Trigeminal neurons control immune-bone cell interaction and metabolism in apical periodontitis

Abstract Apical periodontitis (AP) is an inflammatory disease occurring following tooth infection with distinct osteolytic activity. Despite increasing evidence that sensory neurons participate in regulation of non-neuronal cells, their role in the development of AP is largely unknown. We hypothesized that trigeminal ganglia (TG) Nav1.8+ nociceptors regulate bone metabolism changes in response to AP. A selective ablation of nociceptive neurons in Nav1.8Cre/Diphtheria toxin A (DTA)Lox mouse line was used to evaluate the development and progression of AP using murine model of infection-induced AP. Ablation of Nav1.8+ nociceptors had earlier progression of AP with larger osteolytic lesions. Immunohistochemical and RNAscope analyses demonstrated greater number of macrophages, T-cells, osteoclast and osteoblast precursors and an increased RANKL:OPG ratio at earlier time points among Nav1.8Cre/ DTALox mice. There was an increased expression of IL-1α and IL-6 within lesions of nociceptor-ablated mice. Further, co-culture experiments demonstrated that TG neurons promoted osteoblast mineralization and inhibited osteoclastic function. The findings suggest that TG Nav1.8+ neurons contribute to modulation of the AP development by delaying the influx of immune cells, promoting osteoblastic differentiation, and decreasing osteoclastic activities. This newly uncovered mechanism could become a therapeutic strategy for the treatment of AP and minimize the persistence of osteolytic lesions in refractory cases. Graphical abstract Supplementary Information The online version contains supplementary material available at 10.1007/s00018-022-04335-w.


Introduction
Dental infections are a prevalent global health problem resulting in significant human suffering, premature tooth loss, and decrease in quality of life and productivity [1]. Ingress of microorganisms, that are normally occurring in the oral cavity, into the mineralized dental structures is 1 3 330 Page 2 of 18 marked by robust inflammation and severe pain [2]. As the disease progresses, the inflammation extends apically resulting in the development of apical periodontitis (AP) that is often accompanied by substantial bone remodeling [3]. This change in bone development and differentiation is believed to result from a shift between osteoblastic and osteoclastic activity, favoring the loss of mineralized bone matrix. Some of the complex immunologic events leading to this bone metabolism imbalance include the diffusion of microbial antigens through the apical opening of the root canals with the subsequent activation of cells characteristic for the innate and adaptive immune response [4]. Although the role of immune cells and key cytokines in bone remodeling due to dental infections has been previously investigated [5,6], far less is known for the role of trigeminal afferent neurons, which are abundantly present in relevant tissues, in this pathologic process.
There is an increasing body of evidence that peripheral sensory neurons participate in skeletal homeostasis through elaborate cross-talk with other non-neuronal cells in the appendicular skeleton [7]. Thus, it is agreed that craniodental homeostasis is closely influenced by nociceptors, particularly how these pain-sensing fibers participate in the dysregulation of bone metabolism following dental infections [8,9]. Dental structures are densely innervated with nociceptors, and we have demonstrated that the pharmacological ablation of capsaicin-sensitive neurons led to an early onset of bone loss following dental infections in rats [8]. Despite this novel observation, the mechanism by which peripheral sensory neurons regulate this increased bone loss following dental infection is still largely unknown.
The pharmacological ablation by capsaicin affects neurons expressing the transient receptor vanilloid receptor-1 (TRPV1) only affects a subpopulation of nociceptors [10]. It was demonstrated that a third of the dental innervating neurons are TRPV1 + , while as much as 60% dental nerve fibers contain calcitonin gene-related peptide (CGRP) [11]. Additionally, TRPV1 has been found on multiple cell types including immune cells, which makes the pharmacological intervention less likely to be merely neuronal-selective [12]. This indicates that other tools affecting broader subset of sensory neurons need to be applied for through investigation of bone metabolism regulation by the trigeminal ganglia (TG) sensory neurons. In this respect, TG neurons could be divided into unmyelinated C-fibers, lightly myelinated A-delta fibers (A-δ), and heavily myelinated A-beta (A-β) fibers [13]. Recent studies have further classified sensory neurons using advanced molecular profiling and transgenic tools into various distinct subsets [14][15][16]. Close comparative analysis showed that a hallmark of nociceptors is expression of the tetrodotoxin-resistant (TTX-R) voltage gated sodium channel (Nav1.8), thus making it a suitable molecular marker for manipulating broad spectrum of nociceptors [17]. The use of Cre-loxP system has been efficiently utilized to generate mouse lines with either conditional knockout or reporter mice for specific marker-genes [15,18]. This genetic tool can be applied to ablate defined cell types. Thus, the crossing of Nav1.8 Cre and diphtheria toxin (DTA) Lox mice can lead to double heterozygous offspring in which Nav1.8-positive cells are ablated through DTA expressioninduced toxicity. We have utilized this approach to isolate and study roles of Nav1.8 + nociceptors in apical periodontitis bone loss.
Bone metabolism is a dynamic process and mediated by balanced activities of osteoblasts/osteoclasts cells to maintain the bone mass. In inflammatory pathological conditions such as AP, a shift occurs in the balance leading to abnormal osteoclastic or osteoblastic functions, resulting in bone loss [19]. Bone is innervated by sensory nerves, and there is growing evidence that dorsal root ganglia (DRG) sensory neurons participate in skeletal bone homeostasis and remodeling [20,21]. However, it is known that DRG and TG neurons are cellularly and functionally distinct [14] and the orofacial region is innervated by the TG neurons compared to the rest of the body which is innervated by DRG neurons. In addition, AP resulting from dental infections has a unique pathophysiology with the infected tooth serving as a revoir of a complex microorganism flora whose antigens diffuse through the root foramen resulting in a long-standing robust immunological challenge leading to pain and localized bone loss. Further, the density of periapical innervation has been shown to increase within tissues in AP as result of neuronal sprouting [22], yet the role of sensory innervation in the bone homeostasis of this prevalent disease is largely unknown. Therefore, the aims of the present study were to: (1) to investigate the role of nociceptors (Nav1.8 + neurons) in the development of AP using transgenic selective nociceptor ablation and a well-established murine model of infection-induced AP, and (2) to evaluate TG nociceptors regulation of osteoblastic and osteoclastic function using an in vitro co-culture system.

