Sensory nerve-deficient microenvironment impairs tooth homeostasis by inducing apoptosis of dental pulp stem cells.

Abstract Objectives The aim of this study is to investigate the role of sensory nerve in tooth homeostasis and its effect on mesenchymal stromal/stem cells (MSCs) in dental pulp. Materials and methods We established the rat denervated incisor models to identify the morphological and histological changes of tooth. The groups were as follows: IANx (inferior alveolar nerve section), SCGx (superior cervical ganglion removal), IANx + SCGx and Sham group. The biological behaviour of dental pulp stromal/stem cells (DPSCs) was evaluated. Finally, we applied activin B to DPSCs from sensory nerve‐deficient microenvironment to analyse the changes of proliferation and apoptosis. Results Incisor of IANx and IANx + SCGx groups exhibited obvious disorganized tooth structure, while SCGx group only showed slight decrease of dentin thickness, implying sensory nerve, not sympathetic nerve, contributes to the tooth homeostasis. Moreover, we found sensory nerve injury led to disfunction of DPSCs via activin B/SMAD2/3 signalling in vitro. Supplementing activin B promoted proliferation and reduced apoptosis of DPSCs in sensory nerve‐deficient microenvironment. Conclusions This research first demonstrates that sensory nerve‐deficient microenvironment impairs tooth haemostasis by inducing apoptosis of DPSCs via activin B/SMAD2/3 signalling. Our study provides the evidence for the crucial role of sensory nerve in tooth homeostasis.


| INTRODUC TI ON
Nerve putatively plays a critical role in tissue homeostasis and regeneration, such as bone homeostasis maintaining 1 and salamander limb regeneration. 2 However, how nerve modulates tooth homeostasis remains largely unknown. Our team previously has found abnormal morphology of teeth in a patient with congenital insensitivity to pain with anhidrosis (CIPA), which is a rare inherited disorder of the peripheral nervous system. The patient displays severe oral manifestations, such as dentin hypoplasia and cementogenesis defects, implying the indispensable role of sensory nerve in tooth development. 3 The rodent incisor is mainly innervated by sensory nerve and sympathetic nerve, which is derived from mandibular inferior alveolar nerve (IAN) 4 and sympathetic superior cervical ganglion (SCG) respectively. 5 It has been reported that impeding sensory nerve innervation leads to morphological aberrancy in mouse lower incisor, 6 while which type of nerve mainly regulates incisor homeostasis has not been fully understood yet.
Nerve provides microenvironment for stromal/stem cells through complex signalling to dynamic regulate their behaviour, building highly elaborate structures and thus maintaining tissue homeostasis. 7,8 Studies have illustrated that denervation impedes cellular behaviour in various tissues, 8,9 but whether and how denervation impairs odontogenic stromal/stem cells responsible for tooth homeostasis remains elusive.
The rodent lower incisor as an independent organ grow continuously, offering an excellent model to address the relationship between nerve and odontogenic stromal/stem cells. 10 Additionally, dental pulp stromal/stem cells (DPSCs) as a kind of mesenchymal stromal/stem cells (MSCs) in tooth can differentiate into odontoblasts and pulp cells participating in tooth homeostasis and regeneration. [11][12][13] After sensing injury stimuli, the stem cells in pulp proliferate and migrate to the injured area to maintain tooth homeostasis. 11 In the field of tissue engineering, DPSCs exhibit excellent capacities of proliferation and differentiation when combined with scaffolds such as nanoporous, hydrogel and mesoporous silicon scaffolds; thus, they have been widespread applied in tooth, bone and nerve regeneration procedures. 14 For tissue regeneration, many innovative scaffolds have been proved beneficial to stem cell-based tissue regeneration, including nanosilicates which are mineral-based two-dimensional nanomaterials, graphene which is utilized to create three-dimensional porous foams and bioactive nanomaterial delivery system which can release bioactive molecule as needed. 15,16 However, the direct effect of nerve-deficient microenvironment on DPSCs is largely unclear. Understanding this will shed light on the regeneration mechanism of incisor and further provide novel perspective of pulp regeneration.
In this study, we used rat denervated models to demonstrate that sensory nerve, not sympathetic nerve, maintains tooth homeostasis and dentin formation. Moreover, through isolating rat DPSCs from sensory denervated incisor (IANx-DPSCs), we found denervation led to DPSCs disfunction. Sensory nerve injury induced apoptosis of DPSCs via activin B/SMAD2/3 signalling. Supplementing activin B promoted proliferation and reduced apoptosis of DPSCs in sensory nerve-deficient microenvironment. week-old female Sprague Dawley (SD) rats were used for DPSCs isolation. Six-week-old female SD rats were used for histomorphology assay.

