Divide-and-conquer strategy with engineered “ossification center” organoids for rapid bone healing via recruiting developmental cell community

Divide-and-conquer strategy with engineered “ossification center” organoids for rapid bone healing via recruiting developmental cell community | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Help Center Sign In Submit a Preprint Cite Share Download PDF Article Divide-and-conquer strategy with engineered “ossification center” organoids for rapid bone healing via recruiting developmental cell community Hongwei Ouyang, Xianzhu Zhang, Wei Jiang, Xinyu Wu, Chang Xie, and 15 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5248946/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 04 Jul, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract Current approaches for bone repair are often focused on local delivery of growth factors that are aimed at coupled angiogenesis and osteogenesis. However, delayed revascularization and regeneration of severe bone defects are still challenging. In this study, we engineered an ossification center-like organoid (OCO) that consist of inner-core bone morphogenetic and neurotrophic spheroid generated via MSCs-loaded 3D printing, alongside the interstitially distributed outer-shell proangiogenic neurotrophic phase. Our results demonstrate that collective implantation of OCOs achieved rapid bone bridging with successive OC-like bone ossicles formation across the bone defect in a “divide-and-conquer” way. Single-cell RNA sequencing analysis unveiled a developmentally mimicking stem cell community that dominated with Krt8 + skeletal stem cells (SSCs) was uniquely recruited by the pro-regenerative in-situ organoid fusion and maturation. Particularly noteworthy is the specific expansion of Krt8 + SSCs concomitant with the simultaneous reduction of Has1 + migratory fibroblasts (MFs) 2 weeks post-OCO implantation. Furthermore, cross-species comparisons employing machine learning revealed high resemblance of relative Krt8 + SSCs/Has1 + MFs composition in bone regeneration with that in public data from developmental bone tissues. Our findings advocate an approach akin to “divide-and-conquer” utilizing engineered OC-like organoids for prompt regeneration of large-sized bone defects. Biological sciences/Biotechnology/Regenerative medicine Biological sciences/Biotechnology/Tissue engineering Physical sciences/Engineering/Biomedical engineering Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction As large-sized bone defects exceed the limit of self-repair ability, the subsequent long-term recovery often end up with either delayed union or nonunion 1 . Although the transplantation of autologous bone or artificial grafts is common in the clinic, the donor-site morbidity, infections and poor osteointegration still exist. Synthetic materials that developed for bone repair have resulted in limited potential for vascularization and bone growth within the non-biodegradable bulk constructs 2,3 . Biomedical strategies that mimic the mature bone via the delivery of stem cells or growth factors at high-dosages have generated inconsistent results due to limited cell viability, insufficient vascular supply, and inevitable ectopic bone formation 4 . Thus, there is an urgent demand to develop alternative approaches that are capable of long-term retention of resorbable biomaterials, potent stem cells, and physiologically-relevant soluble factors. Developmental tissue engineering has recently gained much attention through the design of cell-based constructs with the concept of developmental biology for harsh regenerative conditions 5 . Stem cells as well as morphogenetic factors emerging during the natural healing process have been considered for generating an intermediate avascular cartilaginous anlage, and producing new bone through endochondral ossification in vivo 6-8 . Mesenchymal stem cell (MSC) aggregates are capable of forming cell spheroids and producing a mineralized matrix and a bone marrow hematopoiesis-supporting niche; however, the scaffold-free constructs require either extended in vitro chondrogenic induction or native tissue adhesion for their in vivo application 9,10 . Functional biomaterials enable large-scaling of micrometer-sized assemblies and support biophysical cues and molecular signals for sequential endochondral ossification 11,12 . A recent study from our laboratory has developed an injectable microsphere-based callus-like organoid that highly recapitulates the diverse cell compositions and behaviors of the developing bone via digital light-processing (DLP) printing technology 13,14 . While stem cells within the microsphere underwent simultaneous chondrogenesis and osteogenesis and were protected during local delivery, the interstitial pore structure of the organoid further directed native ECM alignment and blood vessel formation, leading to rapid progression of endochondral regeneration 13 . Obviously, the above strategies require high-cost pre-culture in vitro with additional growth factors to ensure the maturation of chondrogenesis before implantation. Bone morphogen presentation in cellular constructs allows for in situ chondrogenic priming and enables ready-to-use implantation without additional in vitro induction. Local delivery of supraphysiological dosages of recombinant human (rh) bone morphogenetic protein-2 (BMP-2) is one of the most popular treatments for large bone defects and is limited by severe ectopic bone formation. Interestingly, developmentally mimicking bone organoids that incorporate exogenous stem cells and BMP-2 induce less ectopic bone formation than BMP-2–containing collagen sponges do 8 , thus providing a promising system for investigating their role in rapid bone regeneration through endochondral ossification in vivo . Emerging studies have been focusing on the optimization of sustainable co-delivery of angiogenic and osteogenic growth factors by integrated scaffolds for bone repair 15-17 . However, it still involves with delayed revascularization and deficient bone formation upon critical-sized defects 16,17 . Our previous work revealed that neurotrophic factors can efficiently promote the in situ recruitment and expansion of the endogenous Msx1-positive skeletal stem cell subpopulation, which dominates the full-thickness bone reconstruction through an endochondral pathway 18 . These findings indicate that neurotrophins have crucial potential for overcoming the bottleneck of regeneration of critical-sized bone defects. Recent studies have reported the role of neurotrophin-directed innervation in promoting vascularization and osteochondral lineage differentiation during the development of both primary and secondary ossification center 19,20 . The endochondral bone is widely and richly innervated by sensory nerve fibers, and nerve dependence, as a persistent determinant in the stem cell niche, is required for bone fracture repair 21 . As reinnervation precedes angiogenesis and osteogenesis during bone repair, disruption of nerve growth factor (NGF)-TrkA signaling strongly bluntes the revascularization and ossification during the bone fracture healing 22 and calvarial regeneration 23,24 . Bio-printed constructs delivering NGF that mimic the ossification center microenvironment for targeted reinnervation have reconstructed the neuro-vascularized networks for subsequent bone regeneration 25 . Moreover, neuropeptides, including Calcitonin Gene-related Polypeptide (CGRP) 26,27 , Substance-P (SP) 28 , Neuropeptide Y (NPY) 29 , Vasointestinal Peptide (VIP) 30 and Neuropeptide VF (NPVF) 31 , are involved in tissue injury repair 28 . Among them, CGRP is known to be a potent amplifier of osteogenic differentiation 26 , whereas SP is capable of recruiting endogenous MSCs 28,32 and promoting angiogenesis 33 . Nevertheless, the bone healing speed and full-thickness regeneration of critical-sized defects are still challenging because of the lack of developmentally biomimetic biomaterials designed for sequential tissue morphogenesis with superimposed neuropeptides. Engineered bone grafts have emerged as alternatives for providing as bioactive carriers with flexible combinations and programmed release patterns of soluble factors that are required for the progress of natural bone repair 16,17,34,35 . However, the insufficient cell-adhering surface of implants and the inconceivable cell migration distance of large injury sizes for endogenous repair greatly limit the in-situ cell proliferation and osteogenesis. Fortunately, the 3D printing technique has endowed the advantages in the manufacturing of microcarriers that encapsulate single cells in thin tunable microgels with spatiotemporally defined patterns of growth factors, as well as abundant cell-adhering surfaces for bone regeneration 36 . In addition, the recruitment and migration of endogenous cells are vital for both niche remodeling and successful bone regeneration in critical-sized defects. However, the efficiency of the microgel-based bioactive cellular constructs as well as the underlying cellular mechanisms at single-cell resolution of such strategies for bone repair remain elusive. In this study, we first demonstrated that CGRP synergistically induced significantly increased levels of osteogenic differentiation in combination with a physiological dose of BMP-2. We report the availability of the “divide-and-conquer” strategy with multiple spherical microgel-based organoids for effective bone regeneration. We showed that the OC-like organoid (OCO) empowered with bioactive neuropeptides achieved long-term retention and fast bone bridging with successive OC-like bone ossicles across the bone defect. Single-cell RNA analysis revealed that a developmental stem cell community, characterized by unique skeletal stem cell (SSC) and niche cell compositions, was orchestrated by the OCO. Notably, the Krt8 + SSCs was specifically expanded while the injury-responsive Has1 + migratory fibroblast subset was concrurrently reduced following OCO implantation. Moreover, cross-species comparisons via machine learning revealed high similarity in Krt8 + SSCs activation between OCO-dominated bone repair and developmental bone structures. In summary, this study illustrates developmentally inspired “ossification center” organoids harnessing endogenous stem cell subpopulations for rapid bone regeneration. Results CGRP synergistically promoted osteogenesis with BMP-2 at physiological dosage. The optimized concentrations for neuropeptides were determined in cell cultures, and no obvious difference in cell proliferation was detected compared with that in the regular growth medium ( Supplementary Fig. 1a-d ). To investigate the role of neuropeptides in promoting osteogenesis, we performed cell differentiation experiments under osteogenic inductions, each supplemented with distinct neuropeptides that were reported to be secreted by peripheral nerves 37 . While VIP and NPY supplementation did not have obvious advantages over the regular osteogenic medium, CGRP, on the other hand, significantly increased osteogenic differentiation, as shown by both ALP/ARS staining ( Fig. 1a and Supplementary Fig. 1e ) and ALPL gene expression ( Fig. 1b and Supplementary Fig. 1f ) after 7 days and 14 days, respectively. To further confirm whether the MSC phenotype was maintained by CGRP, we tested the expression of the cell proliferation marker Ki67 and the cell senescence marker γh2AX. Immunostaining showed that the expression of Ki67 was significantly increased compared with that in the control medium, whereas the expression of γh2AX was significantly reduced after CGRP treatment ( Supplementary Fig. 1g, h ). We further investigated the synergistic effect of the combination of CGRP (10 -8 M) with BMP-2 at a physiological dosage (0.5 μg/ml) on the osteogenic differentiation of MSCs. Results showed that CGRP significantly improved the ALPL and RUNX2 gene expression in MSCs on the basis of BMP-2 after both 7 days and 14 days of osteogenic induction ( Fig. 1c-f ). ALP staining at day 7 ( Fig. 1g, h ) and ARS staining at day 14 ( Fig. 1i, j ) also showed that the enhanced osteogenic differentiation capacity of MSCs induced with CGRP and BMP-2 was greater that of those cultured with either CGRP or BMP-2 alone, indicating the strong osteo-inductivity. The above results showed that CGRP and BMP-2 at low dosages synergistically maximized the osteogenic capacity of MSCs. “All-in-one” bioprinting enables efficient engineering of “ossification center”-like organoid with dual-modular neuropeptides. As innervation has been reported to be indispensable for the development of ossification center and fracture repair through endochondral ossification, in addition to vascularization and mineralization, neuropeptides are hypothesized to potentiate the effects of bone morphogenetic factors. To engineer modular cell constructs that bio-mimic developing ossification centers (OCs), we fabricated “all-in-one” (AIO) spheroids from MSCs-containing 3D-printed microspheres (2×10 6 cells/mL of bioink), together with loading of both CGRP (10 -8 M) and BMP-2 (1 μg/ mL) ( Fig. 2a and Supplementary Fig. 2a ). Microspheres with a diameter of 100 μm showed limited spherical formability and cell encapsulation ( Supplementary Fig. 2b ). Considering the difference in nutrient exchange between cells and their microenvironment, we focused on characterizing cell viability and expansion. These results showed that the microspheres with a diameter of 400 μm had a significantly higher ratio of cell survival and capacity for subsequent cell expansion than did those with a diameter of 200 μm ( Fig. 2b and Supplementary Fig. 2c-e ). To test the synergistic effect of CGRP and BMP-2 on the cell growth and osteogenic differentiation of MSCs in the microspheres, the same initial number of cells was encapsulated in the AIO spheroids either loaded with BMP-2 alone or combined with CGRP ( Supplementary Fig. 2f ). The growth of the cells in the microspheres was observed and assessed by live/dead staining at day 1, 4, 7 and 14. These cells were distributed evenly in the microspheres within the first 7 days, and there was an almost 5-fold increase in the cell number by day 14 in all three groups, with high cell viability and very few dead cells ( Fig. 2c ). The number of cells that expanded during the 14 days was quantified by Imaris, and there were no significant differences in the percentage of proliferating MSCs among the three groups ( Fig. 2d ). Together, the AIO spheroids with a diameter of 400 μm show high biocompatibility and are beneficial for cell assembly. To more directly identify the osteogenic potential of the rapidly expanded MSCs in the AIO spheroids, osteogenic induction was performed after 1 day of pre-treatment in growth medium ( Fig. 2e ). Compared with those in the control or BMP-2-supplemented osteogenic medium, the levels of ALP synthesis and ARS, as well as their inner distribution, were greatly enhanced in the AIO spheroids( Fig. 2f, g ). Consistent with the increased ALP levels and mineralized nodule formation shown by ARS staining, the expression of osteogenic markers, such as RUNX2 and OCN, was also obviously promoted in AIO spheroids after 7 days ( Fig. 2h ) and 14 days of osteogenic induction ( Fig. 2i and Supplementary Fig. 2g ), respectively. After 2 weeks of induction, a much higher percentage of RUNX2- and COL1-posive cells were observed at the early stage, demonstrating increased sensitivety and rapid osteogenesis in response to the synergistic effect of CGRP and BMP-2 ( Fig. 2h, i and Supplementary Fig. 2h, i ). Collectively, these data suggest that the AIO spheroids not only enable high cell viability, but also serve as ideal units for “ossification center”-like organoid assembly with highly efficient osteogenesis. The “divide-and-conquer” strategy achieves fast bone bridging with collective implantation of OC-like bone organoids. GelMA and GelMA/HA-NB hydrogels were synthesized and the differential degradation was tested by subcutaneous implantation ( Supplementary Fig. 3a ). Compared with the GelMA hydrogel, GelMA/HA-NB obviously had a much slower in vivo degradation rate ( Supplementary Fig. 3b, c ), while both hydrogels showed comparable biocompatibility within 4 weeks ( Supplementary Fig. 3d ). To test the ability of the combined hydrogels as a dual-modular drug delivery system, via DLP 3D printing, we developed “inner-core” GelMA/HA-NB microspheres capable of incorporating growth factor A and a uniformly wrapped “outer-shell” module with the GelMA hydrogel containing growth factor B ( Supplementary Fig. 3e ). Because of they are crucial for the ossification center development, BMP-2 (5 μg/mL) and VEGF (5 μg/mL) are capable of distinctly spatiotemporal release at physiological dosages after the integrated incorporation, largely due to their different hydrogel degradation kinetics ( Supplementary Fig. 3f ). Interestingly, the release of VEGF was much faster than that of BMP-2 in the first 2 days, and both VEGF and BMP-2 started to be released slowly and continuously for up to 28 days ( Supplementary Fig. 3g ). To further verify their potential in directing cell migration, we performed live/dead staining to identify MSCs that were seeded on the surface of this dual-modular bioactive construct. The results showed much closer cell-cell contacts and deeper migration distances after the incorporation of BMP-2 and VEGF, than after the addition of the pure hydrogel vehicle ( Supplementary Fig. 3h, i ). Inspired by the spatiotemporal release pattern of the dual-modular hydrogel system in our previous results, we broadened our horizon by adopting novel neuropeptides for recruiting endogenous stem cells to strengthen the new bone formation based on the AIO spheroids. As we have introduced previously, SP is known for recruiting endogenous MSCs to local injuries 28 . Interestingly, after supplementation with growth medium for MSC culture, SP at various concentrations resulted in increased levels of cell proliferation ( Supplementary Fig. 4a ). Among them, SP with a concentration of 1 × 10 -7 M promoted a significantly higher level of VEGF secretion than the other concentrations did ( Supplementary Fig. 4b ), indicating that SP increased the proangiogenic potential of MSCs. Further validation on day 3 via real-time qPCR showed that SP significantly promoted the VEGF gene expression in MSCs ( Supplementary Fig. 4c ). We upgraded the dual-modular hydrogel system with the “outer-shell” module containing SP ( Supplementary Fig. 4d ), which was then coated on a 3D-printed PLA scaffold for cell migration assessment ( Supplementary Fig. 4e ). MSCs that were seeded on the top surface of the integrated bone graft resulted in a larger cell spreading area and more closely intersected cell-cell contacts, as well as a much longer migration distance after SP supplementation ( Supplementary Fig. 4f ). By subcutaneous transplantation, SP obviously promoted the local cell recruitment and red blood cell aggregation, whereas MSC encapsulation further accelerated the effect of SP on vascularization ( Supplementary Fig. 4g ). Moreover, immunostaining of interstitial tissue among the microspheres demonstrated enhanced blood vessel formation with CD31 expression and extensive osteogenesis with RUNX2 when the microspheres were loaded with both “inner-core” MSCs and “outer-shell” SP ( Supplementary Fig. 4h ). The above results indicated that the spatial-peripheral release of the neuropeptide SP could further enhance the cell-recruiting and proangiogenic capacity of the cellular construct for the design of OC-like organoids. To evaluate their potential for bone regeneration, we transplanted the integrated “ossification center”-like organoid (OCO) consisting of AIO spheroids and free SP into a calvarial defect in SD rats using in-situ photo-crosslinking ( Fig. 3a and Supplementary Fig. 5a ). In addition to OCO, the dual-modular hydrogel (Vehicle) and the combination of MSCs and Neurotrophins (Hybrid) were used as parallel controls ( Fig. 3b ). 