Animals
Animals were housed 2-5 per cage and maintained in a light/ temperature control environment with food and water available ad lib. Breeding was performed in our facility, and all experiments were performed on 8-12-week-old mice with equal distribution of male and females.

Apical periodontitis model
All experiments were performed on 8-10-week-old animals with a sample size of eight mice per group per timepoint with equal distribution of female and male mice. In brief, animals were anesthetized via intraperitoneal (IP) injection of ketamine (75 mg/kg)/dexdomitor (1 mg/kg), as described previously [8]. Animals were mounted on a custom jaw retraction apparatus, and two standardized cavities (i.e., mesial and distal of the pulp chamber) were created on the occlusal surface of the mandibular left molar using a sterile high-speed size #1/4 round bur to the depth of the diameter of the bur with the aid of a surgical microscope (Zeiss). The exposure of the dental pulp to the oral environment was confirmed visually with the insertion of an Endodontic K-file #8 (Dentsply Maillefer) into the mesial and distal canals of the tooth. Next, animals received an intraperitoneal injection of Antisedan (1 mg/kg) to reverse the anesthesia, and they were then housed in a climate-controlled environment with free access to water and food until tissue collection.

Tissue collection
At the end of each timepoint (3, 7 or 14 days), animals were briefly anesthetized by isoflurane inhalation (for 60-90 s) and sacrificed by cervical dislocation. The mandibles were dissected, fixed for 2 h in 4% paraformaldehyde in 0.1 M phosphate buffer (PB), rinsed and stored in 0.1 M PB (Ther-moFisher Scientific) until mandibles were scanned with micro-computed tomography (μCT) and later subjected to immunohistochemistry or RNAscope. Next, dissected surrounding bone and teeth were pooled from three mice per sample, and total RNA was isolated following immediate snap freezing in liquid nitrogen, grinding and homogenization in RNA lysis buffer. In addition, trigeminal ganglia were dissected from the animals at different time points and frozen in RNAlater (Thermo Fisher Scientific) at -20 °C until use.

