Research ArticleStem Cells

Deciduous autologous tooth stem cells regenerate dental pulp after implantation into injured teeth

See allHide authors and affiliations

Science Translational Medicine  22 Aug 2018:
Vol. 10, Issue 455, eaaf3227
DOI: 10.1126/scitranslmed.aaf3227

Tooth stem cells regenerate smiles

Dental pulp necrosis is one of the most common pathological conditions that results in tooth loss. However, regeneration of functional dental pulp has proved difficult. In a new study, Xuan et al. implanted ex vivo expanded autologous tooth stem cells from deciduous teeth in two animal models and in human patients. They demonstrated regeneration of dental pulp containing an odontoblast layer, blood vessels, and nerves in the implanted teeth and rescue of sensation to stimuli such as temperature. This work suggests that implantation of tooth stem cells can provide partial recovery of teeth injured by trauma.

Abstract

Pulp necrosis arrests root development in injured immature permanent teeth, which may result in tooth loss. However, dental pulp regeneration and promotion of root development remains challenging. We show that implantation of autologous tooth stem cells from deciduous teeth regenerated dental pulp with an odontoblast layer, blood vessels, and nerves in two animal models. These results prompted us to enroll 40 patients with pulp necrosis after traumatic dental injuries in a randomized, controlled clinical trial. We randomly allocated 30 patients to the human deciduous pulp stem cell (hDPSC) implantation group and 10 patients to the group receiving traditional apexification treatment. Four patients were excluded from the implantation group due to loss at follow-up (three patients) and retrauma of the treated tooth (one patient). We examined 26 patients (26 teeth) after hDPSC implantation and 10 patients (10 teeth) after apexification treatment. hDPSC implantation, but not apexification treatment, led to regeneration of three-dimensional pulp tissue equipped with blood vessels and sensory nerves at 12 months after treatment. hDPSC implantation increased the length of the root (P < 0.0001) and reduced the width of the apical foramen (P < 0.0001) compared to the apexification group. In addition, hDPSC implantation led to regeneration of dental pulp tissue containing sensory nerves. To evaluate the safety of hDPSC implantation, we followed 20 patients implanted with hDPSCs for 24 months and did not observe any adverse events. Our study suggests that hDPSCs are able to regenerate whole dental pulp and may be useful for treating tooth injuries due to trauma.

INTRODUCTION

Regeneration is defined as regrowth of lost or destroyed parts of tissues or organs (1). Regeneration of the heart or liver is through regrowth or repair of injured tissues (2, 3), but regeneration of whole organs remains a major challenge. Our previous study reported that exfoliated deciduous teeth contain multipotent mesenchymal stem cells (MSCs) (4) capable of differentiating into a variety of cell types including neural cells, odontoblasts, stromal cells, and adipocytes. Although the detailed biology of stem cells in exfoliated teeth remains unclear, these cells are able to induce bone formation, generate dentin, and survive in mouse tissues after implantation (4, 5).

Teeth are an ectodermal organ that is derived from sequential reciprocal interactions between oral epithelial cells and cranial neural crest–derived mesenchymal cells. Teeth are unique and complex, containing both hard tissue (dentin and enamel) and soft tissue (pulp). Dental pulp contains vasculature, nerves, and connective tissue to ensure the viability of teeth (6). Nerves in the pulp tissue can mediate pain sensations, regulate blood flow, recruit immunocompetent cells, and serve as an MSC niche (712). Loss of dental pulp terminates the root development of developing permanent teeth, which may destroy the periodontal attachment and results in the loss of teeth. Recent studies have shown that cell-based therapies are capable of regenerating vascular dental pulp tissue in animal models (1316). There is great clinical demand to regenerate pulp tissue for treating a variety of infectious, necrotic, and traumatic dental diseases (1720). For example, a total of 403,149 emergency department visits in the United States during the year 2006 had a primary diagnosis of dental pulp disease (17).

Trauma to immature permanent teeth is relatively common in school-aged children and can result in loss of vital dental pulp, loss of the blood and nerve supply, and impaired root development (21). Although apexification (a procedure that induces tooth root development and closure of the root apex through hard tissue deposition) is the current standard clinical treatment for traumatically injured immature permanent teeth, it fails to restore lost pulp tissue and maintain normal root development (22). Here, we use young permanent teeth that have undergone trauma as a model to assess whether implantation of autologous human deciduous pulp stem cells (hDPSCs) could regenerate lost three-dimensional (3D) pulp tissue, restore pulp function, and promote root development.

RESULTS

Characteristics of hDPSCs obtained from deciduous teeth

We expanded hDPSCs from deciduous teeth of two patients and found no difference in the number of single-colony clusters [colony-forming unit fibroblasts (CFU-F)] (table S1), osteogenesis, and adipogenesis between the two groups (Fig. 1, A to C). hDPSCs formed mineralized nodules after osteogenic induction for 28 days (Fig. 1B). Oil red O–positive cells were also observed after adipogenic induction for 21 days (Fig. 1C). Immunostaining showed that the two groups of hDPSCs were positive for binding of antibodies against CD146 and CD105 and were negative for CD34 antibody staining (Fig. 1D). In addition, the hDPSCs expressed the neuronal markers NeuN and nestin (Fig. 1D), suggesting that hDPSCs may be neural crest–derived MSCs. hDPSCs also expressed the sensory neuron markers calcitonin gene-related peptide (CGRP), transient receptor potential cation channel subfamily M member 8 (TRPM8), and transient receptor potential cation channel subfamily V member 1 (TRPV1) (Fig. 1D).