4 weeks after surgery, only mild new bone formation was observed at the edge of the defect in both the Vehicle and the Hybrid groups, whereas OCO transplantation resulted in nearly complete bone bridging along with honeycomb-like structures ( Fig. 3c ). At 8 weeks post-surgery, although diffusive bone blocks had formed within the defect in the Vehicle group and a much thicker neo-bone had appeared at the inner surface of the defect in the Hybrid group, a large amount of defect area still existed to be filled. Consistent with the results in the early stage, new bone formation by OCO was shown to essentially complete the regeneration of the entire defect area ( Fig. 3d ). Significantly higher levels of bone volume, bone mineral density, as well as new-old bone integration in the OCO group were also observed at 4 weeks ( Fig. 3c, e, f ) and 8 weeks ( Fig. 3d, g, h ), respectively. Histological analysis by Masson’s Trichrome staining revealed that, compared with the control or the Hybrid setting, treatment with the dual-modular hydrogel vehicle alone promoted the matrix reorganization and bone healing to some extent ( Fig. 3i, j ). Interestingly, by combining the advantages of microstructure and bioactive factors, OCO triggered rapid bone healing with successive OC-like bone ossicles across the defect as early as 4 weeks after transplantation ( Fig. 3i and Supplementary Fig. 5b ). Full-thickness (internal and external layers of cortical bone; middle layer of trabecular bone and the bone marrow cavity) reconstruction of the calvarial bone defect was further achieved following in-situ OCO fusion ( Fig. 3j and Supplementary Fig. 5c ). The above results suggest that this “divide-and-conquer” strategy enables fast bone bridging and full-thickness bone regeneration through in-situ fusion of multiple OC-like organoids. In-situ fusion and maturation of OCOs promote concurrent innervation, vascularization, and ossification. To determine whether in-situ OCO grafting facilitates the initiation of OC development and the subsequent osteogenesis, we further examined the biological processes and cellular events during the bone regeneration. Distinct from Vehicle and Hybrid, the OCO recruited more abundant cells and guided extensive ECM deposition around the spherical organoids, as shown by the details of the HE staining results ( Fig. 4a, b ). With the OC-like organoids fusion and development after 4 weeks, the sections harvested from repaired tissues were found to retain a cartilage-like tissue rich in proteoglycan ( Fig. 4c ), which was dynamically absorbed and replaced with bone-like tissue after 8 weeks ( Fig. 4d ), and finally turned into the bone marrow cavity, as shown by OCN and NGFR expression ( Fig. 4e, f ). The initiation and maturation of OCO were then determined by the expression of multiple cellular markers indicating sequential neurogenesis ( Fig. 4g ), angiogenesis ( Fig. 4i ) and osteogenesis ( Fig. 4k ). At 4 weeks after transplantation, there were sporadic β-III tubulin-positive cells in the injured area indicating limited reinnervation in both the Vehicle and Hybrid settings, while OCO stimulated robust aggregation and distribution of both β-III tubulin- and CGRP- positive sensory neuron cells ( Fig. 4h ). Followed by hypertrophic maturation (COLX expression) of the cartilage-like tissue ( Fig. 4c, j ), the vascular invasion that needed for endochondral ossification was also revealed by staining with the endothelial marker CD31( Fig. 4j and Supplementary Fig. 5d ). Along with the enhanced innervation and vascularization, the multi-OCO inner cores were eventually surrounded by more abundant RUNX2- and OCN-positive cells ( Fig. 4l and Supplementary Fig. 5e ) and Collagen I expression, when compared with negligeable distributions in other conditions ( Supplementary Fig. 5f ). These results strongly suggest that the in-situ OCO implantation promotes rapid bone regeneration may partly through an endochondral ossification pathway. These findings collectively indicate that the “ossification center”-like organoids (OCO), comprising the inner-core bone morphogenetic MSCs spheroids generated via DLP 3D printing and an interstitially distributed outer-shell proangiogenic module, have significant potential for preserving bioactive factors and harnessing endogenous stem cells to facilitate bone regeneration. More importantly, the in-situ fusion and maturation of OCO promote bone repair with tissue morphological changes in the early stage characterized by innervation, vascularization and ossification, rather than monotypic fibrotic tissue filling after bone injury. Distinct osteo-lineage cells organize into cell communities that have specific functions To determine the bone regenerative cell atlas in response to the abundant biological processes, we performed single-cell RNA-sequencing analysis of the cell clusters after the in-situ OCO fusion and maturation for 2 weeks after implantation. Cells from the regenerated tissues were isolated, and mRNA libraries were prepared and sequenced ( Fig. 5a ). For quality control, cells with genes detected over 300 genes detected, read counts between 1000 and 25000, and mitochondrial gene expression less than 15% were retained. After filtering, we obtained data from 39471 cells: with 11913 cells in the Defect group, 10953 cells in the Vehicle group, 8208 cells in the Hybrid group, and 8397 cells in the OCO group from 2 weeks after surgery. Following the Seurat pipeline, the data were log-normalized and then scaled. Uniform Manifold Approximation and Projection (UMAP) was calculated to visualize cell heterogeneity in reduced dimensions. Integrated analysis revealed 15 cell clusters with known cell markers that formed during the first two weeks of bone repair ( Supplementary Fig. 6a, b ). On the basis of the understanding of the single cell types in our previous bone repair study as well as in other publications, we quite completely identified a full range of cell types for bone regeneration: B/T cells (expressing Ms4a1 and Cd3g); Monocyte/Macrophage (expressing Cd68 and Adgre1); Endothelial cells (expressing Emcn and Cdh5); Osteo-lineage cells, OLCs (expressing Col1a1 and Prrx1); Pericytes (expressing Mcam and Acta2); and Neural crest cells (expressing Plp1 and Sox10) ( Supplementary Fig. 6c ). All the cell clusters are distributed among the above four conditions, namely the Defect, Vehicle, Hybrid, and OCO groups, without obvious imbalance ( Supplementary Fig. 6d ). The relative proportion of each cluster showed that several OLC clusters underwent considerable variations in response to the OCO maturation, while macrophages, CD8 + T cells and endothelial cells showed only mild differences ( Supplementary Fig. 6e ). As our original intention was to figure out the changes in the recruited osteogenesis-related stem/progenitor cells at single cell resolution, the subsequent data analysis focused mainly on these OLCs. Collectively, these data provide a single-cell atlas of the regenerative bone tissues, which enables us to reveal the cellular heterogeneity of various cell types, especially the osteo-lineage cells of interest. To further dissect the cell heterogeneity and classify the OLC clusters, we merged the OLCs (19019 cells were obtained) in all the groups for sub-clustering ( Fig. 5b, c ). The OLCs were divided into 9 novel cell subpopulations on the basis of the abundant and specific expression of osteogenic genes ( Fig. 5d ) and distinct GO features ( Fig. 5e ). These 9 OLC subpopulations were annotated, in particular, Subcluster 0: Has1 + Migratory Fibroblasts (expressing Has1, Ccl2, Ackr3 and Procr); Subcluster 1: Mfap5 hi Mesenchymal Progenitors (expressing Mfap5, Pdgfra, Fbn1 and Lrp1); Subcluster 2: Thy1 hi Skeletal Stem Cells (expressing Thy1, Cav1, Col5a3 and Cdkn2a); Subcluster 3: Krt8 + Skeletal Stem Cells (expressing Krt8, CD200, Mdk and Sdc4); Subcluster 4: Ucma + Osteochondral Progenitors (Ucma, Fmod, Col11a1 and Clec11a); Subcluster 5: Chondroblasts (expressing Ctgf, Tnmd, Sox9 and Sfrp2); Subcluster 6: Osteoprogenitors (expressing Postn, Sfrp4, Runx2 and Tnn); Subcluster 7: Lyz2 + Fibroblasts (expressing Lyz2, CD74, and Cxcl2); and Subcluster 8: Proliferating Mesenchymal Cells (expressing Ube2C, Cdkn3, and Mki67) ( Fig. 5f and Supplementary Fig. 7a ). To analyze the lineage hierarchy of these heterogenous subpopulations, pseudo-time analysis with RNA velocity was performed to explore the lineage relationships among the osteo-lineage cell subsets. These results displayed that there are three main differentiation trajectories that originate from Krt8 + Skeletal Stem Cells (Subcluster 3), Proliferating Mesenchymal Cells (Subcluster 8) and Mfap5 hi Mesenchymal Progenitors (Subcluster 1), respectively; while Has1 + Migratory Fibroblasts (Subcluster 0) is situated at the downstream of the differentiation hierarchy as a common destination ( Fig. 5g ). Further analysis using 3D RNA velocity from different spatial views also confirmed the differentiation branches originating from Subcluster 3, Subcluster 8 and Subcluster 1, as well as the downstream Subcluster 0 ( Supplementary Fig. 7b-e ). We next sought to understand how cell subclusters assemble into cellular neighborhoods that organize into multi-lineages for in-situ OCO maturation. Together with the 3D RNA velocity, which demonstrated a cluster effect with differentiation continuums: subcluster 3-4-5 ( Supplementary Fig. 7c ), subcluster 8-2-6 ( Supplementary Fig. 7d ), and subcluster 5-1-0 / subcluster 5-0 ( Supplementary Fig. 7e ), Partition-based graph abstraction (PAGA) analysis revealed unique combinations of specific cell subclusters and exhibited cellular continuity. Accordingly, these subclusters are well-organized into 4 independent cell communities (cell subclusters with common molecular features and a spatial mapping of specific cell lineage structures) that each separately featured by the expression of Sdc4 and Fmod ( Fig. 5h ); Thy1 and Cav1 ( Fig. 5i ); Mfap5 and Angptl1 ( Fig. 5j ); and Has1 and Il6 ( Fig. 5k ). These results strongly suggest that OLC subpopulations involved in the bone repair are actually of diverse origins and functionally different yet closely interlinked. The dominant role switch of cell communities with distinct transcriptional signatures and spatial distributions represents bone regeneration or nonunion To better define the above 4 cell communities (CC) with unique characteristics, we first analyzed the differentially expressed genes (DEGs) with community-level specificity showing the segregation of these subpopulations ( Fig. 6a, b ). GO enrichment analysis of the TOP 8 ranked GO terms based on the DEGs showed that CC1 and CC2 were differentially enriched with genes involved in “Respond to Reactive Oxygen Species/Chemical Stress” and “Basement Membrane”, respectively; while CC3 and CC4 were differentially enriched with genes involved in “Ossification” and “Cell-substrate Adhesion”, respectively ( Fig. 6c ). By visualizing these differences with UMAP, we found that cell migration, chemotaxis, and inflammation-associated genes, such as Has1, Ackr3, Il6 and Procr, were enriched in CC1; when fibrosis-associated and immune-regulatory genes, such as Pdgfra, Mfap5, Dpp4 and Lrp1 were enriched in CC2 ( Fig. 6d, e ); in contrast, skeletal stem/progenitor cells and osteochondral progenitors-associated genes, such as CD200, Clec11a, Sdc4, and Fmod were specifically enriched in the CC3; while osteoprogenitors-associated genes, such as Thy, Cav1, Tnn, and Postn were specifically enriched in the CC4 ( Fig. 6f, g ). To this end, we have primarily identified these cell communities from OLCs that were also have spatially distinct distributions, and defined the cellular neighbors for each cell subcluster. We further detected the CC distributions by immunostaining of regenerative tissues during the bone regeneration process. At the early stage of bone repair (2 weeks after surgery), representative CC3 maker (CD200) and CC4 marker (Thy1) were both located interstitially among OCO grafts in the central layer of the regenerated tissues ( Fig. 6h and Supplementary Fig. 8a, b ). Consistent with the gene expressions observed in the UMAP among the 4 communities, CD200 was located in the OCO inner surface, while Thy1 was found relatively in the OCO outer surface, at 4 weeks after bone injury ( Fig. 6i and Supplementary Fig. 8c, d ). In the peripheral layer of the regenerated tissues, except for the relatively few Pdgfra (a CC2 marker) distributions on the surface, abundant CD200 and Thy1 were actively expressed in most of the migrating cells at 2 weeks ( Fig. 6j ), and clear and distinct spatial distributions were observed at 4 weeks ( Fig. 6k ). Interestingly, enormous cells marked by the expression of the CC1 marker Has1 were found nearly throughout the whole fibrotic tissue layer in the untreated Defect group, as early as in the 2 weeks post-injury. Even after 4 weeks, the Has1-positive cells still widely resided on both sides of the fibrotic tissue, showing nonunion of the bone injury. In contrast, few Has1 distribution was detected in both the upper and lower surfaces of the OCO grafts in the peripheral tissue layer during the first 2 weeks, and Has1 was hardly detected after 4 weeks of injury ( Fig. 6l ). We further compared the functional changes in the regulation of the cell communities caused by OCO implantation with those caused by untreated defects. Differential Gene Ontology (GO) analysis based on OCO/Defect DEGs (Differential Regulated Genes) at the CC level ( Supplementary Fig . 9a, b ) revealed that “Regulation of vasculature development”, “Regulation of angiogenesis”, and “rRNA metabolic process” were up-regulated in Cell Community 3 ( Supplementary Fig . 9c ); and “Response to hypoxia”, “Response to corticosteroid”, “Response to mechanical stimulus”, and “Response to transforming growth factor beta” were down-regulated in Cell Community 1 ( Supplementary Fig . 9d ). These results indicated that, in addition to the differences in spatial distribution, the cellular functions of specific cell communities also underwent changes after OCO implantation. The above findings suggested that the dominant role switch of cell communities (CC3 vs CC1) with distinct transcriptional signatures and spatial distributions could more precisely represent the state of bone regeneration or nonunion (OCO vs Defect) ( Fig. 6m ). Together, these OLC communities not only display diverse organizations of transcriptionally different cell subpopulations, but also feature specific cell subpopulations that drive their own community functionality according to the cellular ecosystem. This findings support the concept that stem/progenitor cells may adopt bone developmental characteristics based on their own functions as well as the microenvironmental cells within each cell community. The developmentally mimicking cell community that dominated with Krt8 + SSCs was orchestrated by the pro-regenerative OC-like organoids. To further test whether the OC-like organoids induced rapid bone regeneration by harnessing the endogenous skeletal stem/progenitor cells, we first focused on the differences in the cellular compositions and different modulations of the OLC subclusters. At 2 weeks post-injury, comparable cell numbers of all 9 subclusters were demonstrated among the 4 repair conditions ( Fig. 7a, b ). Remarkably, the proportion of cells in subcluster 3 (Krt8 + Skeletal Stem Cells, CC3 representative) exhibited significant increase in response to OCO implantation compared with that in the Defect, Vehicle, and Hybrid groups. In contrast, subcluster 0 (Has1 + Migratory Fibroblasts, CC1 representative) accounted for a substantial proportion in the untreated Defect group compared to all the other groups ( Fig. 7c and Supplementary Fig. 7f-i ). These results strongly suggested that Krt8 + SSCs and Has1 + Migratory Fibroblasts were predominantly recruited and expanded by pro-regenerative and non-regenerative treatments, respectively. Based on such a marked contrast in the proportion of the osteo-lineage cells under the four repair conditions, we further found that the relative ratio of the two cell proportions (Krt8 + SSCs / Has1 + Migratory Fibroblasts) could more accurately reflect the disparity in the degree of bone regeneration degree among the groups ( Fig. 7d and Fig. 3i, j ). Apparently, distinct osteo-lineage cell compositions and their relative cell proportions are orchestrated in response to various strategies for bone injury repair. To understand the biological significance of the relative ratio of the two cell proportions, we next explored the cell characteristics and differential signatures of the above two OLC subclusters. As shown in our previous results, Krt8, a known marker for keratin filaments in epithelial cells, was identified to be OLCs-specific among all cell types ( Supplementary Fig . 10a, b ) and subcluster 3-specific by comparing subcluster 3 to the rest of the OLCs ( Supplementary Fig . 10c, d ). Moreover, pseudo-time analysis of OLCs showed that Krt8 + SSCs were located in the upstream of the osteochondral lineage of the differentiation trajectory ( Fig. 5g, h and Supplementary Fig . 10e, f ), indicating that despite demonstrating an epithelial signature, the Krt8 + SSCs at the same time possess strong potential for osteochondral lineage differentiation, which supports our previous conclusion that OCO maturation promotes bone regeneration through an endochondral ossification pathway. As the progenitors of CC3, Krt8 + SSCs thus perfectly mark the developmentally mimicking OLC community, which is predominantly recruited after the pro-regenerative OCO implantation. The top 30 GO terms of Krt8 + SSCs showed bone repair-associated characteristics including “Ossification”, “External encapsulating structure”, “Regulation of vascularization”, and “Response to decreased oxygen levels” ( Fig. 7e ). Meanwhile, Has1 + Migratory Fibroblasts, which are mainly distributed at the end of the differentiation trajectory ( Fig. 5g, j and Supplementary Fig . 10e, f ), were active as the major cells of CC1 shown in the untreated bone injury ( Fig. 7c ). The analysis of the top 30 GO terms in Has1 + Migratory Fibroblasts showed that it was more likely to function as an injury-responsive OLC subcluster with a higher capability of migration, which was clearly involved in GO terms, such as “Positive regulation of cytokine production”, “Responsive to reactive oxygen species”, “Responsive to interleukin-1”, “Responsive to chemical stress”, and “Responsive to oxidative stress” ( Fig. 7f ). These findings demonstrated that the implantation of bioactive materials leads to the dominance of the corresponding “effector cell subset” in the specific cell community, which in turn can accurately depict the underlying molecular characteristics of particular bone repair scenario. In consistent with the above results, verification by immunostaining showed that the presence of Krt8-positive OLCs (CC3 representative) as well as CD200 expression (CC3 marker) were greatly increased and distributed within the spherical OCO construct compared with those in the Defect, Vehicle, and Hybrid groups ( Fig. 7g, h ). Interestingly, the Thy1-positive OLCs (CC4 representative) were more likely distributed at the interface of multiple OCO constructs, in which few of them were also krt8 positive at both 2 weeks and 4 weeks post-surgery ( Supplementary Fig . 