Total RNA isolation and RT-PCR
Total RNA (0.1-0.5 µg) was isolated from harvested periapical lesions and trigeminal ganglia tissues (n = 6/group/ time point). In addition, total RNA (1 µg) was also isolated from IDG-SW3 co-cultured with primary TG neurons (see below) using the RNAeasy Kit (Qiagen) according to the manufacturer's recommendation followed by cDNA synthesis using the High-Capacity RNA-to-cDNA kit (Thermo Fisher Scientific).
Real time PCR reactions were conducted using the Taqman™ Fast Advanced Master Mix (Thermo Fisher Scientific) and Taqman™ gene expression assays for in vivo samples and co-culture experiments (see table in supplemental materials) on an ABI 7500 Fast Real-Time PCR System (Thermo Fisher Scientific). The relative expression of each target gene messenger RNA (mRNA) was determined using the comparative delta-delta cycle threshold method (ΔΔCt), after normalization to GAPDH, using the control group as the calibrator as previously reported [25].

Micro-computed tomography
Mandible samples were scanned in a Micro-Computed Tomography Bruker Skyscan1172 (μCT) system (Bruker) to quantify the volume of the osteolytic lesion in the apical area (i.e., progression of AP). Samples were scanned with the following settings: 60 kV, 167 mA, 0.7° rotation step, four frames averaging, and 8 μm voxel size. The region of interest (ROI) included the osteolytic space surrounding the roots and the periodontal ligament space surrounding both roots of the first mandibular molar as a measure of apical lesion volume. The ROI started at the first sagittal slice, where the mesial full root canal space is visible through the apex and included three slices in both buccal and lingual directions with 50 μm interval. This area was manually outlined by blinded observers using CT-Analyser software (Bruker), and an automated thresholding script was used to measure the total volume of the void space. The mean lesion volume of all slices was normalized to the volume of the periodontal space of the contralateral healthy tooth within subjects. Thus, the intact first molar in the contralateral side for every animal was used as its own control to normalize the acquired data presented as percent of the contralateral healthy tissue.

Immunohistochemistry (IHC)
Sectioned mandible samples were processed for IHC with corresponding primary antibodies to evaluate the presence of CGRP + neuronal fibers and to quantify numbers of cluster of differentiation 68 (CD68) + macrophages, cluster of differentiation 3 (CD3) + T-cells, osteocalcin + osteoblasts, and Ki67 + progenitor cells, as well as the expression pattern of RANKL and OPG within apical lesions. In brief, fixed mandible samples were decalcified in a 4% EDTA solution at 4 °C for 3-4 weeks until ready for sectioning. Samples were cryo-protected with 30% sucrose in phosphate buffer overnight before they were embedded in Neg 50 (Richard Allan Scientific). Next, samples were serially sectioned at 20 μm in the longitudinal (Sagittal) plane with a cryostat, and tissue sections were collected onto Superfrost glass slides (Fisher Scientific) and stored at − 20 °C until use.
Samples were permeabilized and blocked using Maxpack Immunostaining Media Kit (Active Motif) and incubated with primary antibodies (see supplemental information) overnight, followed by species-appropriate Alexa Fluor IgG secondary antibodies (Thermofisher; Alexa Fluor 488/568 at a 1:200 concentration). Control sections were processed similarly but without primary antibodies. Next, immune reactivity was detected with a Nikon Eclipse 90i equipped with a C1si laser system confocal microscope (Nikon Instruments). Single images were acquired at 10 × and 20 × magnifications with NIS-elements software (Nikon Instruments) with standardized settings by blinded observers. A minimum of three sections from three independent animals (n = 3) were used for all IHC analyses. Positively stained cells for each target within the periapical lesion were counted using NIH ImageJ software. Cell counts in the periapical lesion were normalized to the contralateral intact healthy tooth. Lastly, sectioned samples were also stained with hematoxylin and eosin (Dako) following basic histological staining protocol.