Fig. 1 hDPSCs regenerate dental pulp in immunocompromised mice.

(A) hDPSCs derived from deciduous canine tooth pulp from two patients (sample 1 and sample 2) formed single CFU clusters in culture. (B) hDPSCs derived from the teeth of two patients formed mineralized nodules when induced in osteogenic culture medium for 28 days. Uninduced hDPSCs failed to form mineralized nodules. Scale bar, 200 μm. (C) hDPSCs derived from the teeth of two patients were able to differentiate into Oil red O–positive adipocytes when cultured under adipogenic induction conditions for 21 days. Uninduced hDPSCs failed to form adipocytes. Scale bar, 50 μm. (D) Immunofluorescence staining showed that hDPSCs expressed CD146, CD105, NeuN, nestin, CGRP, TRPM8, and TRPV1, but not CD34. Scale bar, 20 μm. (E) hDPSC aggregates inserted into empty root canals of human teeth and implanted subcutaneously into immunocompromised mice (n = 12) for 8 weeks were stained with hematoxylin and eosin (H&E) and Masson stain. Implanted hDPSC aggregates regenerated pulp tissue. In the control group, calcium hydroxide was inserted into empty root canals of human teeth and implanted subcutaneously into immunocompromised mice for 8 weeks. After 8 weeks, no pulp tissue was regenerated and only calcified tissue was observed. Scale bar, 50 μm. Enlarged regions of the images show that odontoblasts (black arrows) were present at the margin of the regenerated pulp tissue. Blood vessels (open arrows) were also observed in the regenerated pulp tissue. Scale bar, 20 μm. (F) Left: Image shows dentin sialoprotein–positive odontoblasts (open black arrows) revealed by immunohistochemical staining. Scale bar, 20 μm. Right: Calcein staining showed newly formed dentin (white arrows) in the empty root canal of a human tooth. Scale bar, 0.1 mm.

We placed hDPSC aggregates derived from deciduous teeth of two patients into empty human tooth roots and subsequently implanted them into immunocompromised mice subcutaneously for 8 weeks. Pulp-like structures containing blood vessels and an odontoblast-like layer were observed (Fig. 1E). When calcium hydroxide was inserted into empty human tooth roots as a control and implanted in the same manner, no pulp-like structures were observed at 8 weeks after implantation in this control group (Fig. 1E). We performed immunohistochemistry to show that hDPSC-regenerated pulp tissue bound antibody against dentin sialoprotein, suggesting that pulp tissue contains odontoblasts (Fig. 1F). In addition, we used calcein labeling to show that hDPSCs formed new dentin (Fig. 1F).

hDPSCs differentiate into sensory neurons in rat dorsal root ganglia

Given that functional pulp tissue contains nerves, we wanted to know whether hDPSCs could differentiate into neural cells. We injected hDPSCs into rat dorsal root ganglia in vivo. The results showed that hDPSCs stained for human mitochondria exhibited the morphology of neurons and expressed the TRPV1 channel (sensing heat) and the TRPM8 channel (sensing cold) at 2 months after injection (fig. S1). These results imply that hDPSCs may be capable of differentiating into sensory neurons in vivo.

Pig DPSCs regenerate 3D dental pulp when implanted into minipigs

To confirm that pig DPSC aggregates could regenerate 3D pulp tissue, we isolated DPSCs from minipigs and implanted DPSC aggregates into empty root canals in minipigs in vivo (fig. S2 and table S5). Normal pulp tissue from minipigs was stained for comparison (Fig. 2A). For the control group, calcium hydroxide premixed with iodoform paste was placed into empty root canals in minipigs. After 3 months, the teeth were harvested, and histological analysis showed that implanted pig DPSC aggregates regenerated pulp tissue containing an odontoblast layer and blood vessels (Fig. 2A). However, only calcium hydroxide was observed and no pulp tissue was regenerated in the root canals in the control group (Fig. 2A). Moreover, we observed whole pulp tissue regeneration in minipigs after implantation of pig DPSCs (Fig. 2C). We selected two regions and found overlapping NeuN and 4′,6-diamidino-2-phenylindole (DAPI) staining in regenerated pulp (Fig. 2C). However, normal control dental pulp from minipigs failed to express NeuN (Fig. 2B). These results indicated that pig DPSCs were capable of regenerating functional dental pulp with blood vessels and nerves in a large preclinical animal model (Fig. 2).

Fig. 2 Histological analysis of pig DPSCs implanted into minipigs.