8b, d ). In contrast, the Has1-positive OLCs (CC1 representative) were widely dispersed throughout the whole layers of repaired tissue in the Defect and Hybrid groups, while Has1 expression was significantly reduced in the treatment groups with either pure microspheres (Vehicle) or OC-like organoids (OCO) ( Fig. 7h ). Up to this point, we have shown that microspheres-based repair strategies have structural advantages over pure bioactive ingredients for equalizing the ability of all OLCs to migrate into bone injuries. In the context of cellular compositions, these findings indicate that the relative ratio of Krt8 + SSCs/Has1 + Migratory Fibroblasts could serve as a more precise indicator of rapid bone regeneration with the pro-regenerative OC-like organoids at single-cell resolution, as opposed to traditional studies that only focused on one individual cell subpopulation. Collectively, the specific osteo-lineage cell composition in a developmentally-mimicking cell community with Krt8 + SSCs expansion was cultivated by the pro-regenerative OC-like organoids during rapid bone regeneration. The obvious expansion of Krt8 + SSCs after OCO implantation prompted us to explore its contribution to the in-situ OCO maturation for new bone formation. Bone regenerative Krt8 + SSCs share high similarities to those found in the ossification center during bone development. In order to confirm the rationality of Krt8 + SSCs in presenting the developmentally mimicking OLC community, we carefully analyzed the subsets of both regenerative and developmental tissues involved in new bone formation. With the unique subset specificity of Krt8 shown in our single-cell RNA-seq data ( Supplementary Fig . 10a-d ), we found that Msx1 perfectly labels both Krt8 + SSCs and Has1 + Migratory Fibroblasts by using Krt8 + Msx1 + double positive staining ( Supplementary Fig . 11a, b ), in which Krt8 and Msx1 together better enable the identification of the osteo-lineage cell subset with high potential for osteochondral differentiation 18 , as well as its participation in the endochondral ossification pathway, as shown by Mgp and Alpl expression ( Supplementary Fig . 11b ). To our surprise, immunostaining of the most active repair area was shown in the interstitial spaces (outer area) of OC-like organoids (inner area), in which the Krt8 + Msx1 + double positive cells were mainly located at the bone-forming front ( Supplementary Fig . 11c ). To obtain evidence associated with clinical relevance, we validated the emergence of Krt8 + SSCs in human bone tissues at developmental stages by immunostaining assays. In human child phalangeal bone samples, an ossification center-like structure appeared in the anatomical fingertip and exhibited lacuna cell distribution between the bone-forming front and the immature bone epiphysis, which resembled the limb development of the secondary ossification center ( Fig . 8a ). Krt8 expression was abundantly distributed in the tip side of the bone-forming front, whereas Krt8 + Msx1 + double positive cells were aggregated in the central region of the bone morphogenetic center ( Fig . 8b and Supplementary Fig. 11d ). To more directly confirm the strong consistency of the presence of Krt8-positive cells in bone development, we further investigated the OC-like structure and spatial location of Krt8 + SSCs in the long bone area of the proximal phalanx during skeletal development ( Fig . 8c ). Even though the mature bone epiphysis has already formed, a remarkable number of Krt8 + cells have emerged on the lateral side of the growth plate. Krt8 + cells that reside in the metaphysis side are also Msx1 positive, while those in the epiphysis side are closely inserted into the growth plate cartilage, strongly suggesting the contribution of Krt8 + SSCs to osteochondral lineage cell differentiation during the development of the ossification center in long bones ( Fig . 8d and Supplementary Fig. 11e ). These findings provide novel insights into the pro-regenerative role of Krt8 + SSCs in bone formation during both the bone healing and the OC developmental stages, in which active and rapid osteogenesis occurs through endochondral ossification. Based on these results above, we utilized machine learning to compare the cross-species similarity between bone regeneration and development. We first merged publicly accessible scRNA-Seq data comprising three mouse bone development datasets 38-40 , two human embryonic bone development datasets 13,41 , and one rat bone regeneration dataset 18 ( Fig. 8e ). After homologous gene conversion, 7536 genes were selected as intersected features. In order to select more precise features for model construction, we intersected these genes with cell markers of two distinct subclusters, and finally acquired 38 features (17 genes as featured markers of Krt8 + Skeletal stem cells, and 21 genes as featured markers of Has1 + Migratory Fibroblasts) ( Fig. 8f-h ). We used our bone regeneration data in this study for model construction (80% for training and 20% for testing), Hist Gradient Boosting (HGBoost), Logistic Regression (LR), and Support Vector Machine (SVM) performed well among several models, with the HGBoost model achieving results with the highest accuracy ( Fig. 8i, j, Supplementary Fig. 12b ). We also computed the feature importance of the HGBoost model, and found that specific markers such as Has1, Krt8, and Gsn ranked at the top ( Supplementary Fig. 12a ). To predict divergent OLC subpopulations in external datasets, we applied the HGBoost model (HGBoost) to test in both the bone development data and the bone injury repair data. The prediction results showed that Krt8 + SSCs are involved in both developmental and injury repair processes, whereas the proportion of Krt8 + SSCs was much higher in human embryonic long bone at developmental stage. Interestingly, we found Has1 + migratory fibroblasts in the bone injury repair data but not in the human embryonic long bone data ( Fig. 8k-l ). We also compared data from human embryonic calvarial bone with data from rat bone injury repair, and similar results were also found ( Supplementary Fig. 12c-d ). Thus, cross-species comparisons employing machine learning revealed high resemblance of the relative Krt8 + SSCs/Has1 + MFs composition in bone regeneration with that in public data of developmental bone tissues at embryonic stages. Discussion In this work, we demonstrated that multiple engineered ossification center-like organoids empowered with dual-modular neuropeptides by 3D printing achieved fast bone regeneration in a “divide-and-conquer” way. Concurrent innervation, vascularization, and ossification with a cartilage-like tissue morphology were demonstrated during the in-situ fusion and maturation of OCO, which suggested that the rapid bone regeneration may occur partly through an endochondral ossification pathway. Our scRNA-seq study unveiled that distinct osteo-lineage cells organize into cell communities that have specific molecular functions and spatial distributions. Developmentally mimicking osteo-lineage cell compositions, which include specific expansion of Krt8 + SSCs concomitant with simultaneous reduction of Has1 + migratory fibroblasts, were orchestrated by the pro-regenerative OC-like organoids. Furthermore, cross-species comparisons employing machine learning revealed a remarkable resemblance of relative Krt8 + SSCs after OCO implantation with those in public scRNA-seq datasets of developmental bone tissues. Our findings advocate a promising approach akin to “divide-and-conquer” utilizing engineered OC-like organoids for prompt and efficient regeneration of large-sized bone defects. Current clinical treatments for large bone defects often necessitate extensive wound exposure, leading to increased risks of infection. Injectable biomaterials can be delivered to the injury site in a minimally invasive manner without compromising tissue integrity or the surrounding microenvironment 42 . The strategies of regulating the osteogenic differentiation of stem cells by biomineralized hydrogels 43 or in vitro pre-cultures of osteogenic differentiation 44 have difficulty achieving rapid bone healing and structural reconstruction of critical-sized bone defects, which usually leads to the formation of immature bone tissue. Well-known treatments for bone injury include the absorption of recombinant human BMP-2 onto collagen sponge carriers 45,46 or loading into microspheres 47 to promote the osteogenic differentiation of endogenous stem cells. The therapeutic effect is largely restricted by ectopic bone formation, osteoclast activation and soft tissue inflammation triggered by high doses and the need for additional mechanical stimulation for functional bone reconstruction 4,48 . As vascularization is crucial for early-stage nutrient supply and MSC recruitment, as well as the subsequent matrix remodeling, the combined delivery of vascular endothelial growth factor (VEGF) with BMPs by coupled activation of angiogenesis and osteogenesis 34,35 synergistically enhances bone regeneration. However, difficulties associated with their limited bioactivity (short half-life) in vivo, and the risk of supraphysiological dosage-induced abnormal vascularization and heterotopic ossification 49 have impeded their further clinical application. 50,51 Novel strategies for delivering growth factors (GFs) both with sustainable release and at low dosages to enhance bone regeneration are still under extensive investigation. Bone injury healing is regulated by the sequential signals that promote reinnervation, revascularization and ossification, indicating that reinnervation precedes angiogenesis and ossification during the formation of the bone-callus 22 . Coincidentally, neuropeptides secreted by peripheral nerves were shown to directly regulate bone growth and metabolism, and the corresponding receptors of these neuropeptides have been confirmed to be expressed in osteoblasts and osteoclasts 52,53 . Histological detection has also revealed the existence of neuropeptides from sensory, sympathetic and glutaminergic types in bone tissues 54 . The identified role of magnesium in promoting CGRP-mediated osteogenesis has suggested its therapeutic implications in the repair of osteoporotic bone fracture or other bone diseases 55 . More recently, neuro-immune interactions have shown to be beneficial for non-healing tissue, and delivery of an engineered version of CGRP accelerated wound healing and promoted muscle regeneration 56 . Interestingly, SP was proved to be an injury-inducible factor that acts early in the wound healing process to induce the mobilization of CD29 + stromal-like cells, which also occurs at uninjured sites after intravenous injection. In addition, research has shown no specific binding of biotin-conjugated SP to other tissues, such as liver, dental pulp, skin, spleen and kidney, but binding to the trabecular bone of the femur has been observed 57 . Higher doses of BMP-2 (e.g., greater than 150 μg/mL) even induce osteolysis 58 , while the lowest dose of BMP-2 that can induce bone regeneration is 2-10 μg/mL 59 . Our results in both 2D and 3D culture revealed that CGRP synergistically enhances osteogenic differentiation and cell survival in combination with BMP-2 at a physiological dosage. This finding is crucial for the fabrication of “morphogenetic center” units in OC-like organoid assemblies. Bone repair often progresses from the wound edge to the center 60 , and limited recruitment of osteo-lineage cells with osteogenic potential in critical-sized bone defects is a significant factor leading to bone nonunion 61 . This limitation hinders complete bone formation within the wound in a short timeframe, leading to scattered bone healing and fracture nonunion. One possible explanation for this phenomenon is the inability of stem/progenitor cells to migrate over long distances and the absence of a pro-osteogenic microenvironment 24 . Therefore, engineered osteogenic units are required as spatial bridges to provide a uniform adhering surface for stem cell attachment and an optimal microenvironment for endogenous new bone formation. Given the lack of in-situ osteogenic capacity upon improvement of the cell migration distance in large-sized bone defects, the development of sophisticated tissue-engineered grafts with spatiotemporal release of GFs at physiologically relevant dosages and biophysical support for spatially continuous bone regeneration is highly important. Stem cell encapsulation technology for tissue regeneration has been extensively studied and utilized as an effective way to provide a biomimetic microenvironment for cell preservation and tissue regeneration 62 . Modular microtissues, such as injectable osteogenic microtissues with stem cell incorporation by emulsification 63 , and the simultaneous delivery of osteogenic and vascularized cells that using cryogel building blocks, could potentiate the osteogenic differentiation of MSCs 64 . Furthermore, with improved mechanical properties, the opening of porous core-shell microtissue with demineralized bone matrix particles core modified with BMP-2 and a gelatin shell with BMSCs on the surface provides with another available method for the repair of large bone defects 65 . Together, these studies have shown the potential of the use of microtissue loaded with stem cells or together with GFs for bone regeneration. In the present study, “all-in-one” bioprinting enables efficient engineering of an “ossification center”-like cell construct with one shot of bone morphogenetic factors. Our results have also revealed the structural advantages of hydrogel microsphere-based repair strategies over delivery of pure bioactive GFs delivery for equalizing the migratory ability of all osteo-lineage cells to get into bone injuries ( Fig. 3i, j and Fig. 7d, g, h ). Owing to their specific cell origin and genetic similarity, organoids offer low immunogenicity and are a sustainable alternative tissue source for stem cell transplantation 66 . Recently, the concept of bone organoids has been proposed, including woven bone organoids 67 , callus organoids 68 , trabecular bone organoids 69 , and humanized ossicles 70 etc., which can be utilized to investigate the osteogenic differentiation mechanism and promote bone tissue regeneration. Bone organoids are generated through stem/progenitor cell-driven differentiation, resulting in three-dimensional micro-bone tissues with biomimetic spatial characteristics, self-renewal, and self-assembly capabilities 71 . Matrigel and synthetic hydrogels have been employed to facilitate the assembly of bone organoids in order to achieve oxygen supply, metabolic gradient regulation, and intercellular connectivity 7 . Injectable bone microtissues formed by emulsification and incorporation of stem cells can effectively provide necessary cells for repair 63 . Delivery of frozen gel microspheres containing osteogenic and angiogenic cells can enhance osteogenic differentiation 64 . Furthermore, demineralized bone matrix particles modified with BMP-2 contribute to the repair of highly mechanically deficient bone defects 42 . These studies demonstrate that an “all-in-one” strategy involving stem cells and growth factors serves as an alternative for constructing bone organoids by completing the osteogenesis in advance. However, it is still challenging to replicate the developing bone for building bone organoids solely through mimicking mature bone structure and composition; moreover, there are limitations on cell survival after in vivo transplantation. There has been increasing utilization of cellular structures designed according to developmental biology principles for repairing extensive tissue defects 9 . Along with bone morphogenetic factors, stem cells play crucial roles in generating cartilage intermediates during the bone healing process; they facilitate the recruitment of blood vessels and osteoblasts while promoting endogenous new bone formation via the endochondral ossification pathway in vivo 6,8 . Nevertheless, ongoing studies on these cell constructs are still in the early stages and are characterized by in vitro assembly of scaffold-free elements. Further exploration of the temporal regulation of sequential innervation, vascularization, and ossification in vivo following cell transplantation is needed. In this study, multiple engineered ossification center-like organoids (OCOs) were fabricated as fundamental building blocks for reconstructing bone defects via a “divide-and-conquer” strategy ( Fig. 3i, j and Supplementary Fig. 5b, c ). In addition, the in-situ fusion and maturation of OCO promote bone repair with tissue morphological changes that are characterized by innervation, vascularization and ossification ( Fig. 4h, j, l ). Given the common challenges of cellular complexity and limited specificity for cell sub-clustering in scRNA-seq data analysis of regenerative tissue samples, it is imperative to explore “cell community” perspectives that could substantiate the rationality of subcluster annotations 72 . In tumor cells with extensive heterogeneity and cardiac cell types with complex morphological structures, systematic dissection of single-cell ecosystems helps provide a much more comprehensive understanding of cell heterogeneity with common markers and spatial mapping of their neighboring cell subpopulations 72,73 . Our data revealed two osteo-lineage cell subsets that express one common marker, distinct differentiation origins, distinct cellular community features, and spatial proximities with other cell subsets, suggesting their differential roles in the formation of new bone tissue. Differences in the relative proportions of cell subsets from scRNA-seq data can provide valuable insights into biological significance, which could be statistically challenging owing to the inherent noise in single-cell data and inter-sample variability. Alok K. Maity et al. suggested that leveraging cell attribution information when defining cell communities can effectively reduce noise in single-cell data, eliminating the necessity for batch correction and facilitating the retrieval of cell states for subsequent differential abundance comparisons 74 . Besides, cell composition shifts in relation to the specific healthy or pathological conditions when inferring differentially enriched cell communities. In our study, cell communities were well identified with cell attributions, as specific OLC subset within cell communities are predominantly representative of their corresponding regenerative approaches ( Fig. 6 and Fig. 7 ). Skeletal stem cells (SSCs) subpopulations have been extensively investigated in the past decade with the application of scRNA-seq and lineage-tracing techniques. In postnatal skeletogenesis and impaired osteogenesis, SSCs residing in the bone marrow and growth plate that labeled with Lepr 75 , Gremlin 1 76 , or Gli1 77 , have been identified. Subsequently, SSCs subpopulations from long bone and bone marrow tissues have been identified by a combination of cell surface markers of CD45 − Ter119 − AlphaV + Thy − 6C3 − CD105 − CD200 + in mice 78 and PDPN + CD146 − CD73 + CD164 + in human 79 , respectively. In addition, periosteum-derived Ctsk + SSCs that are found in both the long bones and calvarium undergo intramembranous bone formation at baseline and contribute to the endochondral ossification for cortical bone repair 80 . In our previous bone repair study, Msx1 + SSCs were identified as sources of osteochondral progenitors and osteoprogenitors for full-thickness reconstruction during skull bone regeneration 18 . The above definitions for SSCs greatly differ from our identification on Krt8 + skeletal stem/progenitor cells in the rat calvarium ( Fig . 