RNAscope®
After each timepoint, mandibles were dissected and fixed in 10% Neutral Buffered Formalin (Sigma) for 24 h. The samples were then washed in PBS and placed in ACD Bone Decalcification Buffer (Advanced Cell Diagnostics) at 4ºC for 2 weeks, changing the solution every 2 days. Mandibles were paraffin-embedded, sectioned at 4 μm, and mounted on Superfrost glass slides (Fisher Scientific). The samples were then processed with the RNAscope ® 2.5 HD Duplex Detection Kit (Advanced Cell Diagnostics) following manufacturer's instructions.
A total of three slides per condition (AP or contralateral control) from different samples at each time point was used for each combination of probe pairs. Brightfield images of the periapical lesion were captured at 40 × magnification using a Nikon Ts2 inverted microscope (Nikon Instruments) using NIS-elements software (Nikon Instruments). Relative gene expression within the apical osteolytic lesions was quantified using NIH ImageJ software after color thresholding. Data are presented as area of each of red and green signals over total region of interest.

Osteoblast co-cultures
The osteoblast cell line IDG-SW3 with green fluorescent protein (GFP) expression under the Dentin Matrix Acidic Phosphoprotein 1(DMP1) promoter (Kerafast) or murine calvaria osteoblast precursor cells (MC3T3-E1, subclone 4) (American Type Culture Collection) were cultured at the concentration of (~ 1 × 10 5 /well) in collagen type-1 coated 24-well plated in the same medium of TG neurons but with the addition of ascorbic acid (50 μg/ml) and β-glycerolphosphate (4 mM). Once cells reached 80% confluency, inserts of TG neuronal cultures were transferred to the plates and incubated for 7 or 14 days at 37 °C and 5% CO 2 with media replaced every 2 days and new neuronal inserts every 7 days.

RNA sequencing
Total RNA was isolated from IDG-SW3 cells after 14 days culture in the presence of differentiation media, in the presence or absence of cultured TG neurons from either control or Nav1.8-DTA mice (n = 2 wells/condition) using the RNAeasy Kit (Qiagen) following the manufacturer's recommendation. The total RNA was submitted to the Genome Sequencing Facility at UT Health San Antonio for sequencing and analysis [30]. Briefly, samples were sequenced by 100-bp pair-end module, and Fastq sequences were aligned by tophat2 with ucsc-mm10 genome. Then, gene readCounts were counted by HTSeq, and Reads Per Kilobase of transcript, per Million mapped reads (RPKM) were calculated and corrected by qRNA script. Finally, pairwise comparisons were done by DESeq, and differentially expressed genes (DEGs) for further analysis were selected according to following criteria: RPKM > 1, fold change (FC) ≥ 1.5 and adj p < 0.05 calculated by Benjamin-Hodjberg test. Gene ontology biological process analyses of DEGs were performed using Protein Analysis Through Evolutionary Relationships (PANTHER) Overrepresentation Test (PANTHER 16.0 Released 20210224) [31]. DEGs were compared against all genes in the Mus musculus reference list using a Fisher test and Bonferroni correction. PANTHER software was used to determine enriched biological processes from GO Ontology database (https:// doi. org/ 10. 5281/ zenodo. 50809 93 Released 2021-07-02) and provided outputs including Fold Enrichment scores, number of genes present in the annotation data category, and p values [32].