(A) Pig DPSCs (pDPSCs) were implanted into permanent incisors of minipigs after pulpectomy (n = 3). H&E staining (left) and Masson staining (right) showed that pulp tissue was regenerated 3 months after pDPSC implantation. In the control group, calcium hydroxide instead of pDPSCs was inserted into young permanent incisors in minipigs (n = 3). After 3 months, no pulp tissue was regenerated and only calcium hydroxide was observed. Normal pulp tissue of minipigs was stained for comparison (top). Scale bar, 50 μm. Enlarged images show odontoblasts (black arrow) and blood vessels (open arrow) in select regions of regenerated pulp tissue. Scale bar, 20 μm. (B) Representative histological images showing that pig normal pulp tissue lacks NeuN-positive cells. Scale bar, 100 μm. (C) Pig DPSCs regenerated dental pulp that contained NeuN-positive cells (green); DAPI (blue) was used as a counterstain for nuclei. Scale bar, 200 μm. Two regions of regenerated pulp tissue were selected for higher magnification. Scale bar, 20 μm.

hDPSC implantation regenerates whole dental pulp in patients with tooth trauma

In light of the preclinical results, we conducted a clinical trial between June 2013 and December 2014. We screened 62 participants (each with one traumatized incisor tooth) and randomly allocated 40 patients ages 7 to 12 years to the hDPSC treatment group (30) or the control apexification group (10) in a 3:1 ratio (Fig. 3). Thirty teeth from 30 patients received an autologous hDPSC implant and 10 teeth from 10 control patients received traditional apexification treatment. However, three patients allocated to the hDPSC group were excluded due to loss of follow-up and another was excluded due to additional trauma to the hDPSC-treated tooth. Ultimately, 36 teeth from 36 patients were eligible for a 12-month follow-up, including 26 teeth that had received the hDPSC implant and 10 teeth that received traditional apexification. In the stem cell treatment group, two hDPSC aggregates containing a total of 1 × 108 cells were implanted into each traumatized permanent incisor (fig. S3) and were monitored for 12 months. No complications were identified during the 24 hours after pulp harvesting and hDPSC implantation.

Fig. 3 Clinical trial profile and study timeline.

(A) Design of clinical trial to examine dental pulp regeneration by autologous hDPSCs. (B) Study events and timeline of procedures and testing are shown for participants randomly allocated to the hDPSC implantation group or control group. At the first visit, dental pulp from deciduous teeth was removed and then cultured to obtain hDPSCs for implantation. All participants received radiovisiography (RVG), CBCT, continuous-wave Doppler, and root canal disinfection at the first visit. One month later, the hDPSC implantation group received hDPSCs and the control group received apexification treatment. All participants underwent RVG, CBCT, electric pulp vitality testing, and laser Doppler flowmetry at 6 and 12 months after treatment. Data from the electric pulp vitality test and the laser Doppler flowmetry test were compared between the hDPSC implantation and control groups. *Four patients in the hDPSC implantation group were excluded from analysis.

A total of 36 (90%) of 40 randomly allocated patients were enrolled at the Fourth Military Medical University (FMMU). Four (10%) of the 40 patients came to our hospital within 1 month of dental trauma. Twenty-two (55%) of the 40 patients came to our hospital 1 to 6 months after trauma, and 14 (35%) of the 40 patients came to our hospital 6 months after trauma (Table 1). All of the patients tested negative on the electric pulp vitality test. The electric pulp test reflects the pulpal sensory threshold of the tooth; a lower threshold indicates greater sensation.

Table 1 Baseline characteristics of patients.
View this table:

Laser Doppler flowmetry showed the average blood cell flux (cells per meter squared per second) as a linear readout in relative perfusion units (PU). Relative PU readouts have been shown to correlate linearly with blood flow values. Vascular formation tests yielded an average of 2.81 ± 0.41 PU for the hDPSC-treated group and 3.01 ± 0.44 PU for the control group, respectively, at baseline (Table 1 and table S2). The mean length of the affected tooth roots at baseline was 10.69 ± 1.32 mm in the hDPSC treatment group and 9.99 ± 0.82 mm in the control group (Table 1 and table S2). The mean width of the apical foramen was 3.17 ± 0.69 mm in the hDPSC treatment group and 3.54 ± 0.44 mm in the control group (Table 1). The study population showed no statistically significant differences between groups at baseline (Table 1 and table S2). According to the inclusion criteria, all tested incisors (hDPSC implantation and control) had immature roots with incompletely formed apices. Most of the patients were diagnosed with apical periodontitis, depending on the extent of damage in the periapical area assessed by radiographic examination. Patients showed evidence of percussion pain (an indication of inflammation in the periodontal ligament) on clinical examination. During the first visit, we performed a complete root canal preparation and disinfection. Apical periodontitis was under control before the second visit.

Representative cone-beam computed tomography (CBCT) images of roots were taken before hDPSC implantation or apexification and at 6 and 12 months after treatment (Fig. 4A). In the representative example from a patient who received hDPSC implantation (Fig. 4A), the length of the root was increased and the apical foramen was closed at 12 months after surgery; such changes were not observed in the apexification control group (Fig. 4A). The elongation of the root and closure of the apical foramen indicated that the hDPSC implant supported root development and maintenance. In addition, we reconstructed the 3D images of the treated teeth before and after hDPSC implantation. Representative 3D images showed that the tooth root was elongated at 6 and 12 months after hDPSC implantation relative to baseline (before treatment) (Fig. 4B). In addition, increased dentin thickness was observed in the hDPSC implantation group at 6 and 12 months after treatment, but not in the control apexification group (Fig. 4, B and G, and table S3).

Fig. 4 Pulp regeneration in the incisor teeth of patients after hDPSC implantation.