5f and Fig. 6f ), in which the common expression of CD200 indicates a more closer similarity to the cells from mice 78 . Interestingly, the Krt8 + SSCs identified in this study were also Msx1-positive ( Fig . 8b, d and Supplementary Fig. 10c ), indicating that Krt8 + Msx1 + SSCs better represent cells with higher potential for chondrogenic and osteogenic differentiation both during both bone injury healing and bone development. Strikingly, scRNA-seq data analysis of the alveolar regeneration process showed that airway and alveolar stem cells converge in a Krt8 + transitional stem cell state, which has been demonstrated to not undergo proliferative expansion and likely undergo normal homeostatic turnover 81 . Since our data do not include any epithelial cells, nor do the tissues acquired in single cell suspension, we discovered an epithelial feature of skeletal stem cells that forms a unique cell niche that peaks during the osteogenic phase of bone tissue repair ( Fig. 7e, g and Supplementary Fig. 10b, c ). Although the epithelial–mesenchymal transitions (EMTs) are considered as the characteristics of cell plasticity in embryonic development and cancer metastases, the importance of EMTs or the dual epithelial–mesenchymal properties of osteoprogenitors in tissue repair has become increasingly recognized 82-84 . These highly innovative studies suggest that the activation of transient or partial EMTs process, on the other hand, could confer phenotypic and functional plasticity on cells, which could enhance extraordinary tissue repair following damage 82 . This provides new insight into the mechanisms underlying tissue healing and has the potential to pave the way for novel therapeutics targeting cell plasticity to promote regeneration of currently irreversible tissue damage. Methods Human bone marrow-derived stem cells and human child phalangeal bone tissues. Human BM aspirates were obtained in The Second Affiliated Hospital of Zhejiang University School of Medicine (Zhejiang University, China) with written informed consent from orthopedic individuals with femoral fracture (v 1.3, 2016.8.3). All samples were obtained and used according to standard guidelines approved by the Ethics Committee of the Second Affiliated Hospital of Zhejiang University School of Medicine (Ethics number: 2016-033). Human child phalangeal bone tissues were obtained in Children's Hospital of Zhejiang University School of Medicine (Zhejiang University, China) with written informed consent from individuals after polydactyly resection surgery. Samples were obtained and used according to standard guidelines approved by the ethics committee of Children's Hospital of Zhejiang University (Ethics number: 2020-IRB-007). The bone marrow samples were processed for the following experiments, after the filters used to trap bone spicules and cell aggregates, and were carefully and aseptically washed with cold PBS several times. BMSCs were isolated using a Percoll gradient (Gibco) and cultured in Low-Glucose Dulbecco’s Modified Eagle’s Medium (DMEM; Gibco) containing 10% fetal bovine serum (FBS; Gibco), 1% penicillin/streptomycin (P/S) and incubated at 37 °C, 5% CO2. The culture medium was replaced every other day until about 90% confluence was achieved. Passage 3-5 of the BMSCs were utilized for the in vitro experiments either in 2D or 3D culture model in this study. Rat bone marrow-derived stem cells isolation and culture. The bone marrow from the Sprague Dawley rat (2 weeks old) femora and tibia was flushed in freshly prepared culture medium with L-DMEM medium (DMEM; Gibco) containing 10% fetal bovine serum (FBS; Gibco), and 1% penicillin/streptomycin (P/S; Thermo Fisher Scientific), incubated at 37°C in a humidified atmosphere containing 5% CO 2 . After centrifuged, the bone marrow was resuspended in growth medium and then seeded in 2D or 3D culture system, and samples were analyzed in certain time points. The culture medium was refreshed every 2 days until about 90% confluence was achieved. Passage 3-6 of the BMSCs was utilized for the in vitro experiments in this study. Cell viability assay. 1 X 10 5 cells/well were suspended in DMEM media supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. And the cells and media were incubated for 1, 3, 5 days in 2D culture or 1, 4, 7 days in 3D microsphere-based culture, and the CCK-8 kit (Dojindo, Japan) was chosen to determine cell viability. Osteogenic differentiation assays. For osteogenic differentiation, the cells were plated at confluence in osteogenic medium containing α-DMEM with 10% FBS supplemented with 0.1μM dexamethasone, 0.2 mM L-ascorbic acid, and 10 mM glycerol 2-phosphate disodium salt hydrate (Sigma, St. Louis, MO). For comparisons among different neuropeptides, distinct neuropeptides VIP (10 -7 M), NYP (10 -8 M), and CGRP (10 -8 M) were supplemented in the osteogenic medium. For comparisons among different treatments, CGRP (10 -8 M), BMP-2 (0.5μg/mL), CGRP (10 -8 M) combined with BMP-2 (0.5μg/mL) (CGRP + BMP-2), were supplemented in the osteogenic medium. The medium was changed every 2 days during 1 or 2 weeks. RT-qPCR of osteogenic genes. Osteogenic gene expressions of BMSCs were measured on day 7 and 14 by a Real-time quantitative reverse transcription-polymerase chain reaction (RT-qPCR) system, and four groups were set as above-mentioned. Angiogenic gene VEGF expression of BMSCs was measured by RT-qPCR at day 3 with the treatment of SP (10 -7 M). RNA was extracted from cells and reverse transcribed into complementary DNA (cDNA) using Trizol reagent and PrimeScript RT reagent kit (Takara, Tokyo, Japan) according to manufacturer's instructions. Then diluted cDNA was mixed with SYBR Premix Ex TaqTM (Takara, Tokyo, Japan), forward and reverse primers and RNase free water to perform RT-qPCR. Osteogenic differentiation markers including ALPL and RUNX2 were evaluated, with GAPDH used as housekeeping gene. Relative expression level for each gene (fold change) to that of blank control was calculated. Primer sequences used in this study were listed in Supplementary Table S1 . Alkaline Phosphatase (ALP) activity. BMSCs were seeded in 24-well plate at a density of 2×10 4 cells/well, and 24 h post-seeding the medium was changed to experimental medium. Cells were briefly washed with PBS and fixed for 10 min with 2% paraformaldehyde in PBS (Sigma). Cells were washed twice with PBS and incubated for 20 min at room temperature with 50 mg/mL Naphthol AS-MX phosphate, 0.5% N, N-Dimethylformamide, and 0.6 mg/mL Fast Red Violet LB in 0.1 M Tris-HCl, pH 8.9. Alizarin Red Staining. Cells were washed with cold PBS and fixed with 70% ethanol for 15 min on ice. Cells were then washed with distilled water and stained with 2% alizarin red solution for 5 min. Cells were subsequently washed thoroughly with distilled water and air dried before microscopic visualization. Synthesis of GelMA. GelMA (Methylacrylated Gelatin) was synthesized according to a previous study 18 . Briefly, type A gelatin (Sigma-Aldrich) was dissolved in PBS at 50°C and stirred to make a 10% w/v homogeneous solution. And a 0.1 mL methacrylic anhydride (MA) (Sigma-Aldrich) per gram of gelatin was added to homogeneous gelation solution at a rate of 0.5 mL per minute with a continuous stirring. The mixed solution was allowed to react at 50°C for 3 h with stirring. Then, the GelMA solution was poured into 8–14 kDa cutoff dialysis tubing (VWR Scientific USA) and dialyzed against deionized water for 6 days at 50°C to remove untreated MA and other byproducts. The deionized water was replaced every 1-2 days. The resulted GelMA solution was frozen overnight or longer at -80°C and lyophilized and stored at -20°C for further use. Synthesis of NB and HA-NB. Methyl 4-(4-(hydroxymethyl)-2-methoxy-5-nitrophenoxy) butanoate (mNB) was synthesized based on a previous research 85 . mNB (0.5g, 1.8mmol) and ethylenediamine (1.1mL, 2 mmol, Sigma Aldrich) were dissolved in methanol. Then, the mixture was refluxed overnight until the starting individual components were not detectable by thin layer chromatography (TLC). The solvent was evaporated under vacuum after the reaction was complete. The crude precipitate was then dissolved in methanol and re-precipitated three times using ethyl acetate. The filter cake was dried at 30°C for 12h under vacuum until NB appeared as a light-yellow powder (0.4 g, 1.2 mmol, 66.7%). HA-NB was synthesized according to a published report 85 . Briefly, HA (408 mg, 1 mmol of disaccharide unit, Dongyuan Biotech, Zhenjiang) was dissolved in 50 mL deionized water at room temperature and NB (224 mg, 0.69 mmol) was added followed by HOBt (153 mg, 1 mmol, Sigma-Aldrich). The pH of the mixture was adjusted to pH 4.5, and the 1-(3-Dimethylaminopropyl)-3-ethylcarbodimide hydrochloride (EDC) (200 mg, 1.04, Sigma Aldrich) was added to the mixed solution and then stirred at room temperature for 48h. Then, the solution was loaded into dialysis tubing (Molecular Weight (MW) cutoff 3500, Spectrum®) and dialyzed against diluted HCl (pH 3.5) containing 0.1M NaCl for two days, then dialyzed against deionized water for 2 days. The solution was lyophilized and HA-NB was obtained in powder form. Synthesis and preparation of the photo-initiator. The photo-initiator lithium phenyl-2,4,6- trimethylbenzoylphosphinate (LAP) was synthesized based on our previous research 86,87 . In brief, Dimethyl phenylphosphonite (Ourchem) was reacted with 2,4,6-trimethylbenzoyl chloride (Sigma-Aldrich) via a Michaelis–Arbuzov reaction. At room temperature and under argon gas, 3.2 g (0.018 mol) of 2,4,6-trimethylbenzoyl chloride was added dropwise to an equimolar amount of continuously stirred dimethyl phenylphosphonite (3.0 g). The reaction mixture was stirred for 18h whereupon a four-fold excess of lithium bromide (Aladdin, 6.1g) in 100mL of 2-butanone (Sinopharm Chemical Reagent) was added to the reaction mixture from the previous step, which was then heated to 50°C, a solid precipitate had formed after 10 minutes. Then, the mixture was cooled to ambient temperature and allowed to rest for 4h, and then filtered. The filtrate was washed and filtered 3 times with 2-butanone to remove unreacted lithium bromide, and excess solvent was removed by vacuum. Fabrication of GelMA/HA-NB microcarriers. For precursors of GelMA/HA-NB hydrogels, the freeze-dried GelMA foams and HA-NB foams were dissolved in PBS solution at 40 °C to a final concentration of 5% GelMA and 1.25% HA-NB. For precursors of GelMA/LAP hydrogels, the freeze-dried GelMA foams were dissolved in PBS solution at 40 °C and then added to the photo-initiator LAP to a final concentration of 5% GelMA and 0.1% LAP. In this study, GelMA/HA-NB microcarriers were fabricated by DLP-based 3D bioprinting method. In biocompatibility comparisons, the GelMA/HA-NB microcarriers (2×10 6 cells/mL of the hydrogel) with diameter of 100μm, 200μm, 400μm that encapsulating BMSCs were fabricated. In cell-loaded experiments, the GelMA/HA-NB microcarriers with a diameter of 400μm that encapsulating BMSCs were fabricated (2×10 6 cells/mL of the hydrogel) with no proteins, single BMP-2, or BMP-2 combined with CGRP. Live/dead staining assay. 1 × 10 5 cells were added to per GelMA/HA-NB/LAP to evaluate the cytotoxicity of GelMA/HA microspheres. And the cells were cultured in the microspheres in DMEM media supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin at 37 ◦C and incubated at 5% CO2 for 1, 4, 7, and 14 days. Images were captured by a laser scanning confocal microscope (OLYMPUS IX83-FV3000, Japan) for live/dead assay. After 1, 4 and 7 days of culture, 100μL microcarriers with diameter of 200μm and 400μm were collected and digested with type I collagenase and cells were counted. Growth factors used for Live/dead staining assay including BMP-2 (1μg/mL) and CGRP (10 -7 M). ELISA tests of growth factor release. The microcarriers containing BMP-2 (5μg/ml) was mixed with GelMA hydrogel containing VEGF (5μg/mL) in equal proportion, and the mixture of 200μL was absorbed and cross-linked to cure under UV light for 1 min. The mixture was then immersed in PBS and placed in an incubator at 37°C. PBS was absorbed at regular intervals after soaking, and the release of VEGF and BMP-2 from hydrogel compound structure was detected by Human BMP-2 ELISA kit (NeoBioscience) and Human VEGF ELISA kit (NeoBioscience), respectively. Animal and surgical procedures. SD rats (~ 250 g, 8~10 weeks old, Male) were used in this study. All surgical procedures were performed under 4% isoflurane anesthesia. Surgical sites were sterilized using iodine solution after hair removal using a clipper. In subcutaneous transplantation model, an incision with the length of 1.5 cm was made in the mediodorsal skin and a lateral subcutaneous pocket of SD rats was prepared. For biocompatibility assessment, GelMA and GelMA/HA-NB hydrogel (cylindrical flake: 8 mm in diameter and 1mm in thickness) were implanted under sterile conditions. At 1, 2, and 4 weeks, the rats were sacrificed and the samples were processed for histological analysis. For ectopic ossification experiments, the integrated delivery systems were divided into three groups: (1) GelMA hydrogel + GelMA/HA-NB Ms; (2) GelMA hydrogel (SP) + GelMA/HA-NB Ms; (3) GelMA hydrogel (SP) + GelMA/HA-NB Ms (BMSCs), and mixed according to the volume ratio of 1:2. The mixture of 200μL was absorbed and cross-linked and cured by UV light for 1 min, and then implanted subcutaneously into rats. At 3 weeks, the rats were sacrificed and the samples were processed for histological analysis. In calvarial bone defect model, a cranial defect with a diameter of 5 mm was created in the center of the calvarium using dental trephine. The rats were divided into four groups: (1) Defect control (Defect), (2) Microsphere in hydrogel (Vehicle), (3) Mixed MSCs and Neurotrophins (Hybrid), (3) Ossification center organoid (OCO). Growth factors used for bone injury repair in vivo including BMP-2 (5μg/mL), CGRP (10 -7 M), SP (10 -7 M). At 2 weeks after surgery, the rats were sacrificed and the regenerative tissues in the defect area were harvested and digested for single-cell collection. At 4 weeks and 8 weeks after surgery, the rats were sacrificed and the calvarium were harvested for further histological assessments, respectively. The rats used in this study were fed in separated cages in a temperature, humidity-controlled (~25°C, 50–80%) and 12 h light/dark cycle room. All animals were treated according to standard guidelines approved by the Zhejiang University Ethics Committee (Ethical NO. ZJU20210114). Micro CT analysis and paraffin embedding. After the defect surgery for 4 and 8 weeks, rats (n = 5, per group, per time point) were euthanatized and the calvarial specimens were harvested and fixed overnight with 4% paraformaldehyde at 4 °C. The fixed samples were scanned using micro-CT (U-CT-XUHR, MILabs) at 4μm resolution. The three-dimensional (3D) structures of calvarium were reconstructed through MILabs-Rec interface, and analyzed by IMALYTICS Preclinical 2.1. software of the micro-CT. A cylinder space representing the region of interest (ROI) was designated to evaluate both bone and tissue volume for calculation of bone volume/tissue volume (BV/TV) and Bone density (BMD). Samples were then decalcified with 0.5 M EDTA for 8 weeks and then subjected to paraffin embedding. Histological analysis. The harvested tissue specimens were fixed and then decalcified in 10% ethylenediamine tetra-acetic acid (EDTA, pH = 7.4) solution for two months at room temperature, and then dehydrated through graded alcohol series and embedded in paraffin. Sections of the central segment were cut into 10μm thick slices using a rotary microtome (Leica, Hamburg, Germany). Hematoxylin and Eosin (H&E) staining, Safranin-O staining, and Masson’s Trichrome staining were performed on paraffin sections according to standard protocols, and observed using bright-field microscopy. Immunofluorescent staining. Cells cultured either on dishes or in microspheres were fixed with 4% paraformaldehyde (PFA) for 20 min and then permeabilized with 0.03% Triton X-100 for 10 min at room temperature. After washing with PBS for 3 times, the samples were incubated with blocking solution (1% bovine serum albumin) for 30 min at room temperature to prevent nonspecific binding. The primary antibodies were diluted 200 or 500-fold with blocking solution and added to the cell cultures at 4 °C overnight. Tissues were fixed with 4% PFA and the paraffin-embedded samples were cut into 10-μm-thick sections. The tissue samples were deparaffinized and the antigens were activated by heating the slides in 10 mM citrate buffer (pH 6.0) at 65 °C overnight. After treating the sections with 0.3% H 2 O 2 in MeOH, the samples were incubated with blocking solution (5% bovine serum albumin and 0.03% Triton X-100 in PBS). The samples were treated with primary antibodies at 4 °C overnight, including Ki67 (1:250 dilution, ab16667, Abcam, USA), γh2AX (1:250 dilution, ab22551, Abcam, USA), RUNX2 (1:200 dilution, ab76956, Abcam, UK), Collagen I (1:100 dilution, ab260043, Abcam, UK), OCN (1:50 dilution, MAB1419, R&D Systems, USA), NGFR (1:50 dilution, NBP2-67296, NOVUS, USA), CGRP (1:100 dilution, ab81887, Abcam, UK), beta III Tubulin (1:500 dilution, ab18207, Abcam, UK), Collagen X (1:100 dilution, 14-9771-82, Invitrogen, USA), CD31 (1:100 dilution, ab222783, Abcam, UK), CD200 (1:100 dilution, AF2724, R&D Systems, USA), Thy1 (1:200 dilution, ab181469, Abcam, UK), PDGFR alpha (1:250 dilution, ab203491, Abcam, UK), KRT8 (1:100 dilution, ab53280, Abcam, UK), HAS1 (1:250 dilution, PA5-95599, Invitrogen, USA), MSX1 (1:100 dilution, ab93287, Abcam, UK). After incubation with primary antibody, samples were then incubated with Alexa Fluor® secondary antibodies (G-Rabbit Alexa Fluor® 488, A11008; Goat anti mouse Alexa Fluor® 488, A11001; G-Rabbit Alexa Fluor® 546, A21430-f; Donkey-Mouse Alexa Fluor® 405, ab175658) (diluted 1:500) for 1 h at room temperature. After incubation, the nuclei were stained with 0.1μg/mL DAPI (Invitrogen, USA). After staining, the samples were observed using a confocal microscope (OLYMPUS IX83-FV1000 and FV3000-OSR, Japan). Immunohistochemistry staining. For immune-histochemical staining, sections were prepared, and followed by antigen retrieval by heating the slides in 10 mM citrate buffer (pH 6.0) at 65 °C overnight, inactivation of endogenous peroxidase by hydrogen peroxide with 0.3% H 2 O 2 in MeOH, 5% BSA blocking solution, and was incubated with primary antibodies against OCN (1:50 dilution, MAB1419, R&D Systems, USA), at 4 °C overnight. Then sections were incubated with anti-Mouse secondary antibody conjugated with 1:1000 HRP (Beyotime Institute of Biotechnology). The stained specimens were photographed digitally and viewed under the Digital Slide Scanners. Single-cell RNA Sequencing and data analysis. At 2 weeks after surgery, the rats (male, n=6 biological rats, per group) were sacrificed. The regenerated tissues in calvarial defect area were minced with razor blades and washed several times by 4 °C PBS, and then digested by enzyme mixture (type I collagenase 0.1% and type II collagenase 0.1%, incubated