|Home | About | Journals | Submit | Contact Us | Français|
The anterior eye chamber and the kidney capsule of the mouse have been traditionally used for long-term culture of tooth germ grafts. However, although these sites provide an excellent growth environment, they do not represent real in situ sites for the development of a grafted tooth germ. Here, we describe a protocol to transplant a tooth germ into the mandibular diastema region of mouse embryos using exo utero surgery. Our results demonstrate that the mouse embryonic diastema region represents a normal physiological environment for the development of transplanted tooth germs. Transplanted tooth germs developed synchronically with and became indistinguishable from the endogenous ones. These ectopic teeth were vascularized and surrounded with nerve fibers, and were able to erupt normally. Thus, the exo utero transplantation approach will provide a new avenue to study tooth development and regeneration.
Given the relatively concise morphology and the ease with which it can be accessed, manipulated, cultured, and grafted ex vivo, the mouse tooth germ has long been an excellent model to study vertebrate organogenesis (Thesleff et al., 1996). Several approaches have been developed to study tooth development ex vivo. The in vitro organ culture has been routinely used, in which tooth germs are cultured in the Trowell type organ culture dish (Thesleff et al., 1996). This method provides an easy and quick culture for short-term growth of tooth germ and also provides an openly accessible environment for the manipulation of tooth germ such as addition of growth factors. However, the artificial culture condition limits not only the length of culture time but also the shape and size of explanted tooth germs. The other method that has been used extensively for long-term tooth germ culture is the kidney capsule culture in which tooth germs or organ primordial tissue are grafted and cultured underneath the kidney capsule of adult mice (Morio, 1985). Grafted tooth germs can develop into a well-differentiated tooth (Zhang et al., 2003b). Actually, many parts of mouse body have been tested for hosting grafted tooth germs. These sites include the anterior eye chamber (Kollar and Fisher, 1980; Yoshikawa and Kollar, 1981; Lumsden and Buchanan, 1986), renal subcapsular site (Bartlett and Reade, 1973), subcutaneous tissue (Ivanyi, 1965), cheek pouch (Al-Talabani and Smith, 1979), and spleen (Ishizeki et al., 1987). In addition, chick embryonic wing buds were also shown to be an effective in vivo culture system for tooth development (Koyama et al., 2002). However, even though some of these models provide an excellent growth environment for tooth grafts, they represent a developmental and physiological environment that is distinct from the oral cavity where tooth development normally occurs.
The best and ideal site for ectopic tooth growth is within the oral cavity. The oral cavity of the adult mouse has been successfully used to support the development of grafted tooth germ (Ohazama et al., 2004; Nakao et al., 2007). However, the mouse after birth is full of activity like daily food intake, which may influence and exert mechanical pressure on the normal development of grafted tooth germ. It would be a better choice to nurture a tooth germ in the oral cavity of the mouse at a comparable developmental stage. In mice, a toothless region called diastema exists between the incisor and the first molar. This region is believed to be filled with incisor, canine, and premolars before their disappearance during the rodent evolution (Luckett, 1985; Peterkova et al., 1995). Today, the rudimentary tooth germs still can be observed in the upper diastema of the mouse or sibling vole embryos, and then these tooth germs are removed apoptotically through an unknown mechanism (Keranen et al., 1999). On the other hand, the initiation of tooth development in the mandibular diastema region is evidenced by a weak expression of Shh just anterior to the first molar primordia (Mustonen et al., 2004). Accordingly, disturbance of the normal gene regulation network in this area may results in an extra tooth (Mustonen et al., 2003; Zhang et al., 2003a; Tucker et al., 2004; Kassai et al., 2005). Based on these observations, we reasoned that the mandibular diastema region could be an ideal site for the development of a tooth germ in addition to the endogenous tooth forming loci. To test this hypothesis, we carried out an exo utero surgery to transplant tooth germ. Here, we describe this method and demonstrate that transplanted tooth germs develop normally in the mouse embryonic diastema region, indistinguishable from the endogenous ones.