Osteoblast differentiation and mineralization assays
At the end of the co-culture period of IDG-SW3 with TG cultures (14 days) from either Nav1.8-Cre or nociceptor ablated mice (n = 6 wells/group), the percentage of DMP1-expressing cells was quantified following the detection of GFP-positive cells using an EVOS cell imaging system microscope (Life Technologies) at 20X, following nuclear staining with NucBlue™ Live ReadyProbes™ (Life Technologies).
After 7 days co-culture of MC3T3 cells with TG neurons, formation of hydroxyapatite was detected using the Osteoimage™ mineralization assay (Lonza Biosciences) following the manufacturer recommendations. In addition, calcium accumulation was detected and quantified by Alizarin Red staining using the osteogenic mineralization assay (Millipore-Sigma). Further, alkaline phosphatase activity was measured using the fluorometric alkaline phosphatase assay kit (Abcam) following manufacturer's recommendations. Lastly, conditioned media were collected on the final day of culture, and secreted osteocalcin was quantified using an osteocalcin enzyme-linked immunosorbent assay (ELISA) (MyBioSource). For all experiments, cell viability relative quantification was determined using OZBlue Cell Viability Kit (OZ Biosciences Inc). Fluorescence (Osteoimage and Alkaline Phosphatase (ALP) assays) and absorbance (Alizarin Red Staining and ELISA) of the samples was detected and quantified in a FlexStation 3 Benchtop Multimode Microplate Reader (Molecular Devices). All experiments included 6-8 samples from at least 3 separate experiments. In addition, representative 20 × magnification images were acquired by brightfield and epifluorescence (488 nm wavelength) microscopy at 20 × magnification using an inverted microscope (EVOS cell imaging system).

In vitro osteoclastogenesis co-culture assay
A bone resorption fluorescent assay kit (Cosmo Bio) was used according to manufacturer's recommendations. Murine osteoclast precursor RAW264.7 cells were plated in the lower chamber of 24-well plates coated with fluorescently labelledcalcium phosphate at the density of 1 × 10 4 cells/well in growth medium containing αMEM (Life Technologies), 10% heat inactivated-FBS (Life Technologies), and 1X Glutamine-Penicillin-Streptomycin (Invitrogen) and were allowed to attach for 24 h in 5% CO 2 and 37ºC. Next, Receptor activator of NF-κB (RANK) ligand (RANKL; 100 ng/ml) (Sigma) and liposaccharides (LPS) from Escherichia coli (100 ng/ml) (InvivoGen, CA) were added to the cells to stimulate osteoclast differentiation [33,34]. Cultured primary TG neurons in inserts were transferred to 24-well plates containing the induced RAW 264.7 cells and co-cultured for 5 days at 37 °C and 5% CO 2 (n = 7 wells/group). In addition, RAW264.7 in the presence or absence of RANKL and LPS were cultured to serve as positive and negative controls, respectively. Released fluorescently labeled calcium phosphate in the media supernatant was detected and quantified in a multimode microplate reader FlexStation 3 (Molecular Devices).
Osteoclast differentiation was assessed by quantification of multinucleated cells following nuclear staining of RAW264.7 cells with NucBlue Live Cell Stain (Invitrogen) after 5 days of co-culture. Images at 40 × magnification of fluorescent (350 nm) were acquired and cells with distinct three or more nuclei were manually quantified (n = 8 wells/ condition) using an inverted microscope EVOS cell imaging system.