at 37°C for 40 min) before filtered through a 70μm nylon mesh to obtain single cell suspension. Subsequently, Single-cell suspensions (2×10 5 cells/mL) with PBS (HyClone) were loaded onto microwell chip using the Singleron Matrix® Single Cell Processing System. Next, the scRNA-seq libraries were constructed according to the protocol of the GEXSCOPE single-cell RNA library kits (Singleron). Individual libraries were diluted to 4 nM and pooled for sequencing. Finally, the pools were sequenced on the Illumina NovaSeq 6000 system with 150 bp paired-end reads. For single-cell clustering and annotation,after dimensionality reduction, we performed clustering using FindClusters function offered by Seurat. Genes expressed specifically in each clusters were calculated by Seurat FindAllMarkers function. To identify the biological cell type of each cluster, we performed SingleR 88 analysis, combined with conventional markers of some known cell types. For differentiation trajectory analysis, to map differentiation in bone regeneration , we performed pseudotime analysis with R package Monocle 89 (version 2.20.0). Specifically, we computed trajectory of osteo-lineage cells and compared different states of cells using BEAM function provided by Monocle. RNA velocity was performed with scvelo python package (version 0.1.25).For gene functional annotation analysis,GO enrichment analysis was performed for markers of single-cell clusters using clusterProfiler 90 package. The enriched GO terms were filtered by setting pvalueCutoff to 0.01.For single-cell regulatory network analysis,the analysis of single-cell gene regulatory network was performed using the SCENIC 91 package followed by the standard pipeline. Dot plot shows the cell-type specific regulons with top Regulon Specificity Score (RSS) 92 and their average expression (Z) in the cell subtype.Homologous gene conversion among human, rat and mouse genes was performed by R package babelgene (version 22.9). Machine learning and model optimization in this study was performed using sklearn python package (version 1.0.2). Statistical analysis. Values are expressed as mean ± SD or mean ± SEM unless otherwise indicated in the figure legends. The significance between two groups was analyzed using two-tailed Student’s t-tests. For multiple comparisons, one-way analysis of variance (ANOVA) with Tukey’s post hoc test was used. Statistical analysis was performed using the Graphpad software. P < 0.05 was considered to be significant; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Declarations Data availability The single cell RNA-sequencing data generated in this study have been deposited in the Genome Sequence Archive (GSA) database with accession number CRA018756, which is now publicly available can be accessed from the following link [https://bigd.big.ac.cn/gsa/browse/CRA018756]. All other relevant data supporting the key findings of this study are available within the article and its Supplementary Information files or from the corresponding author upon reasonable request. Acknowledgements This work was supported by the National Key Research and Development Program of China (2023YFB3183000), and NSFC grants (NO. T2121004, 92268203, 82301016). The authors thank Weiliang Shen (from The Second Affiliated Hospital of Zhejiang University) for providing the human bone marrow samples from trauma subjects in traffic accidents or discarded tissue during the operation of fractures. We thank Wei Yin and Junli Xuan (Core Facilities, Zhejiang University School of Medicine) for their assistance with Confocal laser scanning microscope. Author contributions Xianzhu Zhang and Hongwei Ouyang designed the project, performed experiments and wrote the whole manuscript; Xianzhu Zhang, Wei Jiang, Xinyu Wu, Xiaohui Zou, and Hongwei Ouyang helped revised the manuscript; Xianzhu Zhang, Xinyu Wu, and Yi Zhang performed the in vitro cell culture and osteogenic induction assays; Xinyu Wu and Chang Xie, Renjie Liang, and Liying Li completed the hydrogel and microsphere preparation; Xianzhu Zhang, Xinyu Wu, Tao Zhang, Wei Sun, Jingchun Ye, and Youzhi Cai performed the animal experiments and the histological analysis; Chang Xie, Yi Zhang, and Yuqing Gu, and Liying Li helped with the immunofluorescent and immunohistochemical staining of regenerative tissue sections; Xianzhu Zhang, Wei Jiang, and Zihao Hu performed single-cell sequencing and data analysis, and machine learning-based data analysis; Xiaozhao Wang, Wei Wei, and Yi Hong helped with the material preparation and DLP printing; Shufang Zhang, Xiaohui Zou, Yihe Hu and Hongwei Ouyang helped with the results analysis and discussion. Competing interests The authors declare that they have no competing interests. References Zura, R. et al. Epidemiology of Fracture Nonunion in 18 Human Bones. JAMA Surg 151 , e162775, doi:10.1001/jamasurg.2016.2775 (2016). Wang, L. et al. Development of a centrally vascularized tissue engineering bone graft with the unique core-shell composite structure for large femoral bone defect treatment. Biomaterials 175 , 44-60, doi:10.1016/j.biomaterials.2018.05.017 (2018). Chen, S. X. et al. Biomaterials with structural hierarchy and controlled 3D nanotopography guide endogenous bone regeneration. Science Advances 7 , doi:ARTN eabg3089 /sciadv.abg3089 (2021). Ho-Shui-Ling, A. et al. Bone regeneration strategies: Engineered scaffolds, bioactive molecules and stem cells current stage and future perspectives. Biomaterials 180 , 143-162, doi:10.1016/j.biomaterials.2018.07.017 (2018). Bianco, P. et al. The meaning, the sense and the significance: translating the science of mesenchymal stem cells into medicine. Nature Medicine 19 , 35-42, doi:10.1038/nm.3028 (2013). McDermott, A. M. et al. Recapitulating bone development through engineered mesenchymal condensations and mechanical cues for tissue regeneration. Science Translational Medicine 11 , doi:ARTN eaav7756/scitranslmed.aav7756 (2019). Dai, K. et al. Construction of developmentally inspired periosteum-like tissue for bone regeneration. Bone Research 10 , doi:ARTN 1 /s41413-021-00166-w (2022). Herberg, S. et al. Combinatorial morphogenetic and mechanical cues to mimic bone development for defect repair. Science Advances 5 , doi:ARTN eaax2476 /sciadv.aax2476 (2019). Nilsson Hall, G. et al. Developmentally Engineered Callus Organoid Bioassemblies Exhibit Predictive In Vivo Long Bone Healing. Adv Sci (Weinh) 7 , 1902295, doi:10.1002/advs.201902295 (2020). Hasani-Sadrabadi, M. M. et al. An engineered cell-laden adhesive hydrogel promotes craniofacial bone tissue regeneration in rats. Sci Transl Med 12 , doi:10.1126/scitranslmed.aay6853 (2020). Petersen, A. et al. A biomaterial with a channel-like pore architecture induces endochondral healing of bone defects. Nature Communications 9 , doi:ARTN 4430 /s41467-018-06504-7 (2018). Moroni, L. et al. Biofabrication strategies for 3D in vitro models and regenerative medicine. Nature Reviews Materials 3 , 21-37, doi:10.1038/s41578-018-0006-y (2018). Xie, C. et al. High-efficient engineering of osteo-callus organoids for rapid bone regeneration within one month. Biomaterials 288 , doi:ARTN 121741 /j.biomaterials.2022.121741 (2022). He, Q. L. et al. "All-in-One" Gel System for Whole Procedure of Stem-Cell Amplification and Tissue Engineering. Small 16 , doi:ARTN 1906539 /smll.201906539 (2020). Subbiah, R. et al. Osteogenic/Angiogenic Dual Growth Factor Delivery Microcapsules for Regeneration of Vascularized Bone Tissue. Advanced Healthcare Materials 4 , 1982-1992, doi:10.1002/adhm.201500341 (2015). Kim, D. S. et al. Advanced PLGA hybrid scaffold with a bioactive PDRN/BMP2 nanocomplex for angiogenesis and bone regeneration using human fetal MSCs. Sci Adv 7 , eabj1083, doi:10.1126/sciadv.abj1083 (2021). Freeman, F. E. et al. 3D bioprinting spatiotemporally defined patterns of growth factors to tightly control tissue regeneration. Sci Adv 6 , eabb5093, doi:10.1126/sciadv.abb5093 (2020). Zhang, X. et al. Msx1(+) stem cells recruited by bioactive tissue engineering graft for bone regeneration. Nat Commun 13 , 5211, doi:10.1038/s41467-022-32868-y (2022). Tomlinson, R. E. et al. NGF-TrkA Signaling by Sensory Nerves Coordinates the Vascularization and Ossification of Developing Endochondral Bone. Cell Reports 16 , 2723-2735, doi:10.1016/j.celrep.2016.08.002 (2016). Lee, S. Y. et al. NGF-TrkA signaling dictates neural ingrowth and aberrant osteochondral differentiation after soft tissue trauma. Nature Communications 12 , doi:ARTN 4939 /s41467-021-25143-z (2021). Li, Z. et al. Fracture repair requires TrkA signaling by skeletal sensory nerves. Journal of Clinical Investigation 129 , 5137-5150, doi:10.1172/Jci128428 (2019). Li, Z. et al. Fracture repair requires TrkA signaling by skeletal sensory nerves. J Clin Invest 129 , 5137-5150, doi:10.1172/JCI128428 (2019). Meyers, C. A. et al. A Neurotrophic Mechanism Directs Sensory Nerve Transit in Cranial Bone. Cell Rep 31 , 107696, doi:10.1016/j.celrep.2020.107696 (2020). Xu, J. et al. NGF-p75 signaling coordinates skeletal cell migration during bone repair. Sci Adv 8 , eabl5716, doi:10.1126/sciadv.abl5716 (2022). Li, W. T. et al. Bioprinted Constructs that Mimic the Ossification Center Microenvironment for Targeted Innervation in Bone Regeneration. Adv Funct Mater 32 , doi:ARTN 2109871 /adfm.202109871 (2022). Zhang, Y. F. et al. Implant-derived magnesium induces local neuronal production of CGRP to improve bone-fracture healing in rats. Nature Medicine 22 , 1160-1169, doi:10.1038/nm.4162 (2016). Lu, Y. Z. et al. CGRP sensory neurons promote tissue healing via neutrophils and macrophages. Nature , doi:10.1038/s41586-024-07237-y (2024). Hong, H. S. et al. A new role of substance P as an injury-inducible messenger for mobilization of CD29 stromal-like cells. Nature Medicine 15 , 425-435, doi:10.1038/nm.1909 (2009). Zhang, Y. et al. Neuronal Induction of Bone-Fat Imbalance through Osteocyte Neuropeptide Y. Adv Sci 8 , doi:ARTN 2100808/advs.202100808 (2021). Shi, L. et al. Vasoactive Intestinal Peptide Stimulates Bone Marrow-Mesenchymal Stem Cells Osteogenesis Differentiation by Activating Wnt/β-Catenin Signaling Pathway and Promotes Rat Skull Defect Repair. Stem Cells and Development 29 , 655-666, doi:10.1089/scd.2019.0148 (2020). Yu, H. P. et al. Application of the neuropeptide NPVF to enhance angiogenesis and osteogenesis in bone regeneration. Commun Biol 6 , doi:ARTN 197 /s42003-023-04567-x (2023). Wang, L. P. et al. Mice lacking substance P have normal bone modeling but diminished bone formation, increased resorption, and accelerated osteopenia with aging. Bone 144 , doi:ARTN 115806 /j.bone.2020.115806 (2021). Liu, L. J., Dana, R. & Yin, J. Sensory neurons directly promote angiogenesis in response to inflammation via substance P signaling. Faseb J 34 , 6229-6243, doi:10.1096/fj.201903236R (2020). Tang, W. et al. Enhancement and orchestration of osteogenesis and angiogenesis by a dual-modular design of growth factors delivery scaffolds and 26SCS decoration (vol 232, 119645, 2020). Biomaterials 268 , doi:ARTN 120616 /j.biomaterials.2020.120616 (2021). Lee, S. S. et al. Sequential growth factor releasing double cryogel system for enhanced bone regeneration. Biomaterials 257 , doi:ARTN 120223 /j.biomaterials.2020.120223 (2020). Freeman, F. E. et al. 3D bioprinting spatiotemporally defined patterns of growth factors to tightly control tissue regeneration. Science Advances 6 , doi:ARTN eabb5093 /sciadv.abb5093 (2020). Wang, Q., Liu, K. L., Yang, L., Wang, H. H. & Yang, J. BoneClear: whole-tissue immunolabeling of the intact mouse bones for 3D imaging of neural anatomy and pathology. Cell Research 29 , 870-872, doi:10.1038/s41422-019-0217-9 (2019). Kelly, N. H., Huynh, N. P. T. & Guilak, F. Single cell RNA-sequencing reveals cellular heterogeneity and trajectories of lineage specification during murine embryonic limb development. Matrix Biol 89 , 1-10, doi:10.1016/j.matbio.2019.12.004 (2020). Li, J. et al. Systematic Reconstruction of Molecular Cascades Regulating GP Development Using Single-Cell RNA-Seq. Cell Rep 15 , 1467-1480, doi:10.1016/j.celrep.2016.04.043 (2016). Sivaraj, K. K. et al. Regional specialization and fate specification of bone stromal cells in skeletal development. Cell Rep 36 , 109352, doi:10.1016/j.celrep.2021.109352 (2021). He, J. et al. Dissecting human embryonic skeletal stem cell ontogeny by single-cell transcriptomic and functional analyses. Cell Res 31 , 742-757, doi:10.1038/s41422-021-00467-z (2021). Luo, C. et al. Biomimetic open porous structured core-shell microtissue with enhanced mechanical properties for bottom-up bone tissue engineering. Theranostics 9 , 4663-4677, doi:10.7150/thno.34464 (2019). Cui, Z. K. et al. Microporous methacrylated glycol chitosan-montmorillonite nanocomposite hydrogel for bone tissue engineering. Nat Commun 10 , 3523, doi:10.1038/s41467-019-11511-3 (2019). Harvestine, J. N. et al. Osteogenic preconditioning in perfusion bioreactors improves vascularization and bone formation by human bone marrow aspirates. Sci Adv 6 , eaay2387, doi:10.1126/sciadv.aay2387 (2020). Dai, K. et al. Generation of rhBMP-2-induced juvenile ossicles in aged mice. Biomaterials 258 , 120284, doi:10.1016/j.biomaterials.2020.120284 (2020). Herberg, S. et al. Combinatorial morphogenetic and mechanical cues to mimic bone development for defect repair. Sci Adv 5 , eaax2476, doi:10.1126/sciadv.aax2476 (2019). Zhou, P. et al. Loading BMP-2 on nanostructured hydroxyapatite microspheres for rapid bone regeneration. Int J Nanomedicine 13 , 4083-4092, doi:10.2147/IJN.S158280 (2018). Boerckel, J. D. et al. Effects of protein dose and delivery system on BMP-mediated bone regeneration. Biomaterials 32 , 5241-5251, doi:10.1016/j.biomaterials.2011.03.063 (2011). Geuze, R. E. et al. A Differential Effect of Bone Morphogenetic Protein-2 and Vascular Endothelial Growth Factor Release Timing on Osteogenesis at Ectopic and Orthotopic Sites in a Large-Animal Model. Tissue Eng Pt A 18 , 2052-2062, doi:10.1089/ten.tea.2011.0560 (2012). Lee, K., Silva, E. A. & Mooney, D. J. Growth factor delivery-based tissue engineering: general approaches and a review of recent developments. J R Soc Interface 8 , 153-170, doi:10.1098/rsif.2010.0223 (2011). Yan, H. J. et al. Synthetic design of growth factor sequestering extracellular matrix mimetic hydrogel for promoting in vivo bone formation. Biomaterials 161 , 190-202, doi:10.1016/j.biomaterials.2018.01.041 (2018). Muschter, D. et al. Sensory Neuropeptides and their Receptors Participate in Mechano-Regulation of Murine Macrophages. Int J Mol Sci 20 , doi:ARTN 503 /ijms20030503 (2019). Konttinen, Y. T., Imai, S. & Suda, A. Neuropeptides and the puzzle of bone remodeling - State of the art. Acta Orthop Scand 67 , 632-639, doi:Doi 10.3109/17453679608997772 (1996). Elefteriou, F. Neuronal signaling and the regulation of bone remodeling. Cell Mol Life Sci 62 , 2339-2349, doi:10.1007/s00018-005-5175-3 (2005). Zhang, Y. et al. Implant-derived magnesium induces local neuronal production of CGRP to improve bone-fracture healing in rats. Nat Med 22 , 1160-1169, doi:10.1038/nm.4162 (2016). Lu, Y. Z. et al. CGRP sensory neurons promote tissue healing via neutrophils and macrophages. Nature 628 , 604-611, doi:10.1038/s41586-024-07237-y (2024). Hong, H. S. et al. A new role of substance P as an injury-inducible messenger for mobilization of CD29(+) stromal-like cells. Nature Medicine 15 , 425-435, doi:10.1038/nm.1909 (2009). Kim, R. Y. et al. The effect of dose on rhBMP-2 signaling, delivered via collagen sponge, on osteoclast activation and in vivo bone resorption. Biomaterials 35 , 1869-1881, doi:10.1016/j.biomaterials.2013.11.029 (2014). Zara, J. N. et al. High Doses of Bone Morphogenetic Protein 2 Induce Structurally Abnormal Bone and Inflammation In Vivo. Tissue Eng Pt A 17 , 1389-1399, doi:10.1089/ten.tea.2010.0555 (2011). Bixel, M. G. et al. Angiogenesis is uncoupled from osteogenesis during calvarial bone regeneration. Nature Communications 15 , doi:ARTN 4575 /s41467-024-48579-5 (2024). Lu, G. et al. An instantly fixable and self-adaptive scaffold for skull regeneration by autologous stem cell recruitment and angiogenesis. Nat Commun 13 , 2499, doi:10.1038/s41467-022-30243-5 (2022). Kim, H. et al. Mesenchymal stem cell 3D encapsulation technologies for biomimetic microenvironment in tissue regeneration. Stem Cell Res Ther 10 , doi:ARTN 51 /s13287-018-1130-8 (2019). Annamalai, R. T. et al. Injectable osteogenic microtissues containing mesenchymal stromal cells conformally fill and repair critical-size defects. Biomaterials 208 , 32-44, doi:10.1016/j.biomaterials.2019.04.001 (2019). Yuan, Z. et al. Injectable GelMA Cryogel Microspheres for Modularized Cell Delivery and Potential Vascularized Bone Regeneration. Small 17 , e2006596, doi:10.1002/smll.202006596 (2021). Luo, C. et al. Biomimetic open porous structured core-shell microtissue with enhanced mechanical properties for bottom-up bone tissue engineering. Theranostics 9 , 4663-4677, doi:10.7150/thno.34464 (2019). Chen, S., Chen, X., Geng, Z. & Su, J. The horizon of bone organoid: A perspective on construction and application. Bioact Mater 18 , 15-25, doi:10.1016/j.bioactmat.2022.01.048 (2022). Akiva, A. et al. An Organoid for Woven Bone. Adv Funct Mater 31 , doi:ARTN 2010524 /adfm.202010524 (2021). Hall, G. N. et al. Developmentally Engineered Callus Organoid Bioassemblies Exhibit Predictive In Vivo Long Bone Healing. Adv Sci 7 , doi:ARTN 1902295 /advs.201902295 (2020). Park, Y. et al. Trabecular bone organoid model for studying the regulation of localized bone remodeling. Sci Adv 7 , doi:10.1126/sciadv.abd6495 (2021). Grigoryan, A. et al. Engineering human mini-bones for the standardized modeling of healthy hematopoiesis, leukemia, and solid tumor metastasis. Sci Transl Med 14 , eabm6391, doi:10.1126/scitranslmed.abm6391 (2022). Wang, J. et al. Engineering Large-Scale Self-Mineralizing Bone Organoids with Bone Matrix-Inspired Hydroxyapatite Hybrid Bioinks. Adv Mater 36 , e2309875, doi:10.1002/adma.202309875 (2024). Farah, E. N. et al. Spatially organized cellular communities form the developing human heart. Nature 627 , 854-864, doi:10.1038/s41586-024-07171-z (2024). Kang, J. et al. Systematic dissection of tumor-normal single-cell ecosystems across a thousand tumors of 30 cancer types. Nat Commun 15 , 4067, doi:10.1038/s41467-024-48310-4 (2024). Maity, A. K. & Teschendorff, A. E. Cell-attribute aware community detection improves differential abundance testing from single-cell RNA-Seq data. Nat Commun 14 , 3244, doi:10.1038/s41467-023-39017-z (2023). Shen, B. et al. A mechanosensitive peri-arteriolar niche for osteogenesis and lymphopoiesis. Nature 591 , 438-444, doi:10.1038/s41586-021-03298-5 (2021). Worthley, D. L. et al. Gremlin 1 identifies a skeletal stem cell with bone, cartilage, and reticular stromal potential. Cell 160 , 269-284, doi:10.1016/j.cell.2014.11.042 (2015). Shi, Y. et al. Gli1 identifies osteogenic progenitors for bone formation and fracture repair. Nat Commun 8 , 2043, doi:10.1038/s41467-017-02171-2 (2017). Chan, C. K. et al. Identification and specification of the mouse skeletal stem cell. Cell 160 , 285-298, doi:10.1016/j.cell.2014.12.002 (2015). Chan, C. K. F. et al. Identification of the Human Skeletal Stem Cell. Cell 175 , 43-56 e21, doi:10.1016/j.cell.2018.07.029 (2018). Debnath, S. et al. Discovery of a periosteal stem cell mediating intramembranous bone formation. Nature 562 , 133-139, doi:10.1038/s41586-018-0554-8 (2018). Strunz, M. et al. Alveolar regeneration through a Krt8+transitional stem cell state that persists in human