One of the greatest advantages of using avian embryo as a developmental biology model is the accessibility of its embryos throughout the entire developmental process. In the mouse, however, accessing to the developing embryo is always difficult and poses a great challenge due to the complex architecture of the mouse gestation organ. The finding that mouse embryos can develop normally outside the uterine myometrium has substantially extended the accessibility of mouse embryo to a much earlier stage, and moreover, such an exo utero mouse embryo can survive some surgical manipulations to term (Muneoka et al., 1986a,b). Benefiting from these findings, exo utero surgery has evolved to become a powerful tool to study mouse embryonic development (Ngo-Muller and Muneoka, 2000). It has been widely used in studies of nerve system development such as migration of the neural crest cells during early mouse embryogenesis (Serbedzija et al., 1992), limb development and regeneration (Han et al., 2003), as well as in teratology research where effects of chemicals on the developmental mouse embryos can be studied. A recent review has summarized the applications of mouse exo utero development system in biological studies (Hatta et al., 2004).
We were originally seeking to identify an ideal site for the best development and growth of ectopically grafted tooth germ or recombinant tooth germ after genetic modification in vitro (Song et al., 2006). Compared with the kidney capsule, the anterior eye chamber, or other ectopic loci, the oral cavity of the mouse appears to represent the most natural environment for tooth development. The diastema, a toothless region between the incisor and molar, has been demonstrated to host a full array of teeth before they were evolutionally lost. Teeth indeed develop in the mandibular diastema of some genetically modified mice (Mustonen et al., 2003; Tucker et al., 2004; Zhang et al., 2003a; Kassai et al., 2005). We, therefore, deduced that the mandibular diastema of the mouse embryo had an intrinsic ability to support tooth development.
Although exo utero surgery can be carried out on mouse embryos as early as embryonic day (E) 11.5, a better survival rate is achieved with surgery on E13.5 or older embryos (Muneoka et al., 1986a). On the other hand, tooth germs of E13.5 mouse embryos are at the bud stage, which provides an ideal size and solidity for surgical operation compared with that from earlier stages. Thus, we began with transplanting E13.5 mouse tooth germs into the mandibular diastema of E13.5 host embryos. Such an arrangement allows the endogenous tooth primordia to be the control, which provides standard developmental indexes side by side for the evaluation of the ectopic tooth growth. Here, we further took advantage of tooth germs from Rosa26-Egfp mouse embryos, of which tissue can be easily detected and distinguished from the host embryo under the fluorescent microscope by activation through fluorescein isothiocyanate (FITC) -filter set.
For exo utero surgery, the incisor tooth germ was isolated from the E13.5 Rosa26-Gfp mouse embryo. At this stage, the incisor primordia is much better for transplantation experiments due to their smaller size compared with the molar tooth germs. It can be easily grafted into a small incision made in the diastema region of host embryos. As described in detail in the Experimental Procedures section, the same stage embryos were properly exposed from the uterus of an anesthetized female mouse. As shown in Figure 1A, an E13.5 host embryo was exposed after opening the uterus and removing the extraembryonic membrane in a timed-pregnant mouse. Under adequate illumination, details of the embryo can be easily defined under the dissection microscope. Before grafting into the small incision made in the diastema region, the isolated incisor germ was carried on a thin copper wire to the embryonic mandible. The tooth germ was then carefully placed into the incision with the epithelial side face up. By piercing the neighboring host tissue with the copper wire, the transplanted tooth germ was fixed at the incision site. Figure 1B shows a transplanted Rosa26-Egfp incisor primordia fixed at the mandibular diastema region of an E13.5 wild-type host embryo. Under the activation light through the FITC-filter set, the ectopic tooth germ was emanating green fluorescent light.