(A) Representative CBCT images of hDPSC-implanted incisor teeth at 6 and 12 months after treatment. The length of the root (red line) was increased at 6 and 12 months after hDPSC implantation. The apical foramen (blue line) was closed 12 months after hDPSC implantation. In the control group, the length of the root was not increased and the apical foramen was not closed at 12 months after apexification treatment. (B) Representative 3D images of a traumatized immature permanent human incisor tooth before and after implantation of hDPSCs. Frontal images (top) and lateral images (bottom) were constructed using Materialise’s interactive medical image control system (Mimics). Roots were elongated at 6 and 12 months after hDPSC implantation compared to before treatment (white stippled circles). In addition, the amount of dentin was increased at 6 and 12 months after treatment in the hDPSC implantation group (white arrows). (C to F) hDPSC implantation into patient incisor teeth improved vascular formation (C), sensation measured by the electric pulp test (D), root length (E), and width of the apical foramen (F) at 6 and 12 months after implantation. (G) Dentin thickness in implanted incisor teeth was increased at 6 and 12 months after hDPSC implantation compared to baseline. Error bars are means ± SD. Data were analyzed using Student’s t test.

To test the viability of hDPSC-treated teeth, we performed laser Doppler flowmetry and electric pulp testing. At 6 months after treatment, laser Doppler flowmetry showed a mean increase in vascular formation of 6.39 ± 0.83 PU for the hDPSC implantation group and a mean decrease in vascular formation of 0.27 ± 0.58 PU for the control apexification group (P < 0.0001 between-group comparison; Fig. 4C and table S3). At 12 months after treatment, laser Doppler flowmetry showed a mean increase in vascular formation of 7.19 ± 0.77 PU for the hDPSC implantation group and a mean decrease in vascular formation of 0.05 ± 0.48 PU for the control group (P < 0.0001 between-group comparison; Fig. 4C and table S3). Moreover, electric pulp tests showed a mean decrease in sensation of 35.29 ± 6.90 at 6 months after treatment and 43.43 ± 0.86 at 12 months after treatment for the hDPSC implantation group and a mean decrease in sensation of 0.1 ± 0.17 and 0.17 ± 0.16 for the control group at 6 and 12 months, respectively (P < 0.0001 between-group comparison; Fig. 4D and table S3).

CBCT analysis showed that hDPSC implantation increased the length of the treated tooth root by, on average, 4.06 ± 0.82 mm at 6 months after treatment and 5.24 ± 0.92 mm at 12 months after treatment, compared to an increase in the control group of 0.61 ± 0.54 mm at 6 months and 0.88 ± 0.67 mm at 12 months (P < 0.0001 between-group comparison; Fig. 4E and table S3). In addition, the hDPSC implantation group showed a mean decrease in apical foramen width of 1.73 ± 0.49 mm at 6 months after treatment and 2.64 ± 0.73 mm at 12 months after treatment compared to a decrease of 0.44 ± 0.16 mm at 6 months and 0.62 ± 0.22 mm at 12 months in the control apexification group (P < 0.0001 between-group comparison; Fig. 4F and table S3).

Next, we harvested hDPSC-regenerated pulp tissue from the implanted incisor tooth of one patient at 12 months after implantation as the treated tooth had become retraumatized. The histology staining showed that hDPSC implantation had led to regeneration of 3D whole dental pulp tissue containing an odontoblast layer, connective tissue, and blood vessels, similar to normal dental pulp (Fig. 5, A to C). We used immunofluorescence staining to show that the newly regenerated dental pulp expressed the neuronal marker NeuN (Fig. 5B). Overlapping NeuN and DAPI staining indicated nerve generation after hDPSC implantation (Fig. 5B). However, normal control dental pulp cells failed to express NeuN (Fig. 5A).

Fig. 5 Histological analysis of hDPSC-regenerated dental pulp.

(A) Representative histological images show that normal pulp tissue of human teeth lacks NeuN-positive cells. Scale bar, 200 μm. (B) Representative image of a human incisor 12 months after hDPSC implantation shows regenerated pulp tissue with a similar tissue structure to that of normal human pulp tissue. Odontoblasts (black arrows) were observed at the margin of the regenerated pulp tissue. Scale bar, 200 μm. (C) Pulp tissue regenerated after hDPSC implantation contained NeuN-positive cells (red); DAPI (blue) was used to stain nuclei. Odontoblasts were observed at the margin of the regenerated pulp tissue. Two regions of regenerated pulp tissue are shown at higher magnification. Scale bar, 20 μm.

Safety evaluation of patients after hDPSC implantation

We failed to observe any significant side effects 12 months after implantation of hDPSCs. Removal of dental pulp from autologous canine teeth did not influence the resorption of the deciduous root or filling material. To further evaluate the safety of hDPSC implantation, we continued follow-up of the patients after hDPSC implantation for 24 months. In total, six patients were excluded because they moved to a different city, contact was lost, or the implanted tooth was retraumatized. We found that the following blood cell numbers were within the normal range: CD4+ T lymphocytes, CD8+ T lymphocytes, B lymphocytes, and natural killer (NK) cells (Table 2). We also found that blood bilirubin, lactate dehydrogenase, creatinine phosphokinase, uric acid, antistreptolysin O, C-reactive protein, rheumatoid factor, immunoglobulin A (IgA), IgG, IgM, C3, C4, and myocardial enzymes were in the normal range for these 20 patients (Table 2). These data suggested that hDPSC implantation had no effects on immune response, liver function, renal function, and myocardial function at 24 months after treatment. Moreover, in our minipig experiments, we showed that pig DPSC implantation did not alter the numbers of CD3+, CD4+, and CD8+ T cells in minipig blood at 6 months after implantation (fig. S4 and table S5).