lung fibrosis. Nature Communications 11 , doi:ARTN 3559 /s41467-020-17358-3 (2020). Youssef, K. K. & Nieto, M. A. Epithelial-mesenchymal transition in tissue repair and degeneration. Nat Rev Mol Cell Biol , doi:10.1038/s41580-024-00733-z (2024). Weng, Y. et al. A novel lineage of osteoprogenitor cells with dual epithelial and mesenchymal properties govern maxillofacial bone homeostasis and regeneration after MSFL. Cell Res 32 , 814-830, doi:10.1038/s41422-022-00687-x (2022). Storer, M. A. et al. Acquisition of a Unique Mesenchymal Precursor-like Blastema State Underlies Successful Adult Mammalian Digit Tip Regeneration. Dev Cell 52 , 509-524 e509, doi:10.1016/j.devcel.2019.12.004 (2020). Yang, Y. L. et al. Tissue-Integratable and Biocompatible Photogelation by the Imine Crosslinking Reaction. Adv Mater 28 , 2724-2730, doi:10.1002/adma.201505336 (2016). Hong, Y. et al. A strongly adhesive hemostatic hydrogel for the repair of arterial and heart bleeds. Nature Communications 10 , doi:ARTN 2060 /s41467-019-10004-7 (2019). Zhou, F. F. et al. Rapid printing of bio-inspired 3D tissue constructs for skin regeneration. Biomaterials 258 , doi:ARTN 120287/j.biomaterials.2020.120287 (2020). Aran, D. et al. Reference-based analysis of lung single-cell sequencing reveals a transitional profibrotic macrophage. Nat Immunol 20 , 163-+, doi:10.1038/s41590-018-0276-y (2019). Qiu, X. J. et al. Reversed graph embedding resolves complex single-cell trajectories. Nat Methods 14 , 979-+, doi:10.1038/Nmeth.4402 (2017). Yu, G. C., Wang, L. G., Han, Y. Y. & He, Q. Y. clusterProfiler: an R Package for Comparing Biological Themes Among Gene Clusters. Omics 16 , 284-287, doi:10.1089/omi.2011.0118 (2012). Aibar, S. et al. SCENIC: single-cell regulatory network inference and clustering. Nat Methods 14 , 1083-+, doi:10.1038/Nmeth.4463 (2017). Suo, S. B. et al. Revealing the Critical Regulators of Cell Identity in the Mouse Cell Atlas. Cell Reports 25 , 1436-+, doi:10.1016/j.celrep.2018.10.045 (2018). Additional Declarations There is NO Competing Interest. Supplementary Files reportingsummary.pdf Reporting Summary SupplementaryInformation.pdf Cite Share Download PDF Status: Published Journal Publication published 04 Jul, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5248946","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":378376197,"identity":"66939a26-3eda-4d8e-97af-26497c19ba36","order_by":0,"name":"Hongwei Ouyang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAsElEQVRIiWNgGAWjYBACAxDxgYENziZOC+MMkrUw8yDZSBiYSyQfk7b5w5fYwN68TYKh5g5hLZYz0tKkc3jYEht4jpVJMBx7RoTDbueYSedIALVI5JhJMDYcJlKLhQFQi/wbUrQwJIBs4SFWy/1nyZY9B9iM23jSii0SjhGj5czhgzd+/Dkm289+eOONDzVEaAECFgkGhmOQyEwgSgMwJj8wMNQQqXYUjIJRMApGJAAAcbI0rVQtZ7QAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0003-0627-2536","institution":"Zhejiang University","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Hongwei","middleName":"","lastName":"Ouyang","suffix":""},{"id":378376198,"identity":"0ac65b3e-7eeb-41f7-9352-72da07f35789","order_by":1,"name":"Xianzhu Zhang","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Xianzhu","middleName":"","lastName":"Zhang","suffix":""},{"id":378376199,"identity":"cf32623b-fb0c-49e3-88fb-c65bfdad9288","order_by":2,"name":"Wei Jiang","email":"","orcid":"https://orcid.org/0000-0002-6685-7053","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Jiang","suffix":""},{"id":378376200,"identity":"d792b89c-af54-48b2-870c-e88af820721f","order_by":3,"name":"Xinyu Wu","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Xinyu","middleName":"","lastName":"Wu","suffix":""},{"id":378376201,"identity":"b38da6f5-3ccb-4f89-9bdf-69d809659f61","order_by":4,"name":"Chang Xie","email":"","orcid":"https://orcid.org/0009-0000-3946-0600","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Chang","middleName":"","lastName":"Xie","suffix":""},{"id":378376202,"identity":"26376292-2f3e-4156-80e9-bb6a919ec3c6","order_by":5,"name":"Yi Zhang","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Yi","middleName":"","lastName":"Zhang","suffix":""},{"id":378376203,"identity":"b0cb3372-f12b-4fea-bc8e-300a80f3f100","order_by":6,"name":"Yuqing Gu","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Yuqing","middleName":"","lastName":"Gu","suffix":""},{"id":378376204,"identity":"1ba8d18d-4a97-4599-aafb-47f245a89bd9","order_by":7,"name":"Zihao Hu","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Zihao","middleName":"","lastName":"Hu","suffix":""},{"id":378376205,"identity":"a2fef1f4-083f-4d3a-bd1a-a98ed94d7883","order_by":8,"name":"Liying Li","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Liying","middleName":"","lastName":"Li","suffix":""},{"id":378376206,"identity":"bc2254e5-0bcf-4ffe-a4a6-274964e677af","order_by":9,"name":"Renjie Liang","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Renjie","middleName":"","lastName":"Liang","suffix":""},{"id":378376207,"identity":"d9d6493f-b34f-49af-9a3b-2d6ea3368f80","order_by":10,"name":"Tao Zhang","email":"","orcid":"","institution":"Department of Orthopedic Surgery of the Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Tao","middleName":"","lastName":"Zhang","suffix":""},{"id":378376208,"identity":"156e309a-2491-448b-9cd8-be784d709414","order_by":11,"name":"Wei Sun","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Sun","suffix":""},{"id":378376209,"identity":"e6c31084-8252-44ef-b871-a93c2b76be70","order_by":12,"name":"Jinchun Ye","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Jinchun","middleName":"","lastName":"Ye","suffix":""},{"id":378376210,"identity":"354cae8e-9e30-41f0-accf-715ac1064d3e","order_by":13,"name":"Wei Wei","email":"","orcid":"https://orcid.org/0000-0002-2234-4459","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Wei","suffix":""},{"id":378376211,"identity":"91741d7a-a2e1-4bc5-b1d5-f21571c74856","order_by":14,"name":"Xiaozhao Wang","email":"","orcid":"https://orcid.org/0009-0002-0257-4393","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Xiaozhao","middleName":"","lastName":"Wang","suffix":""},{"id":378376212,"identity":"1625eed1-f3bf-480a-aeb0-2dd8a0eac5b1","order_by":15,"name":"Yi Hong","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Yi","middleName":"","lastName":"Hong","suffix":""},{"id":378376213,"identity":"db48cb42-a1fa-43cb-9c25-8ee6dbf7aef2","order_by":16,"name":"Shufang Zhang","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Shufang","middleName":"","lastName":"Zhang","suffix":""},{"id":378376214,"identity":"2559d46f-9435-432c-bf9e-6526acc0f272","order_by":17,"name":"Youzhi Cai","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Youzhi","middleName":"","lastName":"Cai","suffix":""},{"id":378376215,"identity":"b30effe9-e3b2-46dc-b073-539bd6e87154","order_by":18,"name":"Xiaohui Zou","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Xiaohui","middleName":"","lastName":"Zou","suffix":""},{"id":378376216,"identity":"8606f5bc-f9da-457d-82df-b61f0e326a5b","order_by":19,"name":"Yihe Hu","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Yihe","middleName":"","lastName":"Hu","suffix":""}],"badges":[],"createdAt":"2024-10-12 02:20:29","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5248946/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5248946/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-025-61619-y","type":"published","date":"2025-07-04T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":70016293,"identity":"1438ff5b-0e0d-473d-bc49-fdbdc96f1004","added_by":"auto","created_at":"2024-11-27 14:00:39","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":282761,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCGRP synergistically promoted osteogenic differentiation of MSCs with BMP-2 at physiological dosage.\u003c/strong\u003e \u003cstrong\u003ea. \u003c/strong\u003eAlkaline Phosphatase (ALP) staining and \u003cstrong\u003eb\u003c/strong\u003e. mRNA expression level of ALPL at day 7 after treatment with distinct neuropeptides VIP, NYP, or CGRP, when compared with control osteogenic induction. \u003cstrong\u003ec, d.\u003c/strong\u003e mRNA expression level of ALPL and RUNX2 under osteogenic inductions supplemented with CGRP, BMP-2, or CGRP + BMP-2 for 7 days, respectively. \u003cstrong\u003ee, f.\u003c/strong\u003e mRNA expression level of ALPL and RUNX2 under osteogenic inductions supplemented with CGRP, BMP-2, or CGRP + BMP-2 for 14 days, respectively. \u003cstrong\u003eg\u003c/strong\u003e. Alkaline Phosphatase (ALP) staining and \u003cstrong\u003eh\u003c/strong\u003e. quantitative OD values under osteogenic inductions supplemented with CGRP, BMP-2, or CGRP + BMP-2 for 7 days, respectively. \u003cstrong\u003ei\u003c/strong\u003e. Alizarin Red S (ARS) staining and \u003cstrong\u003ej\u003c/strong\u003e. quantitative OD values under osteogenic inductions supplemented with CGRP, BMP-2, or CGRP + BMP-2 for 14 days, respectively. (Exact p value was calculated with one-way ANOVA Tukey’s multiple comparisons test)\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5248946/v1/72472ebfdfb709d8009f258b.png"},{"id":70016294,"identity":"66c202b5-decd-466e-9ce3-bb405da9b14e","added_by":"auto","created_at":"2024-11-27 14:00:39","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":310839,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003e3D printing-based fabrication of “ossification center”-like spheroid with enhanced cell survival and osteogenic potential. a\u003c/strong\u003e.\u003cstrong\u003e \u003c/strong\u003eSchematic diagram of MSCs-contained GelMA/HA-NB hydrogel microspheres loaded with CGRP (10\u003csup\u003e-8\u003c/sup\u003e M) and BMP-2 (0.5μg/mL) by DLP 3D printing. \u003cstrong\u003eb\u003c/strong\u003e. Bright field images of MSCs distribution and cell morphology of inside and outside of the microspheres with the diameter in 400μm after 7 days of \u003cem\u003ein vitro \u003c/em\u003eculture, bar = 200μm. \u003cstrong\u003ec\u003c/strong\u003e. Live/Dead staining (Green, live cells; Red, dead cells) of MSCs cultured in the vehicle microspheres, or either loaded with BMP-2 alone or combined with CGRP for 1, 4, 7 and 14 days, bar = 100μm. \u003cstrong\u003ed\u003c/strong\u003e. Quantitative data of cell number within per microsphere among the vehicle microsphere, BMP-2-loaded and BMP-2+CGRP-loaded microsphere. \u003cstrong\u003ee\u003c/strong\u003e. Workflow of growth medium pre-treatment for 1 day and the subsequent osteogenic induction for 7 days and 14 days. \u003cstrong\u003ef\u003c/strong\u003e. Alkaline Phosphatase (ALP) staining and \u003cstrong\u003eg\u003c/strong\u003e. Alizarin Red Staining under osteogenic differentiation of MSCs cultured in the vehicle microspheres, or either loaded with BMP-2 alone or combined with CGRP for 7 and 14 days, respectively, bar = 50μm. \u003cstrong\u003eh\u003c/strong\u003e and \u003cstrong\u003ei\u003c/strong\u003e. immunostaining of RUNX2 under osteogenic differentiation of MSCs cultured in the vehicle microspheres, or either loaded with BMP-2 alone or combined with CGRP for 7 and 14 days, respectively, bar = 50μm.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5248946/v1/939bbbcac2ea015a4d7974e9.png"},{"id":70016306,"identity":"892714d5-bf5c-4cdd-9da3-09e5665ebb88","added_by":"auto","created_at":"2024-11-27 14:00:39","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":470676,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOCO implantation achieve fast bone-bridging and full-thickness reconstruction with successive OC-like organoids fusion across the bone defects. a. \u003c/strong\u003eSchematic diagram of the ready-to-use fabrication and in-situ crosslinking of OC-like organoids. \u003cstrong\u003eb.\u003c/strong\u003e Schematic illustrating the implantation of the dual-modular hydrogel (Vehicle), Mixed MSCs and Neurotrophins (Hybrid) , and Ossification center organoid (OCO) into calvarial bone defect, respectively. \u003cstrong\u003ec, d.\u003c/strong\u003e μ-CT evaluation of bone regeneration in Defect control (Defect), Vehicle, Hybrid, and OCO groups at 4 weeks and 8 weeks post-surgery of calvarial defect, respectively. n = 5 rats for per group and per time point. \u003cstrong\u003ee.\u003c/strong\u003e Bone volume fraction (BV/TV) (Exact p value calculated with one-way ANOVA Tukey’s multiple comparisons test: **p = 0.0082, *p = 0.0133); and \u003cstrong\u003ef\u003c/strong\u003e. Bone mass density (BMD) (Exact p value calculated with one-way ANOVA Tukey’s multiple comparisons test: *p = 0.0166, **p = 0.0062) of regenerated tissues in defect area at 4 weeks post-surgery. \u003cstrong\u003eg.\u003c/strong\u003e Bone volume fraction (BV/TV) (Exact p value calculated with one-way ANOVA Tukey’s multiple comparisons test: ****p＜0.0001); and \u003cstrong\u003eh.\u003c/strong\u003e Bone mass density (BMD) (one-way ANOVA Tukey’s multiple comparisons test: ****p＜0.0001) of regenerated tissues in defect area 8 weeks post-surgery. n = 5 biologically independent rats. \u003cstrong\u003ei, j. \u003c/strong\u003eMasson’ Trichrome staining of paraffin sections in Defect, Vehicle, Hybrid, and OCO groups at 4 weeks and 8 weeks after calvarial defect (bar = 200μm at low magnification and bar = 50μm at high magnification).\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-5248946/v1/1f926b2491fc6bc31c4a0e36.png"},{"id":70017276,"identity":"a91896dc-f56c-4bfd-bac2-a73b61ba5b3e","added_by":"auto","created_at":"2024-11-27 14:08:39","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":539096,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIn-situ fusion and maturation of OCO recapitulates ossification center development with concurrent innervation, vascularization and ossification. a, b\u003c/strong\u003e.\u003cstrong\u003e \u003c/strong\u003eHE staining of paraffin sections in Defect, Vehicle, Hybrid, and OCO groups at 4 weeks and 8 weeks after calvarial defect (bar = 200μm at low magnification and bar = 50μm at high magnification). \u003cstrong\u003ec, d\u003c/strong\u003e. Safranin-O staining of paraffin sections in OCO group at 4 weeks and 8 weeks after calvarial defect (bar = 20μm). \u003cstrong\u003ee, f\u003c/strong\u003e. Co-immunofluorescent staining of OCN and NGFR expression in OCO group at 4 weeks and 8 weeks after calvarial defect (bar = 50μm), respectively. \u003cstrong\u003eg\u003c/strong\u003e. Schematic of neurogenesis in the central area and \u003cstrong\u003eh\u003c/strong\u003e. Co-immunofluorescent staining of CGRP and β-III tubulin expression (bar = 20μm); \u003cstrong\u003ei\u003c/strong\u003e. Schematic of angiogenesis in the interlayer and \u003cstrong\u003ej\u003c/strong\u003e. Co-immunofluorescent staining of CD31 and COLX expression (bar = 20μm); \u003cstrong\u003ek\u003c/strong\u003e. Schematic of osteogenesis in the outerlayer and \u003cstrong\u003el\u003c/strong\u003e. Immunofluorescent staining of RUNX2 expression (bar = 100μm); in Defect, Vehicle, Hybrid, and OCO groups at 4 weeks post-surgery.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-5248946/v1/5b15f20768ddb1553c92ec19.png"},{"id":70016295,"identity":"3569ad62-e443-4204-8c44-7240a7f4f828","added_by":"auto","created_at":"2024-11-27 14:00:39","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":797251,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSingle-cell RNA-seq reveals osteo-lineage cell heterogeneity and close interlinks in bone repair. a\u003c/strong\u003e. Schematic diagram of sample isolation, mRNA libraries preparation, and single-cell RNA sequencing of regenerative tissues from calvarial defects among different treatments (Defect, Vehicle, Hybrid, and OCO groups) after 2 weeks post-surgery. \u003cstrong\u003eb\u003c/strong\u003e. UMAP shows the distribution of 39471 cells that divided into 15 cell types, of the repaired tissue in all repaired conditions after 2 weeks post-surgery. \u003cstrong\u003ec\u003c/strong\u003e. UMAP shows the distribution of 9 OLC cell subpopulations after sub-clustering. \u003cstrong\u003ed\u003c/strong\u003e. Heatmap shows expression of marker genes of the 9 sub-clusters. Marker genes were selected according to both p value (adjusted p value \u0026lt; 0.05, Wilcoxon Rank Sum test) and fold change (top 5). \u003cstrong\u003ee\u003c/strong\u003e. Dot-plot showing the GO enrichment in the 9 sub-clusters. Markers used for enrichment analysis were computed using FindAllMarkers function in Seurat. according to p value (*p \u0026lt; 0.05, Wilcoxon Rank Sum test) and fold change (\u0026gt;1). Color bar represented the q values performed by R package clusterProfiler (Benjamini–Hochberg). \u003cstrong\u003ef\u003c/strong\u003e. Dot plots showing the expression of marker genes in the 9 OLC sub-clusters. Dot size represented the proportion of cells expressing specific gene in the indicated subsets and color bar represented the gene expression levels. \u003cstrong\u003eg\u003c/strong\u003e. Differentiation trajectory analysis inferred by RNA velocity of all OCL sub-clusters. \u003cstrong\u003eh-k\u003c/strong\u003e. Partition-based graph abstraction (PAGA) showing the connectivity among subsets. The mean expression of cell community-specific genes (\u003cstrong\u003eh\u003c/strong\u003e, Sdc4, Fmod; \u003cstrong\u003ei\u003c/strong\u003e, Thy1, Cav1; \u003cstrong\u003ej\u003c/strong\u003e, Mfap5, Angptl1; \u003cstrong\u003ek\u003c/strong\u003e, Has1, Il6) were shown in abstracted graph. Line thickness indicated the strength of connectivity. Color bar represents the gene expression levels.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-5248946/v1/dcf7d1ab3d6902c26db47de8.png"},{"id":70017766,"identity":"adae76e1-dd27-4d68-a0c3-a1c445846ced","added_by":"auto","created_at":"2024-11-27 14:16:39","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":1156592,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFour OLC communities with distinct transcriptional signatures and spatial distributions were developed during bone repair process. a. \u003c/strong\u003eUMAP shows the distribution of 4 cell communities according to the differentially expressed genes in community-level specificity. \u003cstrong\u003eb\u003c/strong\u003e. Violin plot showing the 4 cell communities with segregation of distinct subpopulations. \u003cstrong\u003ec\u003c/strong\u003e. Barplot showing the Top8 GO enrichment in the 4 cell communities. Markers used for enrichment analysis were selected according to adjusted p value \u0026lt; 0.05 (Wilcoxon Rank Sum test). Color bar represented the q values performed by R package clusterProfiler (Benjamini-Hochberg). \u003cstrong\u003ed-g\u003c/strong\u003e. Feature plots showing the expression of representative marker genes in the 4 cell communities (Community 1: Has1, Ackr3, Il6, Procr; Community 2: Pdgfra, Mfap5, Dpp4, Lrp1; Community 3: CD200, Clec11a, Sdc4, Fmod; Community 4:Thy1, Cav1, Tnn, Postn). \u003cstrong\u003eh, i\u003c/strong\u003e. Co-immunostaining of CD200, Thy1, and Pdgfra expression of paraffin sections of central layer of regenerative tissue in OCO implantation group at 2 weeks and 4 weeks after defect surgery (bar=30μm), respectively. \u003cstrong\u003ej, k\u003c/strong\u003e. Co-immunostaining of CD200, Thy1, and Pdgfra expression of paraffin sections of peripheral layer of regenerative tissue in OCO implantation group at 2 weeks and 4 weeks after defect surgery (bar=30μm), respectively. \u003cstrong\u003el\u003c/strong\u003e. Immunostaining of Has1 expression of paraffin sections of regenerative tissue Untreated Defect group and OCO implantation group at 2 weeks and 4 weeks after defect surgery (bar=200μm), respectively. At least three times of experiments were repeated independently.