Statistics
All experiments were performed at least in triplicate. For μCT analysis, mean volume measurements from each of the exposed tooth were normalized to the contralateral control tooth and data were presented as % void volume. For IHC, mean immunoreactive cell numbers from each sample were normalized to their contralateral control (n = 9-12/time point). For the osteoblast and osteoclast function assays, the data were normalized to results from control group (without co-culture).
Statistical significance in differences of gene expression between control and ablated mice was tested using student's t-test. One-way analysis of variance (ANOVA) with Bonferroni's post hoc test was used to evaluate the behavior and in vitro mineralization and resorption assay data. Two-way (ANOVA) test with post hoc Sidak's test was used to evaluate statistical significance of the osteolytic lesion area. Unpaired t-tests were used to analyze the RNAscope and immunohistochemistry quantification of macrophages, lymphocytes and RANKL/Osteoprotegerin (OPG) ratio. Statistical significance was set at p < 0.05. All statistical analyses were performed using GraphPad Prism 7.0 (GraphPad).
Dental infection resulted in a robust cellular infiltrate at the inflammatory site in control mice ( Fig. 1e; left) which increased with marked greater osteolytic area in Nav1.8-DTA mice ( Fig. 1e; right). Next, the innervation status of the AP lesions and the contralateral healthy PDL space was evaluated by examining the expression of CGRP-immunoreactive (CGRP-IR, peptidergic nociceptor marker) fibers. Healthy teeth have rich innervation of nociceptors in Nav1.8-Cre mice that is substantially increased 14 days following infection due to neuronal sprouting ( Fig. 1f; left). However, there was marked reduction in the density of CGRP-IR fibers in bone surrounding healthy teeth and within the AP lesion at 14 days following infection in the Nav1.8-DTA mice ( Fig. 1f; right).
The ratio of cells expressing RANKL and OPG has been widely used to denote the status of dynamic bone metabolism [35]. Commonly, a high ratio of RANKL/OPG represents active osteolytic activity. In IHC studies, there was a significant higher RANKL/OPG ratio within apical lesions in Nav1.8-DTA animals at 3 days post pulp exposure (~ twofold) (p = 0.01) (Fig. 3e, f). While no significant difference in the RANKL/OPG ratio was found between the two groups at 7 days (p = 0.15) (Fig. 3e), Nav1.8-DTA mice showed downregulation in the RANKL/OPG ratio compared to Nav1.8-Cre mice (~ two-fold) at 14 days (p < 0.01) (Fig. 3e).