Table 2 Safety assessment of 20 patients 24 months after hDPSC implantation

ALT, alanine aminotransferase; AST, aspartate aminotransferase.

View this table:

Digital RVG images showed no inflammation at the periapical area in any of the incisor teeth after hDPSC implantation, and continued root development was observed 24 months after treatment (Fig. 6 and fig. S5). To test the viability of hDPSC-treated teeth, we performed laser Doppler flowmetry and electric pulp testing for the implanted incisor teeth of these 20 patients. Laser Doppler flowmetry analysis showed a mean increase in vascular formation of 7.19 ± 0.77 PU (n = 25) at 12 months after treatment and 8.39 ± 1.35 PU (n = 20) at 24 months after treatment (Fig. 7A and table S4). Electric pulp testing showed a mean decrease in sensation of 43.43 ± 0.86 (n = 25) at 12 months after treatment and 44.55 ± 1.10 (n = 20) at 24 months after treatment (Fig. 7B and table S4). These results indicated that the regenerated dental pulp remained viable in the hDPSC-implanted incisor teeth in 20 patients for up to 24 months after treatment.

Fig. 6 RVG images 12 and 24 months after hDPSC implantation.

To further evaluate the safety of hDPSC implantation, we continued follow-up of 20 patients for 24 months after treatment. Digital RVG images of hDPSC-implanted incisors are shown for 10 patients at 12 and 24 months after hDPSC implantation (images for the remaining 10 patients are shown in fig. S5). The images show no inflammation at the periapical area in any of the incisor teeth after hDPSC implantation. Images show that root length was increased and the apical foramen was closed at 12 and 24 months after treatment. Red arrows indicate incisor teeth before and after hDPSC implantation.

Fig. 7 Dental pulp regeneration after hDPSC implantation into human incisors.

To test the viability of hDPSC-implanted incisor teeth, we performed laser Doppler flowmetry and an electric pulp test for the implanted incisor teeth of 25 patients at 12 months and 20 patients at 24 months after treatment. hDPSC implantation into traumatized patient incisors improved vascular formation as shown by laser Doppler flowmetry (A) and sensation as shown by responses to the electric pulp test (B) at 24 months after treatment compared to 12 months after treatment. Error bars are means ± SD. Data were analyzed using Student’s t test.

DISCUSSION

Our study provides initial experimental and clinical evidence that hDPSCs implanted into injured incisor teeth in the absence of a scaffold promoted 3D dental pulp regeneration and partial tooth recovery. The regenerated pulp tissue contained normal structures such as an odontoblast layer, connective tissue, blood vasculature, and neuronal tissue. The regenerated pulp in the incisor teeth receiving the hDPSC implants showed functional responses to a stimulus in the electric pulp test. Notably, compared to the apexification control group, the hDPSC implant group showed an increase in root length and a reduction in the width of the apical foramen at 12 months after treatment, suggesting that regenerated dental pulp promoted tooth recovery.

Given that tooth root development depends on the residual stem cells in the apical third of the root canal, several studies have assessed the effects of pulp revascularization treatment (2325). This approach provides stem cells from tooth periapical tissue; in this procedure, the periapical tissue is provoked to bleed and forms a clot in the apical foramen after sterilization of the root canal. Because of the limited number of stem cells that reach the root canal, it has been difficult to regenerate whole pulp tissue using that technique (23). hDPSCs have MSC characteristics with an elevated capacity for proliferation and self-renewal (4). In addition, they show an immune phenotype similar to bone marrow MSCs (26). hDPSCs have been successfully used for dental pulp regeneration in animal studies (14). These cells are derived from an accessible tissue source and can be obtained in sufficient numbers to enable their potential clinical application (4). Here, we used hDPSC aggregates containing cells and extracellular matrix (to avoid using a scaffold) to regenerate pulp tissue after implantation into injured incisor teeth in pediatric patients.

hDPSCs are derived from neural crest cells and are capable of differentiating into odontoblasts (4), nerve cells (27), and endothelial cells (28); they have the capacity to form complete dental pulp tissue containing an odontoblast layer, blood vessels, and nerve cells. During tooth development, the tooth receives a sympathetic nerve supply from the superior cervical ganglion (29). Accumulating evidence shows that tooth innervation is also regulated by axon guidance molecules that target the tooth (30). It is possible that paracrine factors produced by the implanted hDPSCs stimulated the innervation of the implanted incisor teeth. hDPSCs may directly contribute to regeneration of pulp and induce an endogenous regeneration response.

