucidate the underlying biological mechanisms that cause morphological changes in IGF1-treated bioengineered teeth.

IGF1 increases the size of tooth germs in culture.
To investigate the effect of IGF1 on the morphogenesis of tooth germs in organ culture, we first treated tooth germs dissected from embryonic day (ED) 14.5 mouse mandibles with IGF1. As shown in Fig. 1a, the tooth germs exhibited increased size. Treatment with 4 µg/mL IGF1 significantly increased the height of tooth germs at days 7 and 14 (Fig. 1b). The width of IGF1-treated tooth germs became larger than that of control tooth germs at days 7 and 14 ( Fig. 1c). Normal dental histological structure, comprising layers of ameloblasts, odontoblasts, enamel, and dentin, was observed in the control and IGF1-treated groups at day 14 of organ culture (Fig. 1d).

IGF1 increases the size of bioengineered tooth germs.
First, to confirm the absence of crosscontamination of dental epithelial and mesenchymal cells (the sources of cells used for the generation of the bioengineered tooth germs) during the isolation procedure, we analysed the gene expression levels of the dental epithelial markers paired-like homeodomain transcription factor 2 (Pitx2), sonic hedgehog (Shh), and fibroblast growth factor 4 (Fgf4), as well as the expression level of the dental mesenchymal marker Fgf10 17 . Expression of Pitx2, Shh, Fgf4, and Fgf10 was observed in the tooth germs (Supplementary Figs S2 and S3). In dental epithelial tissues and cells, Pitx2, Shh, and Fgf4 were expressed, whereas Fgf10 was not ( Supplementary Figs S2 and S3). In addition, Fgf10 was expressed in dental mesenchymal tissues and cells, whereas Pitx2, Shh, and Fgf4 were not (Supplementary Figs S2 and S3). These results indicated that there was no cross-contamination of dental epithelial and mesenchymal cells.
The contact length between reconstituted epithelial and mesenchymal cell layers affects the size of bioengineered tooth germs generated by recombination of epithelium and mesenchyme and grown in an organ culture in vitro set-up 18 . Therefore, we generated tissue recombinants with a uniform contact length in the control and IGF1-treated groups. The bioengineered tooth germs formed a translucent zone between the epithelial and mesenchymal tissues at day 2, showed tooth crown shape at day 7, and grew further by day 14 (Fig. 2a, Supplementary  Fig. S4). The heights and widths of the bioengineered tooth germs increased in a dose-dependent manner upon treatment with IGF1 (Fig. 2b,c). The heights of bioengineered tooth germs treated with 1, 2, 4, and 8 µg/mL IGF1 were significantly greater than the heights of the control group germs at days 7 and 14 (Fig. 2b). The widths of the bioengineered tooth germs treated with 4 µg/mL were significantly larger than those of the control group at days 7 and 14 (Fig. 2c). The widths of the bioengineered tooth germs treated with 8 µg/mL IGF1 were significantly larger than those of the control group at day 7, but not at day 14 (Fig. 2c). The maximum effects of IGF1 on the height and width of the bioengineered tooth germs were observed at 4 µg/mL. These results indicated that IGF1 increased the size of the bioengineered tooth germs.
Histological analysis of the bioengineered tooth germs in the control and IGF1-treated groups showed dental papillae, odontoblasts, and ameloblast layers at day 7, as well as deposition of enamel and dentin matrix at day 14 (Fig. 2d); this indicated no histological differences between the groups. Immunohistochemical staining demonstrated IGF1 receptor (IGF1R) expression in ameloblasts, odontoblasts and dental papilla cells in the bioengineered tooth germs in both control and IGF1-treated groups at day 7 ( Supplementary Fig. S5). The expression of IGF1R was also observed in ameloblasts and odontoblasts in the bioengineered tooth germs at day 14 of culture ( Fig. 2e). Ameloblastin (AMBN), an enamel matrix protein secreted by ameloblasts 19,20 , was expressed in ameloblasts; dentin matrix acidic phosphoprotein 1 (DMP1), the extracellular matrix secreted by ameloblasts and odontoblasts 21 , was expressed in ameloblasts and at the junction area between dentin and predentin; and dentin sialoprotein (DSP), a component of the dentin extracellular matrix 21 , was expressed in predentin in both groups (Fig. 2e). Thus, obvious differences in the expression patterns of these proteins were not recognized between the control and IGF1-treated groups.