| Animals
Six-week-old SD rats were distributed into four groups: IANx group (sectioning the unilateral IAN), SCGx group (removing the unilateral SCG), IANx + SCGx group (sectioning the unilateral IAN and removing the ipsilateral SCG) and the Sham group (performing the same surgical procedures except for resection of the nerve).
The denervation procedure had no impact on food or water uptake of the rat. Each group of rats was, respectively, sacrificed after 1, 2, 3 weeks through heart infusion, then collected the mandible and fixed them in 4% paraformaldehyde.
Four-week-old SD rats were distributed into Sham and IANx groups. Two weeks after surgery, the DPSCs were obtained from Sham and IANx group.

| Inferior alveolar nerve axotomy (IANx)
The IANx surgery was severed as previously described. 17 Under general anaesthesia, an extraoral horizontal incision was made to fully expose the masseter muscle. The bone surface of the mandible was exposed by blunt dissection of the masseter muscle. A small dental round bur was used to remove the cortex bone and expose the IAN.
Then 5 mm length of IAN was removed. The muscle and skin were closed and sutured ( Figure A1(A)).

| Superior cervical ganglionectomy (SCGx)
The SCGx surgery was performed as previously described. 18 Under general anaesthesia, the neck muscles were exposed through a 4-cm vertical incision of neck region. Then, forceps were applied to dissect the deep cervical fascia and partially remove carotid sheath.
After separating the common carotid artery, the SCG which was behind the carotid bifurcation was removed. Ipsilateral blepharoptosis has been used as indicator of the successful removal of SCG ( Figure A1(B)).

| Micro-CT and histological analysis
The samples were collected and analysed by micro-CT (Siemens Inveon, Germany). The longitudinal images of the mandibular incisor were acquired through three-dimensional reconstruction, and the height of contour of mandibular first molar mesial surface was served as a fixed position to obtain cross-sections. The thickness of dentin was analysed via ImageJ 1.47 software. After micro-CT scanning, all mandibles were decalcified by 17% ethylene diamine tetraacetic acid (EDTA) (Pulpdent), embedded with paraffin and sliced in the sagittal plane for haematoxylin and eosin (H&E) (Leica) and Masson trichrome staining (Baso). The thickness of dentin and enamel at the apical were used for statistical analysis.

| Isolation and culture of dental pulp stem cells (DPSCs)
The dental pulp was extracted from the lower incisor after removing apical buds. The DPSCs were obtained with tissue and enzymic digestion. In particular, the pulp was minced into 1 mm 3 and digested in 3 mg/ mL collagenase I (Sigma-Aldrich Corp) at 37°C, 5% CO 2 for 1.5 hours.
Cells were then plated evenly on 6-well plates with α-minimum essential medium (α-MEM; Gibco BRL) supplemented with 20% foetal bovine serum (FBS; Gibco BRL). Recombinant Mouse activin B Protein (R&D systems, 8260-AB) was applied to culture medium in IANx group after cells adhesion at 10 ng/mL as recommended. 20

| Colony-forming assay
To assess the ability to produce colony-forming unit (CFU), singlecell suspensions of DPSCs from Sham and IANx groups (1 × 10 3 cells) were, respectively, seeded into 10-cm-diameter culture dishes. After 14 days of cultivation, cells were fixed with 4% paraformaldehyde for 30 minutes and then stained with 1% toluidine blue.

| Osteogenic/adipogenic differentiation of dental pulp stem cells (DPSCs)
Induction of osteoblasts and adipocytes was performed as previously described. 21,22 After 28 days of osteogenic induction, the cells were characterized by Alizarin red S staining, and total RNA was extracted and analysed for the presence of osteogenic genes (Alp, BioTek) at 520 nm.