At the 12-month follow-up, we did not detect any signs of transplantation rejection or inflammatory responses in the implanted incisor teeth. A recent study showed that human embryonic stem cell–derived retinal pigment epithelium was safe when transplanted into patients with age-related macular degeneration or Stargardt’s macular dystrophy and provided some improvement in visual acuity (31). Furthermore, over 1000 patients receiving MSC treatments did not show serious adverse events (32, 33). Our recent phase 1 clinical trial showed that autologous periodontal ligament stem cell implantation was a safe approach to treat periodontal intrabony defects (34). Although no serious adverse events have been reported in MSC therapy clinical trials, we extended our follow-up to 24 months for 20 of our patients implanted with hDPSCs. At the 24-month follow-up examination, the regenerated dental pulp was found to be viable, and no adverse effects were observed in the implanted incisor teeth. The clinical examination showed that blood measurements were in the normal range for the 20 patients at 24 months after treatment. However, we plan to extend the observation period to examine the health status of our study participants for even longer. Implantation of teeth damaged by trauma with hDPSC aggregates could save young adult dental pulp as indicated in this study by the elongated roots and closed apical foramen of incisor teeth implanted with hDPSCs. Dental pulp tissue containing sensory nerves and blood vessels was regenerated 12 months after hDPSC implantation. After 24 months of follow-up, the regenerated pulp was found to be viable in all 20 patients. Although we found that autologous hDPSC implantation may be an effective approach to regenerate functional pulp in young teeth, we do not know whether allogeneic hDPSCs will be useful for pulp regeneration. We showed that autologous hDPSC implantation is a safe approach for up to 24 months, but longer-term follow-up is required to observe full safety and efficacy. It is interesting that implanted hDPSCs are capable of reorganizing into 3D dental pulp and that the implanted incisor teeth showed some recovery although the mechanisms are still unclear. Such mechanisms will need to be explored in future studies.

MATERIALS AND METHODS

Study design

In the preclinical part of this study, a murine model and a porcine model were used to test whether implantation of dental pulp stem cells from deciduous teeth into injured teeth could regenerate whole dental pulp. Then, an open-label, randomized, controlled phase 1 trial was conducted at the School of Stomatology, FMMU in China. Figure 3 shows a flowchart of the clinical study. A follow-up period of 12 months was chosen because the clinical observation time for traditional apexification treatment is 12 months (35, 36). In addition, we extended the follow-up period to 24 months after treatment for the experimental group to assess the safety of hDPSC implantation. This study was initiated in June 2013 and completed in December 2014. The study protocol no. IRB-REV-2013-002 was approved by the Institutional Review Board at FMMU (see the Supplementary Materials) and was registered with ClinicalTrials.gov (#NCT 01814436).

We recruited 40 patients with traumatic pulp necrosis in an incisor tooth and randomly allocated these patients in a 3:1 ratio to either the stem cell implantation experimental group or the control apexification treatment group. In the experimental group (30 patients), a maxillary deciduous canine was chosen for pulp extraction, and the injured incisor was implanted with hDPSC aggregates derived from the autologous canine tooth pulp. In the control group (10 patients), the injured incisor was treated by apexification using calcium hydroxide as the filling material. The allocation sequence was generated using an online randomization generator (www.randomization.com), concealed by a person not involved with the trial management group, and monitored by the Department of Clinical Research at FMMU. The participants, study personnel, and outcome assessors were all blinded to the treatment allocation, and blinding was maintained until all data had been analyzed.

Subjects

The participants were male and female patients aged 7 to 12 years who had a diagnosis of a traumatized permanent incisor tooth. Parents or guardians of all participants provided written informed consent before clinical treatment and agreed to the proposed stem cell treatment or apexification therapy. We selected patients on the basis of several inclusion and exclusion criteria. Inclusion criteria included patients who were aged 7 to 12 years with mixed dentition, a permanent incisor showing pulp necrosis secondary to trauma, a healthy maxillary deciduous canine tooth, and candidate for stem cell therapy according to the pediatric dentist. Exclusion criteria included poor nutritional condition (serum albumin concentration <2 g/dl), other systemic disease, a history of hereditary disease, severe fear of dental treatment, tooth dysplasia, bruxism and malocclusion, inability to provide the deciduous dental pulp cells, severe dental injuries (complicated crown root or root fracture, tooth displacement), improper oral habits, and withdrawal of informed consent.

Assessments of dental pulp functional recovery

The primary outcomes of the clinical trial included sensation in the hDPSC-implanted incisor teeth as assessed by the electric pulp test, vascular formation as assessed by laser Doppler flowmetry, and formation of sensory nerves and blood vessels as indicated by histological analysis. The secondary outcomes included increased root length and decreased width of the apical foramen based on CBCT analysis. The clinical protocol was composed of seven sessions, and the follow-up sessions took place at 1, 3, 6, 9, and 12 months after treatment. At the first visit, the root canal was prepared and disinfected with conventional endodontic methods in both groups after the baseline CBCT examination. At the second visit 1 month later, we performed the hDPSC implantation for patients in the experimental group and traditional apexification treatment for patients in the control group. At the third, fourth, and sixth visits (1, 3, and 9 months after treatment), clinical examinations and digital RVG were performed on all patients. At the fifth and seventh visits (6 and 12 months after treatment), patients were assessed using CBCT, laser Doppler flowmetry, and an electric pulp test. RVG was used to assess root elongation and the closure of the apical foramen. Laser Doppler flowmetry and electric pulp testing were used to assess vascularization of the treated teeth and the return of sensation in response to temperature stimuli, respectively. CBCT was performed to measure the length of the root and the width of the apical foramen.