Discussion
Apical periodontitis is a prevalent inflammatory disease resulting from dental infection that affects millions worldwide [38]. Pain, loss of function, inflammation and localized bone loss are hallmark features of this cranio-skeletal disease. Although human teeth are one of the tissues most densely with nociceptors [39], the function of the specialized sensory is believed to be mainly sensorial and their role in regulating the cellular events in apical periodontitis is largely unknown. In this study, using a transgenic approach to achieve the conditional ablation of nociceptors (marked as Nav1.8 + neurons), we found that nociceptors regulate early stages of the apical periodontitis initiation and progression by suppressing an initial immune response, which are seen as reduced influx of macrophages and lymphocytes and lower expression of IL-1⍺ and IL-6, promoting osteoblast differentiation and inhibiting osteoclast differentiation and its mediated resorption.
We had previously demonstrated that pharmacological ablation of a subpopulation of nociceptors (TRPV1 + neurons) resulted in accelerated bone loss in a model of apical periodontitis in rats [8]. It is well established that bone homeostasis is result of the net function between osteoblasts and osteoclasts, and that this balance can be shifted towards greater resorption due to inflammation. In this study, a transgenic approach allowed for the targeted ablation of nociceptors to evaluate their participation in the cellular events that mediate bone loss in apical periodontitis. This approach prevents possible off-target effects of large doses of capsaicin in neonatal rats [18]. In addition, this transgenic approach is well-characterized and successfully used in investigate the participation of sensory innervation in different biological processes [18,40,41]. Nav1.8-DTA mice and the control mice showed normal development with no aberrant systemic or dental manifestations including similar bone density and morphology. The successful ablation of a large subset of nociceptors was confirmed by behavior assays with mice displaying longer paw withdrawal latencies to radiant heat stimulus, and significantly less time spent in nocifensive behavior upon application of a capsaicin solution in the eye. At the molecular level, there was a dramatic loss of Nav1.8 and CGRP mRNA in the trigeminal ganglia, representing a significant loss of nociceptors since Nav1.8 is preferentially expressed in nociceptors [17] and CGRP is expressed primarily in a subpopulation of nociceptors (peptidergic) [15]. At the cellular level, the loss of innervation could be confirmed by immunohistochemistry with dramatic reduction of CGRP staining in nerve fibers innervating the periodontal ligament between teeth and the surrounding bone. This innervation pattern is known to have considerable plasticity with robust sprouting of nociceptors within the site of apical periodontitis following infections [22]. This sprouting was indeed observed in control mice but largely abolished in Nav1.8-DTA mice seen as loss of CGRP-positive nerve fibers. Collectively, the results demonstrated that the used transgenic approach resulted in great reduction of nociceptors that could be observed within the tissue at the site of apical periodontitis.
Models of infection-induced apical periodontitis have been widely used to study the initiation, progression and maintenance of apical periodontitis in mice [42,43]. The kinetics of bone loss has been previously reported with the use of either histology or micro-CT as dependent measure [44]. Similar to previous studies, we found a steady increase in osteolytic activity in control mice with a plateau of maximum bone loss after 21 days following pulp exposures to the oral environment (infection). Ablation of nociceptors resulted in larger osteolytic lesions osteolytic lesions developed at earlier time points as detected by histochemical staining and micro-CT.
The development of AP lesions is marked by an increase in inflammatory mediators, including IL-1⍺ expression that is found increased in both animal models [45] and human tissues [46], and that bone resorption is significantly blocked by infusion with neutralizing antibodies against IL-1⍺, confirming its pivotal role in mediating bone resorption in this disease [45]. In this study, we also detected a significant increase of IL-1⍺ expression that was dramatically increased in Nav1.8-DTA mice. Similarly, IL-6 is a pro-inflammatory mediator that has been implicated in apical periodontitis with levels directly proportional the sizes of the osteolytic lesions in both animal models and human tissues [47]. We found a robust upregulation in IL-6 mRNA expression that was dramatically increased in nociceptor ablated mice. This increase in the inflammatory status in apical periodontitis seen in nociceptor ablated mice was also observed by an early increase macrophages and lymphocytes detected by both immunohistochemistry and RNAscope in situ hybridization. Macrophages are important specialized innate immune cells which constitutes the major cell population in AP lesion and are known to synthesize and release pro-resorptive cytokines [6]. They are known to mediate key events during the early defense mechanism and during transitioning to adaptive response by activating T-lymphocytes [5]. In a similar manner, the anticipation of T-lymphocytes increase seen with nociceptor ablation may underlie an important mechanism since they are known secrete RANKL early during AP development [48]. Interestingly, a study using the transgenic ablation of nociceptors to investigate their function in regulating immunological responses in lung infections also found increased inflammatory markers and greater immune cell infiltrate following nociceptor ablation [49]. Thus, nociceptors inhibit macrophage and lymphocyte infiltration in the early stages of AP development and downregulate the expression of IL-1⍺ and IL-6.