IGF1 increases the size and cusp number of bioengineered teeth developed in subrenal capsules.
We examined the effect of IGF1 on the morphology of the bioengineered teeth. Transplantation of bioengineered tooth germs into subrenal capsules resulted in the generation of bioengineered teeth with surrounding hard tissues in the control and IGF1-treated groups (Fig. 3a,b). The height of the bioengineered teeth in the group treated with 4 µg/mL IGF1 (1379.2 ± 177.7 µm, N = 8) was significantly greater than that in the control group (1151.2 ± 156.7 µm, N = 7) (Fig. 3c). The crown width of bioengineered teeth in the group treated with 4 µg/mL IGF1(582.3 ± 54.0 µm, N = 8) was also greater than that of teeth in the control group (503.8 ± 84.6 µm, N = 7) (Fig. 3d). Cusp number is a measurable aspect of tooth shape patterning 22 . Thus, we examined the cusp number of the bioengineered teeth to analyse the effect of IGF1 on bioengineered tooth morphology. The cusp number of bioengineered teeth in the group treated with 4 µg/mL IGF1 (4.5 ± 1.7, N = 8) was greater than that of teeth in the control group (2.4 ± 0.8, N = 7) (Fig. 3e).
IGF1-treated bioengineered teeth erupt from the jawbone with greater size and increased cusp number. We further examined whether the bioengineered teeth developed from the IGF1-treated bioengineered tooth germs could erupt from a jawbone with greater size and additional cusps. Eruption of the bioengineered teeth gradually progressed over 10 weeks after transplantation in both control and IGF1-treated groups without any pathological observations, such as ankylosis or bone defects (Fig. 4a). Quantitative analysis of micro-computed tomography images indicated greater height of bioengineered teeth in the group treated with 4 µg/mL IGF1 (1500.8 ± 119.6 µm, N = 6) than in the control group (1340.6 ± 117.2 µm, N = 7) at 90 days after transplantation (Fig. 4b,c). The crown width of the bioengineered teeth in the group treated with 4 µg/mL IGF1 (677.5 ± 83.8 µm, N = 6) was also significantly greater than that of teeth in the control group (497.9 ± 118.3 µm, N = 7) (Fig. 4b,d). Furthermore, the cusp number of bioengineered teeth in the group treated with 4 µg/mL IGF1 (4.8 ± 0.8, N = 6) was greater than that of teeth in the control group (3.3 ± 1.4, N = 7) (Fig. 4e). Periodontal ligament (PDL) is a fibrous tissue located between the tooth root and alveolar bone 23 . PDL surrounding the bioengineered tooth root was observed in the control and IGF1-treated groups (Fig. 4f). IGF1 did not visibly affect the histology of the PDL (Fig. 4f). We investigated the expression of the nerve fibre marker, neurofilament (NF), and the vascular endothelial cell marker, von Willebrand factor (VWF), in the PDL. NF and VWF were expressed in the PDL in both control and IGF1-treated groups. There were no clear differences in the expression patterns of NF and VWF in the PDL surrounding control and IGF1-treated bioengineered teeth (Fig. 4g,h).

IGF1 does not affect the hardness of bioengineered teeth.
To analyse the effect of IGF1 on the hardness of the bioengineered teeth, we measured Knoop hardness. The Knoop hardness number (KHN) value of the enamel of natural teeth was 433.7 ± 131.1 (N = 6) (Fig. 5a). The KHN values of the enamel of bioengineered teeth developed by renal subcapsular transplantation showed no significant difference between the control group (201.6 ± 19.9 KHN, N = 7) and the group treated with 4 µg/mL IGF1 (210.9 ± 17.9 KHN, N = 5); notably, both were significantly lower than the value of natural tooth enamel (Fig. 5a). The KHN values of the enamel of bioengineered teeth developed by transplantation into jawbones showed no significant difference between the control group (393.1 ± 20.1 KHN, N = 5) and the group treated with 4 µg/mL IGF1 (404.6 ± 19.2 KHN, N = 5); these values were similar to the value of natural tooth enamel (Fig. 5a). The KHN values of the dentin of the bioengineered teeth developed in subrenal capsules were 81.7 ± 9.5 (N = 7) in the control group and 81.1 ± 8.5 (N = 5) in the group treated with 4 µg/mL IGF1 (Fig. 5b). The KHN values of the dentin of the bioengineered teeth developed in jawbones were 89.7 ± 4.6 (N = 5) in the control group and 79.9 ± 4.2 (N = 5) in the group treated with 4 µg/mL IGF1 (Fig. 5b). These dentin KHN values were comparable to those of natural teeth (90.7 ± 7.1 KHN, N = 6) (Fig. 5b).