Histological analysis of a full-term host embryo shows a well-developed tooth germ in the transplanted site (Fig. 2A). Both epithelial and mesenchymal structures of the transplanted incisor are comparable to that of the endogenous incisors (Fig. 2B). To confirm the supportive role of host tissue in the ectopic tooth development, we further examined blood vessel invasion and innervation of the transplanted tooth. By immunohistochemical staining using antibody against laminin, a molecular marker for blood vessel endothelial cells, we were able to detect blood vessels in the dental pulp of both the endogenous and the transplanted teeth (Fig. 2C,D), clearly demonstrating the presence of blood vessels in the transplanted tooth. The invasion of blood vessels provides the transplanted tooth with nutrients through the blood supplies from the host. Moreover, using antibodies against neurofilament, we have detected nervous termini around the transplanted tooth, indicating that the transplanted tooth will have the similar potential as the endogenous teeth to be innervated (Fig. 2E,F). Thus, the transplanted tooth germ develops synchronically with the endogenous tooth in the host embryo. Based on the facts that blood vessels are not present in the developing mouse tooth before the cap stage (E14.5; Luukko et al., 2005) and that tooth innervation does not occur until birth (Tsuzuki and Kitamura, 1991; Kettunen et al., 2005), we conclude that these blood vessel cells and neuron cells within or surrounding the transplanted tooth are derived from the host.
Of interest, when E13.5 embryos were used as hosts for tooth transplantation, all the hosts (21 in total) that were monitored died within 1 day after caesarian section. Examination of gross morphology identified the cleft palate phenotype in all the surgical embryos, explaining the postnatal lethality. It is known that the palate shelves elevated from a perpendicular to the horizon position and fused to each other above the tongue between E13.5 and E14.5. Therefore, our surgical operation in the oral cavity of E13.5 embryo must have disturbed such an elevation and/or fusion process of the palate shelves, resulting in the cleft palate formation.
Because the palate shelves have already fused in E14.5 mouse embryos, we decided to use E14.5 mouse embryos as host for tooth transplantation, in hope of getting living pups with an ectopic tooth in their mouth. As expected, none of the pups (25 in total) that received tooth germ transplantation at E14.5 developed a cleft palate phenotype. The mice were fed by a foster mother immediately after C-section, and grew up as normal mice. Figure 3 shows such a weaned mouse carrying an ectopic incisor growth in the mandibular diastema. These results clearly demonstrate that the transplanted tooth germ can develop normally in the diastema region of mouse embryo and can eventually erupt and form as a normal tooth in the host mouse oral cavity.
The mechanism of how some vertebrates lost their teeth during evolution remains elusive. It could be a result of loss of gene function that is critical for tooth development. In avian, the remnant of developing tooth primordia can be found in the oral cavity too, and the abortion of tooth development is at least partially due to the loss of odontogenic Bmp4 expression (Chen et al., 2000). On the other hand, tooth agenesis could also be a result of an active inhibitory mechanism that is present in the oral cavity to repress tooth development. It is evidenced that Gas1, a diffusible protein acting as Shh antagonist, is able to sequester Shh and, therefore, to block Shh function in the mouse diastema region. Such an antagonistic mechanism may be used to delineate the border of tooth forming sites and the diastema region at the beginning of tooth development, or required to constantly inhibit tooth development within the diastema (Cobourne et al., 2004). Nevertheless, we show here that, at or after E13.5, the mandibular diastema of mouse embryo serves as an excellent site to support tooth development. Our results suggest that the grafted tooth germ is able to override any inhibitory mechanism or can develop independent of the inhibitory mechanism. Alternatively, the constant inhibitory mechanism discussed above may have disappeared in the diastema region at the time when tooth germ is grafted.
Similar to many other ectopic sites in living animals, particularly the mouse kidney capsule, the embryonic diastema region appears to be excellent for grafted tooth development and differentiation. However, the approach we describe here also has several advantages, such as (1) it makes it possible to study tooth eruption of embryonic lethal mutant mice, and (2) it permits studies on tissue interactions between grafted tooth germ and the host supportive tissues in the oral cavity. On the other hand, this approach has its limitation as well. The size of a graft is a concern. For example, an E13.5 molar germ was proven too big to be grafted properly. In conclusion, while the diastema region of the mouse mandible does not grow endogenous teeth, it does fully support tooth formation when ectopic tooth germ is grafted.