Osteoclastogenesis and bone resorption is initiated and maintained by the agonist action of RANKL on its receptor in osteoclast precursor cells. Several inflammatory mediators, also known to be present in AP, have been shown to be co-activators of this process such as TNF-α and IL-1 [50]. This effect is counteracted by OPG that acts as a decoy soluble receptor sequestering available RANKL. Thus, the overall resorptive activity is highly dependent on RANKL/ OPG ratio [51]. Indeed, we found concomitant upregulation of RANKL/OPG expression with the phase of active osteolytic activities of the lesion. This is in agreement with previous study evaluating the levels of RANKL/OPG ratio in relation to the bone destruction and the expression of proresorption inflammatory mediators [35]. They also found that in naïve animals, the RANKL/OPG ratio peaked at 14-21 days post-AP induction and then decreased afterward during chronic stage. However, this increase in RANKL/ OPG ratio was significantly anticipated in Nav1.8-DTA mice with peak seen at 7 days agreeing with the quantification of bone loss in ablated mice and correlated with the increased early inflammatory infiltrate and the increase in TRAP mRNA detected by in situ hybridization. Further, the direct modulation of neurons on osteoclasts was investigated by co-culture experiments of RAW264.7 cells with trigeminal neurons using an established bone resorption assay [34]. In this assay, RANKL and lipolysaccharides (LPS) were used to stimulate the osteoclastic differentiation while in presence of trigeminal cultures from control or Nav1.8-DTA mice. It is noteworthy, that besides promoting RAW264.7 differentiation in vitro [52], LPS is one of the primary bacterial antigens in dental infections with direct correlation with symptoms and size of bone resorptive lesions in AP [53]. Our results revealed that TG neurons inhibit osteoclastic activities were inhibited by TG neurons from control mice, but this effect was partially reversed in co-cultures with TG from Nav1.8-DTA mice. This effect does not appear to be due to reduced differentiation since there was not a difference in number of multinucleated cells between both co-cultures. Instead, the results represent a reduction in osteolytic activity mediated by soluble factors released from neuronal cultures in the semipermeable culture inserts. Although the neuronal derived factors mediating this effect were not identified, there is evidence that CGRP suppresses the differentiation and function of osteoclasts [54]. Collectively, we demonstrated that nociceptors downregulate osteoclastic function.
There is increasing evidence that support an intimate cross-talk between sensory innervation and osteoblasts in both bone homeostasis [55] and in pathological processes that result in the dysregulation of bone turnover [56]. In this study, an early increase in the osteoblastic precursor RUNX2 mRNA was detected at 3 days post infection in Nav1.8-DTA mice. Interestingly, this increase was mirrored by a decrease in osteocalcin-positive cells and the progressive increase of Ki67 + positive cells without the increase in co-expression between these two markers. These results suggest that nociceptors promote early differentiation of osteoblasts and the commitment of rapidly dividing progenitor cells (Ki67 +) [57] into an osteoblast phenotype (osteocalcin +) in apical periodontitis. Indeed there is evidence that sensory innervation can drive the commitment of undifferentiated mesenchymal cells The reporter osteoblast precursor cell line IDG-SW3 was used co-culture experiments to monitor the expression of DMP-1 by GFP, and due to its well-established genetic profile as it transition from early osteoblasts to late osteocytes [58]. Co-culture with TG neurons from control mice resulted in increased expression of DMP-1 that was dramatically decreased when cells were co-cultured with TG neurons from nociceptor ablated mice. RNA sequencing revelated that several genes related to osteoblast differentiation and function were significantly downregulated in cells cocultured with nociceptor ablated TG neurons. Notably, the downregulation of the putative osteoblast marker BGLAP (osteocalcin) and the mature osteoblast/osteocyte marker Phex [59] supports that differentiation into mature osteoblasts was halted when co-cultured with Nav1.8-DTA TG neurons. Collectively, the transcriptomic analysis of IDG-SW3 co-culture experiments shown a robust downregulation in osteoblast differentiation and genes involved in mineralization and bone regulation by Nav1.8-DTA neurons compared to cells exposed to TG neurons from control mice. Evidence of the direct stimulation of osteoblast differentiation by nociceptors was also obtained by of murine calvaria osteoblast precursors, MC3T3-E1 [60] with TG neurons. Ablation of nociceptors resulted in reduced formation of hydroxyapatite deposits after 7 days in culture and reduced detection of secreted osteocalcin. Interestingly, there were no differences in calcium deposit staining by alizarin red and a greater alkaline phosphatase (ALP) activity was detected in co-cultures with Nav1.8-DTA neurons. This increase in ALP activity detected after 7 days in presence of differentiation factors is likely related to younger osteoblasts still entering the secretory phase, known to have higher ALPL expression [61], due to their retarded differentiation in absence of nociceptors. These effects are likely mediated by an entourage of secretory factors that have not yet been identified. Altogether, these experiments demonstrate that nociceptors promote osteoblast differentiation and mineralization in vivo and in vitro.
In conclusion, our findings demonstrate, for the first time, that TG nociceptors regulate the development of AP by immunomodulation observed as suppression of immune cell infiltration and decrease in cytokine production and increase of osteoblast differentiation while reducing osteoclastic activities. This newly uncovered neuroimmune mechanism of nociceptor regulation of bone turnover due to infections has broad implications to understanding infection mediated osteolysis, and could become a therapeutic target for the treatment of AP to reduce inflammation, while maximizing osseous healing after adequate treatment and minimize the persistence of osteolytic lesions in refractory cases.