IGF1 increases the number of Fgf4-expressing enamel knots during the development of bioengineered tooth germs. The formation of enamel knots, a cluster of non-dividing epithelial cells, is a critical biological process for the patterning of tooth cusps 24,25 . To analyse the effect of IGF1 on enamel knot formation, we measured the expression of the enamel knot marker, Fgf4, during the development of the bioengineered tooth germs 24,25 . Fgf4 expression was not detectable at day 1 (Fig. 6a). Its expression was observed in both groups at day 2; notably, the number of Fgf4-expressing sites increased during the development of the bioengineered tooth germs (Fig. 6a). The number of Fgf4-expressing sites at day 5 in the group treated with 4 µg/mL IGF1 (4.3 ± 0.5, N = 6) was significantly larger than that of the control group (3.2 ± 0.4, N = 6) (Fig. 6b,c).   IGF1 directly upregulates the proliferation of both mesenchymal and epithelial cells derived from tooth germs. To investigate the mechanisms of the IGF1-mediated enhancement in the size of bioengineered teeth, we analysed the effect of IGF1 on the proliferation of dental mesenchymal and epithelial cells isolated from ED14.5 mouse tooth germs. The total cell number and percentage of bromodeoxyuridine (BrdU)-positive mesenchymal cells were significantly higher in the IGF1-treated group than in the control group ( Fig. 7a−d). A significant increase in the total cell number and percentage of BrdU-positive epithelial cells was also observed in the IGF1-treated group (Fig. 7e-h).

IGF1 stimulates the differentiation of mesenchymal and epithelial cells derived from tooth germs.
Previous reports indicate that the interaction of IGF1 and bone morphogenetic protein 2 (BMP2) signalling contributes to the development of various tissues, including tooth, bone, and cartilage [26][27][28] . Both Igf1 and Igf1r were expressed in dental epithelial and mesenchymal tissues in the ED14.5 tooth germs, while Bmp2 was expressed solely in dental epithelial tissue ( Supplementary Figs S6 and S7). We then examined the effect of IGF1 on the odontoblast differentiation of dental mesenchymal cells, focusing on its potential interaction with BMP2. Phase-contrast micrographs revealed a polygonal cell shape in the control and IGF1-treated groups at day 10 of culture (Fig. 8a). In contrast, mesenchymal cells in the BMP2-treated and BMP2 plus IGF1-treated groups were spindle-shaped (Fig. 8a). Dentin sialophosphoprotein (Dspp), Dmp1, and microtubule-associated protein 1b (Map1b) are known as odontoblast markers 21,29 . Expression of Dmp1, Map1b, and Dspp was clearly induced by the simultaneous administration of IGF1 and BMP2, whereas treatment with IGF1 or BMP2 alone did not induce expression of these markers (Fig. 8b-d). Next, we examined the effect of IGF1 and BMP2 on the expression of insulin-like growth factor-binding proteins (IGFBPs), which regulate IGF1 bioactivity 30 . The expression levels of Igfbp2, 3, and 4 were significantly reduced, while the expression of Igfbp6 was significantly enhanced by BMP2 treatment, as well as by simultaneous administration of IGF1 and BMP2 (Fig. 8e-h).
We examined the effect of IGF1 on the differentiation of dental epithelial cells. The epithelial cells reached confluence and showed polygonal shape at day 4 of culture (Fig. 8i). IGF1 treatment did not visibly alter the shape of the epithelial cells (Fig. 8i). Ambn is considered a marker of ameloblast differentiation 19,20 . Analysis of Ambn expression demonstrated significantly greater expression of Ambn in the IGF1-treated group than in the control group (Fig. 8j-l).