Safety assessment

Complications and adverse events during postoperative healing were recorded for all patients. We also performed clinical examination of the patients with hDPSC implantation 24 months after treatment. The teeth with regenerated pulp were assessed by RVG examination, laser Doppler flowmetry, and electric pulp testing. Blood tests were also performed for each patient. Blood test analysis included (i) a decrease/increase in the percentage of T lymphocyte subsets and T lymphocyte subset counts, (ii) a decrease/increase in the percentage of B lymphocytes and B lymphocyte counts, and (iii) a decrease/increase in the percentage of NK cells and NK cell counts. At the time of blood collection, liver function, renal function, and myocardial enzyme examinations were also performed for each patient. Moreover, the concentrations of antistreptolysin O, C-reactive protein, rheumatoid factor, IgA, IgG, IgM, C3, and C4 were measured in the serum using enzyme-linked immunosorbent assay at the Department of Clinical Laboratory, School of Stomatology, FMMU. Because of the limitation of swine antibody, we only examined CD3+, CD4+, and CD8+ T cells in DPSC-treated minipigs.

Histology

The tooth of one patient was excluded from analysis due to re-experiencing of dental trauma 12 months after hDPSC implantation. The newly formed pulp-like tissue in the incisor tooth of one implanted patient was removed via pulpectomy and examined histologically. The specimen was immediately formalin-fixed, paraffin-embedded, and cut into 4-μm slices for hematoxylin and eosin staining and immunohistochemistry staining using antibodies against NeuN (1:200; Sigma-Aldrich). All of the antibodies were used according to the manufacturer’s instructions.

Cell culture

The pulp tissue for hDPSC isolation was harvested using standard sterile techniques. hDPSCs were isolated and expanded in culture using good manufacturing practice–grade reagents in a certified laboratory, as described previously (4). Autologous hDPSCs were obtained from the patient’s maxillary deciduous canine tooth. After local anesthesia, the access cavity was prepared with a round diamond drill, and the dental pulp was extracted and prepared for culture. The root canal was gently flushed with 20 ml of a 3% sodium hypochlorite (NaOCl) solution and subsequently flushed with saline solution. Next, the root canal was filled with calcium hydroxide premixed with iodoform paste and covered with glass ionomer. The deciduous pulp tissue was digested, and nonadherent cells were discarded after 24 hours. The cells were expanded in culture for 2 weeks and showed normal morphology of MSCs. The method of cell aggregate culture was as described previously (37). Briefly, hDPSCs from the second passage were digested with trypsin (Sigma-Aldrich), and a total of 5 × 105 cells were seeded into one well of a six-well plate (Corning) and cultured for 24 hours in culture medium. When the cells reached about 80% confluence, the culture medium was changed to new medium containing vitamin C (100 μg/ml) (Sigma-Aldrich), at which point the cells were at a density of 1 × 106 cells per well. The culture medium was refreshed every 2 days. After 10 days of incubation, a white membrane structure could be observed and the cell aggregate became thicker with the passage of time. The Supplementary Materials provide further details regarding the cell manufacturing process.

Immunocytochemical staining

Surface marker expression for hDPSCs was determined by antibody staining against MSC-defining surface markers CD146 (1:200; Abcam) and CD105 (1:200; Abcam), as well as antibodies against hematopoietic cell marker CD34 (1:200; Abcam). Antibodies against the odontoblast marker dentin sialoprotein (1:200; Sigma-Aldrich) and neuron markers NeuN (1:200; Sigma-Aldrich), nestin (1:1000; Abcam), CGRP (1:200; Abcam), TRPM8 (1:200; Abcam), and TRPV1 (1:1000; Alomone Labs) were also used. All antibodies were used according to the manufacturer’s instructions.

Animal studies

To further verify that hDPSCs were able to regenerate pulp structure, we conducted additional experiments in female immunocompromised mice (Vital River Laboratory Animal Technology Co. Ltd). Briefly, third molar teeth were harvested from human patients by the Oral-Maxillofacial Department in the School of Stomatology, FMMU. Periodontal ligament tissues were carefully scraped away along with partial removal of outer cementum, inner dental pulp tissue, and predentin. Next, deionized water was used to treat the tooth roots for 5 to 6 hours, and during this period, the teeth were concussed for 5 to 6 min every hour by an ultrasonic cleaner (Sheng Yuan). The deionized water was changed once every hour. Then, the teeth were exposed to 17% EDTA for 3 to 4 min, washed in deionized water for 5 min, exposed to 5% EDTA for 2 min and washed in deionized water for 5 min, maintained in sterile phosphate-buffered saline (PBS) (Gibco BRL) with penicillin (100 U/ml) (Gibco BRL) and streptomycin (100 mg/ml) (Gibco BRL) for 72 hours, washed in sterilized deionized water for 5 min, and finally stored in routine media with penicillin (50 U/ml) and streptomycin (50 mg/ml) at 4°C. Then, we inserted hDPSC aggregates into the root canals of human teeth and implanted the root subcutaneously in immunocompromised mice for 8 weeks. To label the newly formed dentin, calcein (20 mg/kg body weight; Sigma-Aldrich) was intraperitoneally injected into the nude mice at 10 and 3 days before sacrifice.