\u003cstrong\u003e m\u003c/strong\u003e. Schematic diagram of 4 cell communities with spatially distinct distributions between Defect and OCO group during bone repair process.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-5248946/v1/d1c8bb4069fcb0544cdbb739.png"},{"id":70017279,"identity":"d745439c-b215-47e6-9f5d-7893fb4394bd","added_by":"auto","created_at":"2024-11-27 14:08:39","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":1011546,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDevelopmental Krt8\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e skeletal stem cells is predominantly recruited by the pro-regenerative OC-like organoids during bone regeneration. a\u003c/strong\u003e. UMAP of the 9 OLC subclusters (19019 cells) of regenerative bone tissues in Defect, Vehicle, Hybrid, and OCO groups at 2 weeks post-surgery of calvarial defect. \u003cstrong\u003eb\u003c/strong\u003e. Split UMAP of OLC subclusters in Defect (5782 cells), Vehicle (5725 cells), Hybrid (4180 cells), and OCO (3332 cells) groups. \u003cstrong\u003ec\u003c/strong\u003e. Relative proportion of cells among the above four groups in each subclusters. \u003cstrong\u003ed\u003c/strong\u003e. Relative proportion of Krt8\u003csup\u003e+\u003c/sup\u003e SSCs / Has1\u003csup\u003e+\u003c/sup\u003e Migratory Fibroblasts proportions among four groups\u003cstrong\u003e. e\u003c/strong\u003e. Top 30 GO terms of Krt8\u003csup\u003e+\u003c/sup\u003e Skeletal Stem Cells subset; \u003cstrong\u003ef\u003c/strong\u003e. Top 30 GO terms of Has1\u003csup\u003e+\u003c/sup\u003e Migratory Fibroblasts subset. Markers used for enrichment analysis were selected according to adjusted p value \u0026lt; 0.05 (Wilcoxon Rank Sum test). Color bar represented the q values performed by R package clusterProfiler (Benjamini-Hochberg). \u003cstrong\u003eg\u003c/strong\u003e. Co-immunostaining of Krt8, CD200, and Thy1 expression of paraffin sections of bone regenerative tissue in Defect, Vehicle, Hybrid, and OCO groups at 4 weeks after defect surgery (bar=30μm). \u003cstrong\u003eh\u003c/strong\u003e. Co-immunostaining of Has1 and CD200 expression of paraffin sections of bone regenerative tissue in Defect, Vehicle, Hybrid, and OCO groups at 4 weeks after defect surgery (bar=30μm). At least three times of experiments were repeated independently.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-5248946/v1/b6cdc52c5d35545e08d1e80c.png"},{"id":70017767,"identity":"f1415af2-11d9-4997-a7a5-c51b15f90704","added_by":"auto","created_at":"2024-11-27 14:16:39","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":947359,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCross-species comparisons by machine learning revealed high resemblance of Krt8\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e SSCs in bone regeneration with that in bone development. a\u003c/strong\u003e. Safranin-O staining (bar = 100μm) and \u003cstrong\u003eb\u003c/strong\u003e. co-immunofluorescent staining of Krt8 and Msx1 of paraffin sections in Human child distal phalangeal bone tissues around the epiphysis side (bar = 200μm; bar = 50μm). \u003cstrong\u003ec\u003c/strong\u003e. Safranin-O staining (bar = 100μm) and \u003cstrong\u003ed\u003c/strong\u003e. co-immunofluorescent staining of Krt8 and Msx1 of paraffin sections in Human child proximal phalangeal bone tissues between the epiphysis side and metaphysis side (bar = 200μm; bar = 50μm). \u003cstrong\u003ee\u003c/strong\u003e. Venn plot showing genes from internal and external datasets. \u003cstrong\u003ef\u003c/strong\u003e. Venn plot showing the process of feature selection. \u003cstrong\u003eg\u003c/strong\u003e. Flow chart illustrating construction of our bone-forming OLCs identity machine learning classifier. \u003cstrong\u003eh\u003c/strong\u003e. Bar plot showing all 38 gene features used in our model, avg_log\u003csub\u003e2\u003c/sub\u003eFC represents fold change computed by FindAllMarkers function. \u003cstrong\u003ei\u003c/strong\u003e. UMAPs of subclustered OLCs used for training and testing. UMAPs colored by subpopulations and HGBoot predicted OLC states. \u003cstrong\u003ej\u003c/strong\u003e. Training and testing set of internal bone regeneration dataset in this study, with 80% for training set and 20% for testing set. \u003cstrong\u003eK \u003c/strong\u003eand\u003cstrong\u003e l\u003c/strong\u003e. Integrated and split UMAPs of external datasets (human embryonic long bone and rat bone injury repair) showing the prediction results of OLCs (Krt8+ Skeletal stem cells and Has1+ Migratory Fibroblasts) by the HGBoost model.\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-5248946/v1/5ba63d927f8b9a0be914d567.png"},{"id":86057366,"identity":"70488308-2733-49fb-a8ab-2af6734ab1b8","added_by":"auto","created_at":"2025-07-05 07:07:05","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":9090869,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5248946/v1/c97b824f-0996-41e4-ac2e-45d53b4d1cef.pdf"},{"id":70016309,"identity":"1d183517-5ace-411c-990b-346bb7fe3b11","added_by":"auto","created_at":"2024-11-27 14:00:40","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":2418801,"visible":true,"origin":"","legend":"\u003cp\u003eReporting Summary\u003c/p\u003e","description":"","filename":"reportingsummary.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5248946/v1/e9949bc2d59272b85e38f175.pdf"},{"id":70016310,"identity":"5403ddf7-2abd-44ca-a7e6-6d3eedff86ca","added_by":"auto","created_at":"2024-11-27 14:00:40","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":7635847,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cbr\u003e\u003c/p\u003e","description":"","filename":"SupplementaryInformation.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5248946/v1/e3456ac5b7ec867bdf6f40e0.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Divide-and-conquer strategy with engineered “ossification center” organoids for rapid bone healing via recruiting developmental cell community","fulltext":[{"header":"Introduction","content":"\u003cp\u003eAs large-sized bone defects exceed the limit of self-repair ability, the subsequent long-term recovery often end up with either delayed union or nonunion\u003csup\u003e1\u003c/sup\u003e. Although the transplantation of autologous bone or artificial grafts is common in the clinic, the donor-site morbidity, infections and poor osteointegration still exist. Synthetic materials that developed for bone repair have resulted in limited potential for vascularization and bone growth within the non-biodegradable bulk constructs\u003csup\u003e2,3\u003c/sup\u003e. Biomedical strategies that mimic the mature bone via the delivery of stem cells or growth factors at high-dosages have generated inconsistent results due to limited cell viability, insufficient vascular supply, and inevitable ectopic bone formation\u003csup\u003e4\u003c/sup\u003e. Thus, there is an urgent demand to develop alternative approaches that are capable of long-term retention of resorbable biomaterials, potent stem cells, and physiologically-relevant soluble factors.\u003c/p\u003e\n\u003cp\u003eDevelopmental tissue engineering has recently gained much attention through the design of cell-based constructs with the concept of developmental biology for harsh regenerative conditions\u003csup\u003e5\u003c/sup\u003e. Stem cells as well as morphogenetic factors emerging during the natural healing process have been considered for generating an intermediate avascular cartilaginous anlage, and producing new bone through endochondral ossification \u003cem\u003ein vivo\u003c/em\u003e\u003csup\u003e6-8\u003c/sup\u003e. Mesenchymal stem cell (MSC) aggregates are capable of forming cell spheroids and producing a mineralized matrix and a bone marrow hematopoiesis-supporting niche; however, the scaffold-free constructs require either extended \u003cem\u003ein vitro\u003c/em\u003e chondrogenic induction or native tissue adhesion for their \u003cem\u003ein vivo\u003c/em\u003e application\u003csup\u003e9,10\u003c/sup\u003e. Functional biomaterials enable large-scaling of micrometer-sized assemblies and support biophysical cues and molecular signals for sequential endochondral ossification\u003csup\u003e11,12\u003c/sup\u003e. A recent study from our laboratory has developed an injectable microsphere-based callus-like organoid that highly recapitulates the diverse cell compositions and behaviors of the developing bone via digital light-processing (DLP) printing technology\u003csup\u003e13,14\u003c/sup\u003e. While stem cells within the microsphere underwent simultaneous chondrogenesis and osteogenesis and were protected during local delivery, the interstitial pore structure of the organoid further directed native ECM alignment and blood vessel formation, leading to rapid progression of endochondral regeneration\u003csup\u003e13\u003c/sup\u003e. Obviously, the above strategies require high-cost pre-culture \u003cem\u003ein vitro\u003c/em\u003e with additional growth factors to ensure the maturation of chondrogenesis before implantation.\u003c/p\u003e\n\u003cp\u003eBone morphogen presentation in cellular constructs allows for \u003cem\u003ein situ\u003c/em\u003e chondrogenic priming and enables ready-to-use implantation without additional \u003cem\u003ein vitro\u003c/em\u003e induction. Local delivery of supraphysiological dosages of recombinant human (rh) bone morphogenetic protein-2 (BMP-2) is one of the most popular treatments for large bone defects and is limited by severe ectopic bone formation. Interestingly, developmentally mimicking bone organoids that incorporate exogenous stem cells and BMP-2 induce less ectopic bone formation than BMP-2–containing collagen sponges do\u003csup\u003e8\u003c/sup\u003e, thus providing a promising system for investigating their role in rapid bone regeneration through endochondral ossification \u003cem\u003ein vivo\u003c/em\u003e. Emerging studies have been focusing on the optimization of sustainable co-delivery of angiogenic and osteogenic growth factors by integrated scaffolds for bone repair\u003csup\u003e15-17\u003c/sup\u003e. However, it still involves with delayed revascularization and deficient bone formation upon critical-sized defects\u003csup\u003e16,17\u003c/sup\u003e. Our previous work revealed that neurotrophic factors can efficiently promote the in situ recruitment and expansion of the endogenous Msx1-positive skeletal stem cell subpopulation, which dominates the full-thickness bone reconstruction through an endochondral pathway\u003csup\u003e18\u003c/sup\u003e. These findings indicate that neurotrophins have crucial potential for overcoming the bottleneck of regeneration of critical-sized bone defects.\u003c/p\u003e\n\u003cp\u003eRecent studies have reported the role of neurotrophin-directed innervation in promoting vascularization and osteochondral lineage differentiation during the development of both primary and secondary ossification center\u003csup\u003e19,20\u003c/sup\u003e. The endochondral bone is widely and richly innervated by sensory nerve fibers, and nerve dependence, as a persistent determinant in the stem cell niche, is required for bone fracture repair\u003csup\u003e21\u003c/sup\u003e. As reinnervation precedes angiogenesis and osteogenesis during bone repair, disruption of nerve growth factor (NGF)-TrkA signaling strongly bluntes the revascularization and ossification during the bone fracture healing\u003csup\u003e22\u003c/sup\u003e and calvarial regeneration\u003csup\u003e23,24\u003c/sup\u003e. Bio-printed constructs delivering NGF that mimic the ossification center microenvironment for targeted reinnervation have reconstructed the neuro-vascularized networks for subsequent bone regeneration\u003csup\u003e25\u003c/sup\u003e. Moreover, neuropeptides, including Calcitonin Gene-related Polypeptide (CGRP)\u003csup\u003e26,27\u003c/sup\u003e, Substance-P (SP)\u003csup\u003e28\u003c/sup\u003e, Neuropeptide Y (NPY)\u003csup\u003e29\u003c/sup\u003e, Vasointestinal Peptide (VIP)\u003csup\u003e30\u003c/sup\u003e and Neuropeptide VF (NPVF)\u003csup\u003e31\u003c/sup\u003e , are involved in tissue injury repair\u003csup\u003e28\u003c/sup\u003e. Among them, CGRP is known to be a potent amplifier of osteogenic differentiation\u003csup\u003e26\u003c/sup\u003e, whereas SP is capable of recruiting endogenous MSCs\u003csup\u003e28,32\u003c/sup\u003e and promoting angiogenesis\u003csup\u003e33\u003c/sup\u003e. Nevertheless, the bone healing speed and full-thickness regeneration of critical-sized defects are still challenging because of the lack of developmentally biomimetic biomaterials designed for sequential tissue morphogenesis with superimposed neuropeptides.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eEngineered bone grafts have emerged as alternatives for providing as bioactive carriers with flexible combinations and programmed release patterns of soluble factors that are required for the progress of natural bone repair\u003csup\u003e16,17,34,35\u003c/sup\u003e. However, the insufficient cell-adhering surface of implants and the inconceivable cell migration distance of large injury sizes for endogenous repair greatly limit the in-situ cell proliferation and osteogenesis. Fortunately, the 3D printing technique has endowed the advantages in the manufacturing of microcarriers that encapsulate single cells in thin tunable microgels with spatiotemporally defined patterns of growth factors, as well as abundant cell-adhering surfaces for bone regeneration\u003csup\u003e36\u003c/sup\u003e. In addition, the recruitment and migration of endogenous cells are vital for both niche remodeling and successful bone regeneration in critical-sized defects. However, the efficiency of the microgel-based bioactive cellular constructs as well as the underlying cellular mechanisms at single-cell resolution of such strategies for bone repair remain elusive.\u003c/p\u003e\n\u003cp\u003eIn this study, we first demonstrated that CGRP synergistically induced significantly increased levels of osteogenic differentiation in combination with a physiological dose of BMP-2. We report the availability of the “divide-and-conquer” strategy with multiple spherical microgel-based organoids for effective bone regeneration. We showed that the OC-like organoid (OCO) empowered with bioactive neuropeptides achieved long-term retention and fast\u0026nbsp;bone bridging with\u0026nbsp;successive OC-like bone ossicles across the bone defect. Single-cell RNA analysis revealed that a developmental stem cell community, characterized by unique skeletal stem cell (SSC) and niche cell compositions, was orchestrated by the OCO. Notably, the Krt8\u003csup\u003e+\u003c/sup\u003e SSCs was specifically expanded while the injury-responsive Has1\u003csup\u003e+\u003c/sup\u003e migratory fibroblast subset was concrurrently reduced following OCO implantation. Moreover, cross-species comparisons via machine learning revealed high similarity in Krt8\u003csup\u003e+\u003c/sup\u003e SSCs activation between OCO-dominated bone repair and developmental bone structures. In summary, this study illustrates developmentally inspired “ossification center” organoids harnessing endogenous stem cell subpopulations for rapid bone regeneration.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eCGRP synergistically promoted osteogenesis with BMP-2 at physiological dosage.\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe optimized concentrations for neuropeptides were determined in cell cultures, and no obvious difference in cell proliferation was detected compared with that in the regular growth medium (\u003cstrong\u003eSupplementary\u003c/strong\u003e \u003cstrong\u003eFig. 1a-d\u003c/strong\u003e). To investigate the role of neuropeptides in promoting osteogenesis, we performed cell differentiation experiments under osteogenic inductions, each supplemented with distinct neuropeptides that were reported to be secreted by peripheral nerves\u003csup\u003e37\u003c/sup\u003e. While VIP and NPY supplementation did not have obvious advantages over the regular osteogenic medium, CGRP, on the other hand, significantly increased osteogenic differentiation, as shown by both ALP/ARS staining (\u003cstrong\u003eFig. 1a\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Supplementary\u003c/strong\u003e \u003cstrong\u003eFig. 1e\u003c/strong\u003e) and ALPL gene expression (\u003cstrong\u003eFig. 1b\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Supplementary\u003c/strong\u003e \u003cstrong\u003eFig. 1f\u003c/strong\u003e)\u0026nbsp;after 7 days and 14 days, respectively. To further confirm whether the MSC phenotype was maintained by CGRP, we tested the expression of the cell proliferation marker Ki67 and the cell senescence marker \u0026gamma;h2AX. Immunostaining showed that the expression of Ki67 was significantly increased compared with that in the control medium, whereas the expression of \u0026gamma;h2AX was significantly reduced after CGRP treatment (\u003cstrong\u003eSupplementary\u003c/strong\u003e \u003cstrong\u003eFig. 1g, h\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe further investigated the synergistic effect of the combination of CGRP (10\u003csup\u003e-8\u003c/sup\u003e M) with BMP-2 at a physiological dosage (0.5 \u0026mu;g/ml) on the osteogenic differentiation of MSCs. Results showed that CGRP significantly improved the ALPL and RUNX2 gene expression in MSCs on the basis of BMP-2 after both 7 days and 14 days of osteogenic induction (\u003cstrong\u003eFig. 1c-f\u003c/strong\u003e). ALP staining at day 7 (\u003cstrong\u003eFig. 1g, h\u003c/strong\u003e) and ARS staining at day 14 (\u003cstrong\u003eFig. 1i, j\u003c/strong\u003e) also showed that the enhanced osteogenic differentiation capacity of MSCs induced with CGRP and BMP-2 was greater that of those cultured with either CGRP or BMP-2 alone, indicating the strong osteo-inductivity. The above results showed that CGRP and BMP-2 at low dosages synergistically maximized the osteogenic capacity of MSCs.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026ldquo;All-in-one\u0026rdquo; bioprinting enables efficient engineering of \u0026ldquo;ossification center\u0026rdquo;-like organoid with dual-modular neuropeptides.\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAs innervation has been reported to be indispensable for the development of ossification center and fracture repair through endochondral ossification, in addition to vascularization and mineralization, neuropeptides are hypothesized to potentiate the effects of bone morphogenetic factors. To engineer modular cell constructs that bio-mimic developing ossification centers (OCs), we fabricated \u0026ldquo;all-in-one\u0026rdquo; (AIO) spheroids from MSCs-containing 3D-printed microspheres (2\u0026times;10\u003csup\u003e6\u003c/sup\u003e cells/mL of bioink), together with loading of both CGRP (10\u003csup\u003e-8\u003c/sup\u003e M) and BMP-2 (1 \u0026mu;g/ mL) (\u003cstrong\u003eFig. 2a and Supplementary Fig. 2a\u003c/strong\u003e). Microspheres with a diameter of 100 \u0026mu;m showed limited spherical formability and cell encapsulation (\u003cstrong\u003eSupplementary Fig. 2b\u003c/strong\u003e). Considering the difference in nutrient exchange between cells and their microenvironment, we focused on characterizing cell viability and expansion. These results showed that the microspheres with a diameter of 400 \u0026mu;m had a significantly higher ratio of cell survival and capacity for subsequent cell expansion than did those with a diameter of 200 \u0026mu;m (\u003cstrong\u003eFig. 2b and Supplementary Fig. 2c-e\u003c/strong\u003e). To test the synergistic effect of CGRP and BMP-2 on the cell growth and osteogenic differentiation of MSCs in the microspheres, the same initial number of cells was encapsulated in the AIO spheroids either loaded with BMP-2 alone or combined with CGRP (\u003cstrong\u003eSupplementary Fig. 2f\u003c/strong\u003e). The growth of the cells in the microspheres was observed and assessed by live/dead staining at day 1, 4, 7 and 14. These cells were distributed evenly in the microspheres within the first 7 days, and there was an almost 5-fold increase in the cell number by day 14 in all three groups, with high cell viability and very few dead cells (\u003cstrong\u003eFig. 2c\u003c/strong\u003e). The number of cells that expanded during the 14 days was quantified by Imaris, and there were no significant differences in the percentage of proliferating MSCs among the three groups (\u003cstrong\u003eFig. 2d\u003c/strong\u003e). Together, the AIO spheroids with a diameter of 400 \u0026mu;m show high biocompatibility and are beneficial for cell assembly.