Discussion
In the present study, we analysed the effect of IGF1 on the morphology of bioengineered teeth. We demonstrated that IGF1 treatment significantly increased the size of bioengineered tooth germs while maintaining normal dental histological architecture. IGF1 binds to IGF1R with high affinity and activates its signalling pathway 31,32 . IGF1R expression was observed in ameloblasts, odontoblasts, and dental papilla cells during the development of bioengineered tooth germs; these IGF1R expression patterns were similar to those of normal tooth germs [33][34][35] . Our findings suggest that IGF1 increases the size of bioengineered tooth germs by acting on both epithelial and mesenchymal cells. Moreover, the IGF1-treated bioengineered teeth, which were developed from bioengineered tooth germs in subrenal capsules and jawbones, showed an increase not only in size but also in cusp number. These findings indicated for the first time that IGF1 is a competent agent for increasing the size and cusp number of bioengineered teeth. Importantly, IGF1 treatment did not interfere with the formation of PDL surrounding tooth roots and the acquisition of significant hardness of enamel and dentin. PDL has essential roles in tooth function, such as perception, adaptation to mechanical load, and homeostasis of teeth and periodontal tissues 36 . Sufficient tooth hardness is required to exert masticatory potential 37 . Therefore, the IGF1-treated bioengineered teeth could exert normal physiological function as related to PDL and tooth hardness. The biological effects of pharmaceutical agents are dependent on their concentration. We demonstrated that IGF1 increased the size of bioengineered teeth in a dose-dependent manner, representing maximum effects at 4 μg/mL. The physiological concentration of IGF1 in the serum of 4-8-week-old mice is approximately 0.4-1 μg/mL 8,38 . Therefore, the concentration of IGF1 used in the present study is higher than the physiological level. However, the histological structure and the expression patterns of IGF1R and enamel and dentin matrix proteins in the bioengineered tooth germs treated with 4 μg/mL IGF1 showed no abnormal changes when compared with the control group. In addition, pathological findings such as cell necrosis and tissue degeneration were not observed. Therefore, the concentration of IGF1 used in the present study had no adverse effect on the bioengineered tooth germs. We transplanted the bioengineered tooth germs, which had been cultured with IGF1 for 7 days, into mouse subrenal capsules and jawbones. Although exogenous IGF1 was not administered after the transplantation, the generated bioengineered teeth showed increased size and cusp number. Kim et al. 39 cultured tooth germs extracted from ED13.5 mice with the administration of a caspase inhibitor for 2 days, and subsequently transplanted them into mouse subrenal capsules. As a result, teeth with shorter crown heights and wider mesiodistal diameters were generated despite no further administration of the inhibitor to the mice. These findings indicate that the ex vivo effects of agents on the morphology of bioengineered tooth germs are maintained during their development into bioengineered teeth in the living body. This concept is significant for the future clinical application of IGF1 to regulate the morphology of bioengineered teeth.
During tooth germ development, epithelial cells form the enamel knot at the tip of the cap-shaped tooth germ (cap stage) 40 . During the progression of tooth development, the number of enamel knots increases and the location of future cusps is determined 25 . BMP2 expression in dental epithelium induces the formation of enamel knots 41 . BMP4 expression in dental mesenchyme also contributes to enamel knot formation via upregulation of p21 and Msh homeobox 1 (Msx1) 41 . However, the molecular mechanism of enamel knot formation is not fully understood. In the present study, IGF1 increased the number of Fgf4-expressing sites in bioengineered tooth germs, indicating that IGF1 directly and/or indirectly contributes to the production of enamel knots. FGF4 expression in enamel knots upregulates the proliferation of peripheral dental epithelial and mesenchymal cells during tooth morphogenesis 24 . Kratochwil et al. 42 reported that FGF4 expression in enamel knots induces the expression of SHH in dental epithelium, which is an essential factor for tooth morphogenesis, via upregulation of FGF3 in dental mesenchyme. Our data suggest that the IGF1-induced FGF4 in enamel knots could activate proliferation and downstream signalling, subsequently increasing the size and cusp number of bioengineered teeth. The IGF1-dependent increase in the number of enamel knots could subsequently increase the cusp number in bioengineered teeth. On the other hand, generation of more space between enamel knots during growth of tooth germs would allow the formation of new enamel knots, regulated by activators and inhibitors of enamel knot formation 43 . Thus, it is possible that the IGF-induced increase in the size of bioengineered tooth germs indirectly increases the total number of enamel knots via activator-inhibitor loops. We need further study to elucidate regulatory mechanisms of the IGF-induced increase in the size and cusp number of bioengineered teeth.