| Western blot analysis
See Appendix 1.

| Statistical analysis
All data were expressed as mean (±SD) from at least three independent experiments and analysed by two-tailed unpaired Student's t test or one-way ANOVA test using GraphPad Prism 5.0. Values of P < .05 were considered statistically significant.

| Sensory nerve mainly maintains the tooth homeostasis
The rat lower incisors are mainly innervated by sensory nerve fibres derived from IAN and sympathetic nerve fibres derived from SCG.
To clarify the role of sensory and sympathetic nerve in supporting tooth homeostasis, we established rat models including sectioning the unilateral sensory nerve (IANx) and sympathetic nerve (SCGx) ( Figure A1). Firstly, we observed the incisors changes in 1, 2 and 3 weeks (Figure 1, Figure A2). Compared to Sham ( Figure 1A) and SCGx group ( Figure 1C), the incisors from IANx ( Figure 1B) and IANx + SCGx group ( Figure 1D) turned chalky after 2 weeks of surgery. Thus, we set 2 weeks after surgery as a time point to do further research. Micro-CT of the four groups showed some irregular masses of calcifications in pulp of IANx ( Figure 1F) and IANx + SCGx group ( Figure 1H), but not in Sham ( Figure 1E 23 The finding indicated that it is the sensory nerve, not sympathetic nerve, maintains the tooth homeostasis.

| Sensory nerve contributes to dentin formation
Through micro-CT analysis, we acquired cross-sections through a fixed position to compare the thickness of dentin wall of the four F I G U R E 1 Sensory nerve maintains the phenotype of incisor. A-D, The phenotype of rat incisors after 2 wk of surgery. The right incisor in IANx (B) and IANx + SCGx (D) groups turn chalky (indicated by boxes), while the Sham (A) and SCGx group (C) do not exhibit the similar phenotype. E-H, The micro-CT shows the longitudinal images of the incisors from four groups after 2 wk of surgery. The irregular calcified masses (indicated by red arrows) can be observed in IANx (F) and IANx + SCGx group (H), not in SCGx (G) or Sham group (E). Scale bar: 2 mm. I-L, The cross-sections are sampled at comparable positions, indicated by yellow lines in E-H. Scale bar: 1 mm. The area of dentin walls (indicated by yellow arrows) is measured. M, Compared to Sham, the thickness of dentin in the other three groups displays a reduction. (*P < .05; **P < .01; n = 5). Besides, IANx and IANx + SCGx groups show thinner dentin wall than SCGx group. (**P < .01; n = 5) groups. Compared to Sham group, the other three groups showed a reduction of dentin thickness ( Figure 1I-M, P < .05). Among them, the thickness of dentin wall in IANx and IANx + SCGx group displayed a significant decrease ( Figure 1M, P < .01). Histological examination also exhibited distinct thinner layer of enamel and dentin, respectively, at the incisor apical of IANx and IANx + SCGx group

| Dental pulp stem cells from sensory nerve-deficient microenvironment display impaired properties
To investigate the influence of microenvironment lack of sensory innervation on DPSCs, we isolated the DPSCs from sensory nerve- MSCs markers CD29 and CD105 and negative for hemopoietic stem cells marker CD45 ( Figure A4(B)). Moreover, the CFU analysis displayed that the IANx-DPSCs exhibited lower CFUs rates than Sham-DPSCs ( Figure 4A,B, P < .01). In addition, compared to Sham group, IANx-DPSCs showed an impaired proliferative capacity ( Figure 4C, P < .005) and higher percentage of apoptotic cells ( Figure 4D,E, P < .01).
To further compare the multi-differentiation capacity of the two groups, we analysed the osteogenic and adipogenic differ-

| Supplementing activin B promotes proliferation and reduces apoptosis of DPSCs from denervated microenvironment
Given sensory nerve-deficient microenvironment impairs DPSCs properties, sensory nerve may secrete bioactive compounds to constitute favourable microenvironments for DPSCs. It has been F I G U R E 3 Sensory nerve contributes to dentin formation. A, HE and Masson staining shows the area of incisor apical at comparable positions. IANx and IANx + SCGx group exhibits thinner enamel (indicated by arrows) and dentin (indicated by double-headed arrows) walls than Sham and SCGx group. Scale bars:

