Identifying the transcriptional profile of neurons specifically associated with calvarial bone injury

Introduction: Skeletal sensory nerves play critical roles in bone healing. After injury, early nerve infiltration stimulates bone formation followed by nerve retraction to baseline levels in healing injuries. Previous studies identified bone-forming roles for nerve-secreted factors, such as calcitonin-gene related peptide (CGRP). In vivo, these factors may function indirectly (e.g. via vasculature) or directly through osteoprogenitor signaling to promote bone regeneration. In non-healing injuries, however, nerves do not retract, suggesting that nerves and nerve factor persistence may have a negative impact on bone healing. As such, understanding the complex interplay between nerves, vasculature, osteoprogenitor differentiation, and, ultimately, mineral deposition is important for developing therapeutic strategies for the treatment of non-healing bone injuries. In craniofacial bones, such as the calvaria, there has been a dearth of studies that have elucidated the neural signaling interactions that govern bone healing. One barrier has been that calvarial nerve cell bodies are located in distal trigeminal ganglia which innervate multiple craniofacial regions and tissues in addition to bone, resulting in difficulty identifying bone-specific nerves that secrete factors to signal osteoprogenitors in the defect region. To address this challenge, we propose to use viral retrograde tracing to label, and subsequently isolate, nerve nuclei specifically associated with calvarial sub critical- and critical-sized defect injuries to temporally investigate nerve phenotypes and interactions with osteoprogenitors during bone healing.

Materials and

Methods: We performed sub critical- (1-mm) and critical-sized (4-mm) calvarial defect injuries on the right parietal bones of C57BL/6J mice using a trephine drill bit without disturbing the underlying dura mater (Figure 1A). Following injury, we harvested calvaria from post fracture day (PFD) 7-112 to observe infiltration and retraction patterns of calvarial nerves at early and late timepoints after injury. We used quantitative light sheet microscopy (QLSM) and confocal imaging to characterize the densities of TUBB3+ and NeuF+ nerves in both defects, as well as the densities of Osx+ osteoprogenitors and CD31+ and Emcn+ vasculature. To investigate osteoprogenitors differentiation and proliferation patterns, we performed immunofluorescence staining for osteocalcin and Ki67. To investigate neuroskeletal signaling interactions between healing and non-healing defects, we developed a retrograde tracing method combined with single cell RNA-sequencing (scRNA-seq). To identify the timing required for complete tracing, AAV-PHP.S (tdTomato+) (Addgene #59462) was injected with Fast Blue stain into the right parietal bone. Ipsilateral and contralateral trigeminal ganglia were imaged via whole mount confocal imaging after injection. To investigate the baseline transcriptional signatures and cell-cell interactions, the retrograde tracing and scRNA-seq workflow was performed on uninjured parietal bones and their corresponding trigeminal ganglia. Subsequent studies will harvest cells from the trigeminal ganglia and sub critical- and critical-sized defect regions for scRNA-seq. All statistical tests were performed in GraphPad Prism with either a two-way ANOVA with Bonferroni’s post-hoc test or a two-tailed t-test.

Results, Conclusions, and Discussions: At early timepoints following injury, nerves rapidly infiltrate into 1-mm and 4-mm defects at similar densities; however, at late timepoints (PFD 56 and PFD 112), TUBB3+ and NeuF+ nerves retract to baseline in 1-mm defects yet fail to retract in 4-mm defects (Figure 1B-D). CD31hiEmcnhi blood vessels and Osx+ osteoprogenitors are also elevated in 4-mm defects at PFD56 (Figure E). Immunofluorescent staining of defect regions revealed significantly reduced OCN+ staining, suggesting reduced differentiation of osteoprogenitors in non-healing defects, regardless of their higher density (Figure F). AAV-PHP.S tdTomato+ retrograde tracing initially reaches the trigeminal ganglia at D3 but continues to increase in intensity until D14 and remain constant through D21 (Figure 1G). Following 10X Genomics sequencing of uninjured parietal bone and ipsilateral trigeminal ganglia, 19,552 parietal bone and 16,492 trigeminal ganglia cells were processed using the Seurat platform. We identified 20 clusters, including nerves, Schwann cells, mesenchymal cells, endothelial cells, and immune cells (Figure 1H). In addition, we visualized a unique population of tdTomato+ calvarial neurons (Figure 1I). Following ligand-receptor analysis, we identified the top interactions between tdTomato+ nerves and parietal bone mesenchymal cells, which included FGF1-FGFR1.

Given that nerves did not fully retract in 4-mm defects yet are known to be required at early stages of healing and that osteoprogenitors were also elevated in density, but do not exhibit high levels of differentiation, we hypothesized that nerves encourage osteoprogenitor proliferation, maintaining stemness and inhibiting osteogenic differentiation. Thus, to further understand the role of calvarial nerves in coordinating bone formation, we developed a high throughput method to identify calvarial nerve transcriptional signatures using retrograde tracing and scRNA-seq. We identified a unique tdTomato+ nerve population which secrete neuroskeletal signaling factors that may modulate osteoprogenitor proliferation and differentiation. Specifically, neural FGF-1 signaling has been shown to mediate mesenchymal stem cell proliferation in tooth homeostasis and is also known to activate SOX2 expression and maintain stemness in osteoprogenitors, inhibiting their differentiation. Thus, we aim to further validate these signaling interactions in defect models to confirm whether critical-sized defect healing can be activated through late inhibition of nerve-derived factors.

Acknowledgements (Optional):