Regarding the role of IGF1 in the proliferation of tooth germ cells, Young et al. 15 showed an inducible effect of IGF1 on mitotic activity in the dental epithelial cells of ED16 mouse molar tooth germs. Caton et al. 44 reported that IGF1 induces proliferation of dental mesenchymal cells isolated from ED18 mouse molar tooth germs. However, the effect of IGF1 on the proliferation of mesenchymal and epithelial cells in stage of tooth germ development earlier than ED16 has not been clarified. In the present study, we revealed for the first time, that IGF1 directly enhanced the proliferation of both dental mesenchymal and epithelial cells isolated from the tooth germs of ED14.5 mice in vitro. In the molar tooth germ of ED14 mice, epithelial tissue invaginates around the mesenchymal tissue, reaching the cap stage 45 . Subsequently, the size of tooth germs increases as the dental epithelial tissue further invaginates and elongates into mesenchymal tissue by active cell proliferation; the tooth germ then forms a bell-shaped crown at ED16 (bell stage) 45 . Thus, our data suggest that IGF1 contributes to increased tooth germ size by stimulating the proliferation of dental mesenchymal and epithelial cells between the cap and bell stages. The direct effect of IGF1 on proliferation may provide a mechanism to increase the size of bioengineered teeth. Dental mesenchymal cells begin to differentiate into odontoblasts by signallings from the enamel knot in the bell stage 46 ; dental epithelial cells begin to differentiate into ameloblasts following odontoblast differentiation 46 . During the late bell stage, differentiated odontoblasts and ameloblasts form dentin and enamel, which are dental hard tissues, and teeth develop with increasing size 38 . Thus, we analysed the effect of IGF1 on odontoblast and ameloblast differentiation of dental mesenchymal and epithelial cells isolated from ED14.5 mouse molar tooth germs, in order to further elucidate the mechanism for the increased size of bioengineered teeth.
In the present study, we showed that expression of odontoblast markers Dmp1, Dspp, and Map1b was not induced by IGF1 in dental mesenchymal cells. Expression of these markers was also not induced by BMP2. In contrast, simultaneous administration of IGF1 and BMP2 resulted in significant induction of these marker genes. Bègue-Kirn et al. 47 also showed that the combination of IGF1 and BMP2 promotes odontoblast differentiation by histological analysis of organ-cultured dental papilla tissue isolated from ED17 mouse tooth germs; IGF1 or BMP2 alone does not promote this differentiation. These findings suggest that BMP2 is essential for IGF1-induced differentiation of dental mesenchymal cells into odontoblasts.
We then analysed expression of IGFBPs, which regulate the biological action of IGF1, to elucidate the mechanism by which simultaneous administration of IGF1 and BMP2 induces differentiation of dental mesenchymal cells into odontoblasts. We found decreased expression of Igfbp2, 3, and 4; notably, expression of Igfbp6 was increased by BMP2, with or without IGF1 administration. Takahashi et al. 28 showed that BMP2 decreased the expression of IGFBP3 in neonatal bovine chondrocytes. These findings suggest that BMP2 is a modulator of IGFBP expression. IGFBP2, 3, and 4 inhibit the bioactivity of IGF1 by interfering with IGF1 binding to its receptor 30 . Thus, activation of IGF1 signalling due to reduced expression of IGFBP2, 3, and 4 by BMP2 constitutes the mechanism by which simultaneous administration of IGF1 and BMP2 induces differentiation of dental mesenchymal cells into odontoblasts. Additionally, IGFBP6 has high binding affinity for IGF2 and suppresses the activity of IGF2 30 . However, IGFBP6 has low binding affinity for IGF1 and its effect on IGF1 activity remains uncleared 30 . Moreover, although IGFBP6 suppresses the differentiation of oligodendrocyte precursor cells, and osteoblasts 48,49 , its role in the differentiation of dental cells has not been reported. Further studies are necessary to clarify the role of BMP2-induced IGFBP6 in odontoblast differentiation.