| D ISCUSS I ON
Nerve axons integrate in the tooth germ participating in tooth development. 25 Nerve is crucial in tooth development and homeostasis. 6,26 Our team previously has found abnormal morphology of teeth in a patient with congenital insensitivity to pain with anhidrosis (CIPA), which is a rare inherited disorder of the peripheral nervous system.
The patient displays severe oral manifestations including dentin hypoplasia and cementogenesis defects, 26  a few parasympathetic nerve fibres which are still under controversial. 25 The sensory nerve primarily governs the sense of pain, the pressure and vasodilatation regulation, 4,27 while the sympathetic nerve tends to regulate vasoconstriction functions. 5, 28 Though it has been reported IANx leads to slower incisor eruption and tooth discoloration, 23 whether sympathetic nerve participate in incisor growth remains unclear. In this study, we utilized IANx, SCGx and IANx + SCGx models to investigate the contribution and interaction of sensory and sympathetic nerve in tooth growth. The results show the incisor from IANx + SCGx group displayed the similar changes with IANx group such as disorganized pulp structure, irregular calcification, as well as less dentin and enamel formation (Figures 1-3), indicating sympathetic nerve barely influences the tooth homeostasis.
Although no obvious structure change was noted in the SCGx group, the thickness of dentin reduced comparing with Sham group, suggesting it participates in the process of dentin formation (Figures 1-3). To our knowledge, our study is the first to provide comparable evidence that it is sensory nerve, not sympathetic nerve, that mainly In the present study, we find that although the adverse microenvironment has been removed, DPSCs from sensory nerve-deficient microenvironment still display impaired capacities ( Figure 4).
Besides, the expression level of activin B dramatically reduced in IANx pulp tissue ( Figure 5A), but not in DPSCs ( Figure A5(A)).
Importantly, supplementing activin B partly rescues the impaired properties of DPSCs ( Figure 5F-H  In summary, we first demonstrate that sensory nerve ablation has adverse impact not only on tooth homeostasis, but also on the properties of DPSCs. Though the DPSCs from sensory nerve-deficient microenvironment exhibit impaired behaviour, they can be alleviated by activin B treatment. The present study exhibits that sensory nerve-deficient microenvironment impairs tooth haemostasis by inducing apoptosis of DPSCs via activin B/SMAD2/3 signalling, providing a novel perspective for the interaction between sensory nerve and odontogenic MSCs.

CO N FLI C T O F I NTE R E S T
The authors have no conflicts of interest to declare.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets used and/or analysed during the current study are available from the corresponding author on reasonable request.

F I G U R E A 2
The incisor phenotype changes with time. A, After 1 wk of surgery, there were no differences in the four groups. After 3 wk of surgery, the right incisor of IANx and IANx + SCGx groups turn chalky (indicated by boxes)

F I G U R E A 3
The efficiency of denervation. A, Immunofluorescence staining shows the expression of CGRP (red) decrease in IANx and IANx + SCGx group, while the expression of TH (green) decrease in SCGx and IANx + SCGx group. Scale bar: 20 μm, OB: odontoblast, * indicates disorganized structures F I G U R E A 4 The morphology and characteristics of DPSCs. A, The Sham group exhibits a long spindleshaped morphology and grew robustly. However, the IANx group exhibits long spindle and polygonal shape without good growth state. Scale bars: 100 μm B, Flow cytometry analysis of DPSCs from IANx and Sham group shows that the MSCs surface markers CD29 and CD105 are positive of DPSCs, while the hematopoietic stem cell surface marker CD45 is negative

F I G U R E A 5
The activin signalling related genes in DPSCs. A, Except activin A, the expression level of activin B, Acvr2a and Acvr2b show no significant difference between IANx and Sham group. (**P < .01, n = 3). B, After activin B treatment, the expression level of activin A, Acvr2a and Acvr2b is increased in IANx group. (*P < .05, **P < .01, ***P < .005, n = 3)