Tooth germs at the bud stage were used as donors in this study. Embryos at E13.5 were collected from timed pregnant CD-1 or Rosa26-Egfp mice. The incisor tooth germs were isolated from the mandibles and kept in ice-cold PBS for use. A fine copper pin was used to pierce through each tooth germ at the mesenchyme, making it easier to transfer the tooth germ from the dish to the surgical embryo with forceps. Exo utero surgery was performed as described previously (Ngo-Muller and Muneoka, 2000). In this study, we used E13.5 and E14.5 embryos as recipients for tissue transplantation. Briefly, timed pregnant CD-1 mice at 13.5 or 14.5 gestation were anesthetized by intraperitoneal injection of Nembutal at a dose of 10 μl/g body weight. After properly anesthetized, the mouse was placed on a surgical platform. The skin and body wall of the abdomen were sequentially opened by scissors and forceps. The uterus was subsequently opened with iridectomy scissors at the opposite of placental side in a continuous longitudinal incision. By rolling a dry cotton-tipped applicator between the placenta and the uterus, the unwanted embryos were removed with minimal bleeding from the uterus. Usually, four embryos were left in the abdomen of each female for grafting. Blood and debris were washed away with lactated Ringer’s solution before further operation. Extra care needs to be taken to avoid damaging the major blood vessels in the extraembryonic membrane when an incision was made in the membrane. The head of the host embryo was allowed to pop out through the incision on the extraembryonic membrane, and then was properly positioned with the help of sterile cotton balls as cushions. A small slit was made by fine forceps within the diastema region of the host’s mandible. Donor tooth germ was inserted into the incision with the dental epithelial side facing the oral cavity, and then was fixed by fine copper pins. After completion of tissue transplantation, the extraembryonic membrane was closed by suturing. The abdomen cavity was then sutured after cotton balls were removed and the abdomen cavity was totally cleaned with lactated Ringer’s solution. Typically, the mouse wakes up within 1 hr after surgery. A warm heating pad was used to help the mouse maintain body temperature during this process.
The pregnant female mouse was housed and closely monitored in an isolated cage until term. To collect experimental embryos, the host female mouse was killed, and surgical pups were collected through C-section. For further development purposes, the operated embryos were stimulated to begin breathing by rubbing cotton-tipped applicator on their tail tips. Any fluids in their mouth were also cleaned with a cotton-tipped applicator. When they were breathing regularly and moving about, the experimental newborns were transferred to the cage with a nursing mother that had delivered her litter at the same day or 1 day before. An animal protocol for the exo utero surgery was approved by the IACUC of Tulane University. The entire surgery procedure strictly followed the animal protocol.
The recipient mice were killed at birth. The heads were harvested and fixed in 4% paraformaldehyde (PFA)/ PBS, de-calcified by demineralization in 0.1 M ethylenediaminetetraacetic acid/PBS for a week with several changes of solution. Samples were then dehydrated through graded ethanol, cleared with xylene, embedded in paraffin, and sectioned at 10 μm. Sections were processed with Azan di-chromic staining, as described previously (Presnell and Schreibman, 1997).
The recipient mice were collected at the delivery day, and the heads were removed and processed for frozen sections. OCT-embedded samples were sectioned at 10 μm, and the cryostat sections were fixed for 30 min at room temperature with 4% PFA and analyzed by indirect immunohistochemical staining. The rabbit anti-mouse laminin antibody (Chemicon, Temecula, CA) was used as the primary antibody specific for blood vessel endothelial cells, and the rabbit anti-mouse neurofilament 200 antibody (Sigma, St. Louis, MO) was used as the primary antibody for nerve cells. The horseradish peroxidase-coupled goat anti-rabbit IgG were used as the secondary antibody (Sigma). Immunohistochemical staining was performed according to the protocol provided with the antibodies by the manufacturers.
Y.P.C. and K.M. were supported by the NIH.
Grant sponsor: the NIH; Grant number: DE15123; Grant number: DE16623; Grant number: HD43277.