In the present study, we showed that the expression of the enamel matrix gene Ambn was enhanced upon IGF1 treatment of cultured dental epithelial cells. Takahashi et al. 50 showed that IGF1 increased the expression of Ambn in organ-cultured ED15 mouse molar tooth germs. However, it is not clear whether IGF1 directly increases the expression of Ambn in dental epithelial cells, because tooth development progresses through various signalling crosstalk mechanisms between epithelial and mesenchymal tissues. The present study showed that the expression of Ambn was directly enhanced by IGF1 in dental epithelial cells. Ambn is expressed earlier than other enamel matrix genes, such as Enam and Amel, during tooth germ development 51 . Because Ambn-null mice exhibited severe enamel hypoplasia and reduced expression of Amel, Ambn may play an essential role in normal ameloblast differentiation 20 . Therefore, IGF1-induced Ambn would induce the ameloblasts differentiation during the development of bioengineered tooth germs.
IGF1 plays a role in increasing the size of various organs, such as the kidneys, liver, spleen, and bones, during postnatal growth of rats and mice 52,53 . Therefore, our findings contribute to a novel strategy for the morphNological control of bioengineered organs (not strictly limited to teeth) by the activation of IGF1 signalling.

Methods
Ethics statement. C57BL/6 mice were purchased from CLEA Japan (Tokyo, Japan). Animal experiments were performed in accordance with the Regulations for Animal Experiments and Related Activities at Tohoku University. All animal protocols were approved by the Institutional Animal Care and Use Committee of the Tohoku University Environmental and Safety Committee. Transplantation of bioengineered tooth germs into subrenal capsules. Bioengineered tooth germs isolated from collagen gel after 7-day-organ culture were placed into plastic ring-shaped devices with 2.5-mm internal diameter and 1.3-mm thickness. These were transplanted into subrenal capsules of 7-week-old C57BL/6 mice under general anaesthesia, as described previously 54 .

Generation of bioengineered tooth germs from single cells. Bioengineered
Transplantation of tooth germs into jawbones. Transplantation of bioengineered tooth germs into jawbones was performed as described previously 3 . Upper first molars of 5-week-old C57BL/6 mice were extracted under general anaesthesia and maintained for 3 weeks to allow natural repair of the tooth cavity and oral epithelium. Then, the oral mucosa at the extraction sites was incised and bony holes (0.8 mm in diameter and 1.5 mm in depth, one hole per site) were made by a dental micro motor (PAL; Morita, Osaka, Japan). Bioengineered tooth germs were isolated from collagen gel after 7-day-organ culture, marked at the top of the dental epithelium with methylene blue to confirm the direction, and transplanted into the bony holes. The incised oral mucosa was sutured with 8-0 nylon (BEAR Medic, Chiba, Japan). To analyse the expression of NF and VWF, deparaffinized sections were treated with 0.15% Triton-X100 in PBS for 30 min at room temperature. After sections were blocked with 1% goat serum (Vector Laboratories) in PBS for 1 h at room temperature, they were incubated with anti-NF (1:200; Chemicon, Temecula, CA, USA) or anti-VWF (1:1000; Dako, Glostrup, Denmark) antibodies for 2 h at room temperature. After sections were washed with 0.1% Tween 20 in PBS, they were treated with goat anti-rabbit IgG Alexa 488 (1:1000; Invitrogen) for 1 h at room temperature. Nuclei were stained with 4' ,6-diamidino-2-phenylindole (DAPI; 1:1000; KPL, Gaithersburg, MD, USA). Samples were visualized using a confocal laser scanning microscope system (C2si; Nikon Instech, Tokyo, Japan).