In addition, we also performed pulp regeneration experiments in female minipigs (Shichuang Technology Miniature Pig Breeding Base). We harvested the pulp tissue of deciduous incisor teeth from minipigs, isolated pig DPSCs, and performed the DPSC aggregate implantation into young permanent incisors in which the pulp was removed before implantation. All animal procedures were performed according to the guidelines of the Animal Care Committee of FMMU (Xi’an, China). The root and pulp tissue from serial sections stained with hematoxylin and eosin or Masson’s trichrome were used to evaluate the regenerated pulp tissue. Immunohistochemistry staining used antibodies against NeuN (1:200; Sigma-Aldrich); antibodies were used according to the manufacturer’s instructions.

Injection of hDPSCs into rat dorsal root ganglia in vivo

Following the methods in the study conducted by the Ferreira laboratory in 2007 (38), the injecting needle was prepared by inverting the position of a gingival needle (30 gauge) relative to its plastic syringe connector, which was tightly connected to the metal piece. After shaving the fur on the lower back, the immunocompetent rats (FMMU, Xi’an, China) were anesthetized using 1% sodium pentobarbital (50 mg/kg body weight, ip) and placed over a small cylinder to elevate the lumbar region. To expose the point of puncture, the skin over the rat’s lumbar region was sliced. The point of puncture was defined at 1.5 cm laterally to the vertebral column, about 0.5 cm caudal from a virtual line passing over the rostral borders of the iliac crests. In sequence, the injecting needle was inserted through the muscle, toward the intervertebral space (L3-L4 or L4-L5), until the tip touched the lateral region of the vertebrae. To reach the space between the transverse processes of the vertebrae, delicate movements of the needle were made until the bone resistance was diminished and a paw flinch reflex was observed. The paw flinch reflex was used as a sign that the needle tip had penetrated the dorsal root ganglia. Next, 3 × 105 hDPSCs were injected into the dorsal root ganglia. Here, the dorsal root ganglia in the intervertebral space between L3-L4 and L4-L5 were injected.

Statistical analysis

The difference between groups in the change of sensation based on the electric pulp vitality test at 12 months after hDPSC implantation was estimated to be 10, and the pooled variance was 1; the between-group difference in the change of vascular formation based on laser Doppler flowmetry at 12 months after hDPSC implantation was estimated to be 1.4, and the pooled variance was 1. Therefore, the minimum number of patients estimated to be necessary for this trial to enable detection of such differences at a two-sided type I error rate of 0.05 (α = 0.05) and power of 0.90 (β = 0.10) was 32 in total (24 in the hDPSC implantation group and 8 in the control apexification group). In anticipation of a 20% dropout rate, we recruited 40 patients in total.

All patients who followed the study protocol were included in the analysis according to their original group assignments (per-protocol analysis). Continuous variables are presented as means and SDs, and categorical data are presented as numbers and percentages. Baseline characteristics were compared between the two groups using Student’s t test for continuous variables, Mann-Whitney test for ordinal variables, and chi-square test or Fisher’s exact test for categorical variables where appropriate. The primary and secondary outcomes were analyzed using Student’s t test.

SPSS software package version 16.0 (SPSS) was used for data analysis. P values less than 0.05 were considered significant.

SUPPLEMENTARY MATERIALS

www.sciencetranslationalmedicine.org/cgi/content/full/10/455/eaaf3227/DC1

Fig. S1. hDPSCs differentiate into sensory neurons after intraganglion injection in rats.

Fig. S2. Characteristics of minipig DPSCs.

Fig. S3. Implantation of hDPSC aggregates in patients.

Fig. S4. Number of CD3+, CD4+, and CD8+ T cells in minipigs after pDPSC implantation.

Fig. S5. RVG images 12 and 24 months after hDPSC implantation.

Table S1. Source data for Fig. 1.

Table S2. Source data for Table 1.

Table S3. Source data for Fig. 4.

Table S4. Source data for Fig. 7.

Table S5. Source data for figs. S2 and S4.

Project protocol

References (39, 40)

REFERENCES AND NOTES

Acknowledgments: We thank S. Bai and Z. Dong for assistance with CBCT data analysis and L. Wang for assistance with statistical analysis. Funding: This work was supported by the National Key Research and Development Program of China (grant nos. 2016YFC1101400 and 2017YFA0104800), the Nature Science Foundation of China (grant nos. 81570937, 31401255, 81271117, and 31501121), and a Schoenleber pilot grant from the University of Pennsylvania School of Dental Medicine. Author contributions: K.X., B.L., H.G., and W.S. performed the surgical procedures and animal experiments, cultured cells, collected data, analyzed the data, and produced all figures and tables. X.K. and X.H. performed the immunofluorescence staining. J.S. and Y.Z. generated the randomization codes and designed the study database. A.L. participated in the animal surgery. L.L., S.L., W.L., and C.H. helped with data analysis. S.S. and Y.J. designed the experiments, oversaw the collection of results and data interpretation, and drafted the paper. All authors have seen and approved the final version of the paper. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data associated with this study can be found in the paper or the Supplementary Materials.
View Abstract

Navigate This Article