Micro
Immunocytochemistry. Epithelial cells isolated from the mandibular first molar tooth germs of ED14.5 C57BL/6 mice were seeded at a density of 1.5 × 10 5 cells/cm 2 on glass-bottomed dishes coated with 20 µg/mL fibronectin (Sigma) and 100 µg/mL type IV collagen (Nitta Gelatin) and cultured in CnT-Prime medium (CELLnTEC, Bern, Switzerland). On the next day, 500 ng/mL IGF1 was added into the medium. After 4-day culture, cells were fixed using 4% PFA for 15 min and treated with 0.1% Triton-X100 in PBS for 15 min at room temperature. After cells were blocked with 3% bovine serum albumin (Sigma) in PBS for 1 h at room temperature, they were incubated with an anti-AMBN antibody (1:100; Santa Cruz Biotechnology Inc.) overnight at 4 °C. Cells were treated with goat anti-rabbit IgG Alexa 488 (1:1000; Invitrogen) for 1 h at room temperature. Nuclei were stained with DAPI (1:1000; KPL). Signals were visualized using C2si (Nikon Instech). Integrated density measurements of five microscopic fields in each sample were performed using ImageJ software (https://imagej.nih.gov/ij/). Probe preparation and whole-mount in situ hybridization. Digoxigenin-labelled RNA probes were prepared with the MAXIscript Kit (Thermo Fisher Scientific, Waltham, MA, USA), in accordance with the manufacturer's instructions. The DNA fragment of Fgf4 (NM_010202, located between 117 and 731) was amplified by polymerase chain reaction (PCR), using primers shown in Supplementary Table S1. The fragment was subcloned into the pGEM-T vector (Promega, Madison, WI, USA) and used to generate sense and antisense probes. Bioengineered tooth germs after 5-day organ culture were fixed in 4% PFA overnight at 4 °C, treated with 6% H 2 O 2 and 0.1% Tween 20 in PBS at room temperature, and incubated with 10 µg/mL proteinase K (Roche, Mannheim, Germany) for 3 min at 30 °C. After post-fixation and prehybridization, samples were hybridized overnight at 70 °C with digoxigenin-labelled RNA probes. Samples were blocked with 10% normal sheep serum (Sigma) and then incubated with anti-digoxigenin antibody conjugated with alkaline phosphatase (Roche), overnight at 4 °C. Nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate (Roche) were used for signal detection.
BrdU assay. Mesenchymal and epithelial cells were isolated from the mandibular molar tooth germs of E14.5 C57BL/6 mice. Mesenchymal cells were seeded on glass-bottomed dishes at a density of 2.0 × 10 4 cells/cm 2 and cultured in DMEM supplemented with 10% FBS, 100 units/mL of penicillin, and 100 µg/mL of streptomycin. On the next day, IGF1 (500 ng/mL) was added into the medium. Epithelial cells were seeded at a density of 7.5 × 10 4 cells/cm 2 on glass-bottomed dishes coated with 20 µg/mL fibronectin (Sigma) and 100 µg/mL type IV collagen (Nitta Gelatin); the cells were cultured in CnT-Prime medium (CELLnTEC). After 3-day culture of mesenchymal cells and 2-day culture of epithelial cells, cells were incubated with DMEM supplemented with 1% BrdU Labeling Reagent (Invitrogen) for 4 h at 37 °C, then fixed with 70% ethanol for 15 min at 4 °C. The cells were treated with 0.1% Triton-X100 in PBS for 20 min at room temperature, then with Denaturing Solution (BrdU Staining Kit, Invitrogen) for 30 min at room temperature. Cells were treated with 0.1% Triton-X100 and 5% normal goat serum in PBS for 1 h at room temperature and incubated with Alexa Fluor 488-conjugate BrdU antibody (1:100; Thermo Fisher Scientific) overnight at room temperature. Nuclei were stained with DAPI (1:1000; KPL). Samples were visualized using C2si (Nikon Instech). Counts of total and BrdU-positive cell numbers were performed using five microscopic fields in each sample.

PCR.
Total RNA was isolated from whole mandibular first molar tooth germ, dental mesenchymal and epithelial tissues, and mesenchymal and epithelial cells isolated from tooth germs in E14.5 mice, using the RNeasy Kit (Qiagen, Valencia, CA, USA). cDNA was synthesized from 0.4 µg total RNA using the PrimeScript RT Reagent Kit  Supplementary Table S2.
Statistical analysis. Data were statistically analysed by Student's t-test to compare differences between two groups. For more than two groups, we used an analysis of variance (ANOVA) followed by the Tukey-Kramer test. P values < 0.05 was considered statistically significant.