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo more directly identify the osteogenic potential of the rapidly expanded MSCs in the AIO spheroids, osteogenic induction was performed after 1 day of pre-treatment in growth medium (\u003cstrong\u003eFig. 2e\u003c/strong\u003e). Compared with those in the control or BMP-2-supplemented osteogenic medium, the levels of ALP synthesis and ARS, as well as their inner distribution, were greatly enhanced in the AIO spheroids(\u003cstrong\u003eFig. 2f, g\u003c/strong\u003e). Consistent with the increased ALP levels and mineralized nodule formation shown by ARS staining, the expression of osteogenic markers, such as RUNX2 and OCN, was also obviously promoted in AIO spheroids after 7 days (\u003cstrong\u003eFig. 2h\u003c/strong\u003e) and 14 days of osteogenic induction (\u003cstrong\u003eFig. 2i and Supplementary Fig. 2g\u003c/strong\u003e), respectively. After 2 weeks of induction, a much higher percentage of RUNX2- and COL1-posive cells were observed at the early stage, demonstrating increased sensitivety and rapid osteogenesis in response to the synergistic effect of CGRP and BMP-2 (\u003cstrong\u003eFig. 2h, i and Supplementary Fig. 2h, i\u003c/strong\u003e). Collectively, these data suggest that the AIO spheroids not only enable high cell viability, but also serve as ideal units for \u0026ldquo;ossification center\u0026rdquo;-like organoid assembly with highly efficient osteogenesis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe \u0026ldquo;divide-and-conquer\u0026rdquo; strategy achieves fast bone bridging with collective implantation of OC-like bone organoids.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGelMA and GelMA/HA-NB hydrogels were synthesized and the differential degradation was tested by subcutaneous implantation (\u003cstrong\u003eSupplementary Fig. 3a\u003c/strong\u003e). Compared with the GelMA hydrogel, GelMA/HA-NB obviously had a much slower \u003cem\u003ein vivo\u003c/em\u003e degradation rate (\u003cstrong\u003eSupplementary Fig. 3b, c\u003c/strong\u003e), while both hydrogels showed comparable biocompatibility within 4 weeks (\u003cstrong\u003eSupplementary Fig. 3d\u003c/strong\u003e). To test the ability of the combined hydrogels as a dual-modular drug delivery system, via DLP 3D printing, we developed \u0026ldquo;inner-core\u0026rdquo; GelMA/HA-NB microspheres capable of incorporating growth factor A and a uniformly wrapped \u0026ldquo;outer-shell\u0026rdquo; module with the GelMA hydrogel containing growth factor B (\u003cstrong\u003eSupplementary Fig. 3e\u003c/strong\u003e). Because of they are crucial for the ossification center development, BMP-2 (5 \u0026mu;g/mL) and VEGF (5 \u0026mu;g/mL) are capable of distinctly spatiotemporal release at physiological dosages after the integrated incorporation, largely due to their different hydrogel degradation kinetics (\u003cstrong\u003eSupplementary Fig. 3f\u003c/strong\u003e). Interestingly, the release of VEGF was much faster than that of BMP-2 in the first 2 days, and both VEGF and BMP-2 started to be released slowly and continuously for up to 28 days (\u003cstrong\u003eSupplementary Fig. 3g\u003c/strong\u003e). To further verify their potential in directing cell migration, we performed live/dead staining to identify MSCs that were seeded on the surface of this dual-modular bioactive construct. The results showed much closer cell-cell contacts and deeper migration distances after the incorporation of BMP-2 and VEGF, than after the addition of the pure hydrogel vehicle (\u003cstrong\u003eSupplementary Fig. 3h, i\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eInspired by the spatiotemporal release pattern of the dual-modular hydrogel system in our previous results, we broadened our horizon by adopting novel neuropeptides for recruiting endogenous stem cells to strengthen the new bone formation based on the AIO spheroids. As we have introduced previously, SP is known for recruiting endogenous MSCs to local injuries\u003csup\u003e28\u003c/sup\u003e. Interestingly, after\u0026nbsp;supplementation with growth medium for MSC culture, SP at various concentrations resulted in increased levels of cell proliferation\u0026nbsp;(\u003cstrong\u003eSupplementary Fig. 4a\u003c/strong\u003e). Among them, SP with a concentration of 1\u0026nbsp;\u0026times;\u0026nbsp;10\u003csup\u003e-7\u0026nbsp;\u003c/sup\u003eM promoted a significantly higher level of VEGF secretion than the other\u0026nbsp;concentrations\u0026nbsp;did\u0026nbsp;(\u003cstrong\u003eSupplementary Fig. 4b\u003c/strong\u003e), indicating that SP increased the proangiogenic potential of MSCs. Further validation on day 3 via\u0026nbsp;real-time qPCR showed that SP significantly promoted the VEGF gene expression in MSCs (\u003cstrong\u003eSupplementary Fig. 4c\u003c/strong\u003e). We upgraded the dual-modular hydrogel system with the \u0026ldquo;outer-shell\u0026rdquo; module containing SP (\u003cstrong\u003eSupplementary Fig. 4d\u003c/strong\u003e), which was then coated on a 3D-printed PLA scaffold for cell migration assessment (\u003cstrong\u003eSupplementary Fig. 4e\u003c/strong\u003e). MSCs that were seeded on the top surface of the integrated bone graft resulted in a larger cell spreading area and more closely intersected cell-cell contacts, as well as a much longer migration distance after SP supplementation (\u003cstrong\u003eSupplementary Fig. 4f\u003c/strong\u003e). By subcutaneous transplantation, SP obviously promoted the local cell recruitment and red blood cell aggregation, whereas MSC encapsulation further accelerated the effect of SP on vascularization (\u003cstrong\u003eSupplementary Fig. 4g\u003c/strong\u003e). Moreover, immunostaining of interstitial tissue among the microspheres demonstrated enhanced blood vessel formation with CD31 expression and extensive osteogenesis with RUNX2 when the microspheres were loaded with both \u0026ldquo;inner-core\u0026rdquo; MSCs and \u0026ldquo;outer-shell\u0026rdquo; SP (\u003cstrong\u003eSupplementary Fig. 4h\u003c/strong\u003e). The above results indicated that the spatial-peripheral release of the neuropeptide SP could further enhance the cell-recruiting and proangiogenic capacity of the cellular construct for the design of OC-like organoids.\u003c/p\u003e\n\u003cp\u003eTo evaluate their potential for bone regeneration, we transplanted the integrated \u0026ldquo;ossification center\u0026rdquo;-like organoid (OCO) consisting of AIO spheroids and free SP into a calvarial defect in SD rats using in-situ photo-crosslinking (\u003cstrong\u003eFig. 3a and Supplementary Fig. 5a\u003c/strong\u003e). In addition to OCO, the dual-modular hydrogel (Vehicle) and the combination of MSCs and Neurotrophins (Hybrid) were used as parallel controls (\u003cstrong\u003eFig. 3b\u003c/strong\u003e). 4 weeks after surgery, only mild new bone formation was observed at the edge of the defect in both the Vehicle and the Hybrid groups, whereas OCO transplantation resulted in nearly complete\u0026nbsp;bone bridging along with honeycomb-like structures\u0026nbsp;(\u003cstrong\u003eFig. 3c\u003c/strong\u003e). At 8 weeks post-surgery, although diffusive bone blocks had formed within the defect in the Vehicle group and a much thicker neo-bone had appeared\u0026nbsp;at the inner surface of the defect in the Hybrid group,\u0026nbsp;a large amount of defect area\u0026nbsp;still existed to be filled.\u0026nbsp;Consistent with the results in the early stage, new bone formation by OCO was shown to essentially complete the regeneration of the entire defect area\u0026nbsp;(\u003cstrong\u003eFig. 3d\u003c/strong\u003e). Significantly higher levels of bone volume, bone mineral density, as well as new-old bone integration in the OCO group were also observed at\u0026nbsp;4 weeks (\u003cstrong\u003eFig. 3c, e, f\u003c/strong\u003e) and 8 weeks (\u003cstrong\u003eFig. 3d, g, h\u003c/strong\u003e), respectively. Histological analysis by Masson\u0026rsquo;s Trichrome staining revealed that, compared with the control or the Hybrid setting, treatment with the dual-modular hydrogel vehicle alone promoted the matrix reorganization and bone healing to some extent (\u003cstrong\u003eFig. 3i, j\u003c/strong\u003e). Interestingly, by combining the advantages of microstructure and bioactive factors, OCO triggered rapid bone healing\u0026nbsp;with\u0026nbsp;successive OC-like bone ossicles across the defect as early as 4 weeks after transplantation (\u003cstrong\u003eFig. 3i and Supplementary Fig. 5b\u003c/strong\u003e). Full-thickness (internal and external layers of cortical bone; middle layer of trabecular bone and the bone marrow cavity) reconstruction of the calvarial bone defect was further achieved following \u003cem\u003ein-situ\u003c/em\u003e OCO fusion (\u003cstrong\u003eFig. 3j and Supplementary Fig. 5c\u003c/strong\u003e). The above results suggest that this \u0026ldquo;divide-and-conquer\u0026rdquo; strategy enables fast bone bridging and full-thickness bone regeneration through \u003cem\u003ein-situ\u003c/em\u003e fusion of multiple OC-like organoids.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIn-situ fusion and maturation of OCOs promote concurrent innervation, vascularization, and ossification.\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo determine whether in-situ OCO grafting facilitates the initiation of OC development and the subsequent osteogenesis, we further examined the biological processes and cellular events during the bone regeneration. Distinct from Vehicle and Hybrid, the OCO recruited more abundant cells and guided extensive ECM deposition around the spherical organoids, as shown by the details of the HE staining results (\u003cstrong\u003eFig. 4a, b\u003c/strong\u003e). With the OC-like organoids fusion and development after 4 weeks, the sections harvested from repaired tissues were found to retain a cartilage-like tissue rich in proteoglycan (\u003cstrong\u003eFig. 4c\u003c/strong\u003e), which was dynamically absorbed and replaced with bone-like tissue after 8 weeks (\u003cstrong\u003eFig. 4d\u003c/strong\u003e), and finally turned into the bone marrow cavity, as shown by OCN and NGFR expression (\u003cstrong\u003eFig. 4e, f\u003c/strong\u003e). The initiation and maturation of OCO were then determined by the expression of multiple cellular markers indicating sequential neurogenesis (\u003cstrong\u003eFig. 4g\u003c/strong\u003e), angiogenesis (\u003cstrong\u003eFig. 4i\u003c/strong\u003e) and osteogenesis (\u003cstrong\u003eFig. 4k\u003c/strong\u003e). At 4 weeks after transplantation, there were sporadic \u0026beta;-III tubulin-positive cells in the injured area indicating limited reinnervation in both the Vehicle and Hybrid settings, while OCO stimulated robust aggregation and distribution of both\u0026nbsp;\u0026beta;-III tubulin- and CGRP- positive sensory neuron cells (\u003cstrong\u003eFig. 4h\u003c/strong\u003e). Followed by hypertrophic maturation (COLX expression) of the cartilage-like tissue (\u003cstrong\u003eFig. 4c, j\u003c/strong\u003e), the vascular invasion that needed for endochondral ossification was also revealed by staining with the endothelial marker CD31(\u003cstrong\u003eFig. 4j and Supplementary Fig. 5d\u003c/strong\u003e). Along with the enhanced innervation and vascularization, the multi-OCO inner cores were eventually surrounded by more abundant RUNX2- and OCN-positive cells (\u003cstrong\u003eFig. 4l and Supplementary Fig. 5e\u003c/strong\u003e) and Collagen I expression, when compared with negligeable distributions in other conditions (\u003cstrong\u003eSupplementary Fig. 5f\u003c/strong\u003e). These results strongly suggest that the in-situ OCO implantation promotes rapid bone regeneration may partly through an endochondral ossification pathway.\u003c/p\u003e\n\u003cp\u003eThese findings collectively indicate that the \u0026ldquo;ossification center\u0026rdquo;-like organoids (OCO), comprising the inner-core bone morphogenetic MSCs spheroids generated via DLP 3D printing and an interstitially distributed outer-shell proangiogenic module, have significant potential for preserving bioactive factors and harnessing endogenous stem cells to facilitate bone regeneration. More importantly, the in-situ fusion and maturation of OCO promote bone repair with tissue morphological changes in the early stage characterized by innervation, vascularization and ossification, rather than monotypic fibrotic tissue filling after bone injury.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDistinct osteo-lineage cells organize into cell communities that have specific functions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo determine the bone regenerative cell atlas in response to the abundant biological processes, we performed single-cell RNA-sequencing analysis of the cell clusters after the in-situ OCO fusion and maturation for 2 weeks after implantation. Cells from the regenerated tissues were isolated, and mRNA libraries were prepared and sequenced (\u003cstrong\u003eFig. 5a\u003c/strong\u003e). For quality control, cells with genes detected over 300 genes detected, read counts between 1000 and 25000, and mitochondrial gene expression less than 15% were retained. After filtering, we obtained data from 39471 cells: with 11913 cells in the Defect group, 10953 cells in the Vehicle group, 8208 cells in the Hybrid group, and 8397 cells in the OCO group from 2 weeks after surgery. Following the Seurat pipeline, the data were log-normalized and then scaled. Uniform Manifold Approximation and Projection (UMAP) was calculated to visualize cell heterogeneity in reduced dimensions.\u003c/p\u003e\n\u003cp\u003eIntegrated analysis revealed 15 cell clusters with known cell markers that formed during the first two weeks of bone repair (\u003cstrong\u003eSupplementary Fig. 6a, b\u003c/strong\u003e). On the basis of the understanding of the single cell types in our previous bone repair study as well as in other publications, we quite completely identified a full range of cell types for bone regeneration: B/T cells (expressing Ms4a1 and Cd3g); Monocyte/Macrophage (expressing Cd68 and Adgre1); Endothelial cells (expressing Emcn and Cdh5); Osteo-lineage cells, OLCs (expressing Col1a1 and Prrx1); Pericytes (expressing Mcam and Acta2); and Neural crest cells (expressing Plp1 and Sox10) (\u003cstrong\u003eSupplementary Fig. 6c\u003c/strong\u003e). All the cell clusters are distributed among the above four conditions, namely the Defect, Vehicle, Hybrid, and OCO groups, without obvious imbalance (\u003cstrong\u003eSupplementary Fig. 6d\u003c/strong\u003e). The relative proportion of each cluster showed that several OLC clusters underwent considerable variations in response to the OCO maturation, while macrophages, CD8\u003csup\u003e+\u003c/sup\u003e T cells and endothelial cells showed only mild differences (\u003cstrong\u003eSupplementary Fig. 6e\u003c/strong\u003e). As our original intention was to figure out the changes in the recruited osteogenesis-related stem/progenitor cells at single cell resolution, the subsequent data analysis focused mainly on these OLCs. Collectively, these data provide a single-cell atlas of the regenerative bone tissues, which enables us to reveal the cellular heterogeneity of various cell types, especially the osteo-lineage cells of interest.\u003c/p\u003e\n\u003cp\u003eTo further dissect the cell heterogeneity and classify the OLC clusters, we merged the OLCs (19019 cells were obtained) in all the groups for sub-clustering (\u003cstrong\u003eFig. 5b, c\u003c/strong\u003e). The OLCs were divided into 9 novel cell subpopulations on the basis of the abundant and specific expression of osteogenic genes (\u003cstrong\u003eFig. 5d\u003c/strong\u003e) and distinct GO features (\u003cstrong\u003eFig. 5e\u003c/strong\u003e). These 9 OLC subpopulations were annotated, in particular, Subcluster 0: Has1\u003csup\u003e+\u003c/sup\u003e Migratory Fibroblasts (expressing Has1, Ccl2, Ackr3 and Procr); Subcluster 1: Mfap5\u003csup\u003ehi\u003c/sup\u003e Mesenchymal Progenitors (expressing Mfap5, Pdgfra, Fbn1 and Lrp1); Subcluster 2: Thy1\u003csup\u003ehi\u003c/sup\u003e Skeletal Stem Cells (expressing Thy1, Cav1, Col5a3 and Cdkn2a); Subcluster 3: Krt8\u003csup\u003e+\u003c/sup\u003e Skeletal Stem Cells (expressing Krt8, CD200, Mdk and Sdc4); Subcluster 4: Ucma\u003csup\u003e+\u003c/sup\u003e Osteochondral Progenitors (Ucma, Fmod, Col11a1 and Clec11a); Subcluster 5: Chondroblasts (expressing Ctgf, Tnmd, Sox9 and Sfrp2); Subcluster 6: Osteoprogenitors (expressing Postn, Sfrp4, Runx2 and Tnn); Subcluster 7: Lyz2\u003csup\u003e+\u003c/sup\u003e Fibroblasts (expressing Lyz2, CD74, and Cxcl2); and Subcluster 8: Proliferating Mesenchymal Cells (expressing Ube2C, Cdkn3, and Mki67) (\u003cstrong\u003eFig. 5f\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Supplementary Fig. 7a\u003c/strong\u003e). To analyze the lineage hierarchy of these heterogenous subpopulations, pseudo-time analysis with RNA velocity was performed to explore the lineage relationships among the osteo-lineage cell subsets. These results displayed that there are three main differentiation trajectories that originate from Krt8\u003csup\u003e+\u003c/sup\u003e Skeletal Stem Cells (Subcluster 3), Proliferating Mesenchymal Cells (Subcluster 8) and Mfap5\u003csup\u003ehi\u003c/sup\u003e Mesenchymal Progenitors (Subcluster 1), respectively; while Has1\u003csup\u003e+\u003c/sup\u003e Migratory Fibroblasts (Subcluster 0) is situated at the downstream of the differentiation hierarchy as a common destination (\u003cstrong\u003eFig. 5g\u003c/strong\u003e). Further analysis using 3D RNA velocity from different spatial views also confirmed the differentiation branches originating from Subcluster 3, Subcluster 8 and Subcluster 1, as well as the downstream Subcluster 0 (\u003cstrong\u003eSupplementary\u003c/strong\u003e \u003cstrong\u003eFig. 7b-e\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe next sought to understand how cell subclusters assemble into cellular neighborhoods that organize into multi-lineages for in-situ OCO maturation. Together with the 3D RNA velocity, which demonstrated a cluster effect with differentiation continuums: subcluster 3-4-5 (\u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;Fig. 7c\u003c/strong\u003e), subcluster 8-2-6 (\u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;Fig. 7d\u003c/strong\u003e), and subcluster 5-1-0 / subcluster 5-0 (\u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;Fig. 7e\u003c/strong\u003e), Partition-based graph abstraction (PAGA) analysis revealed unique combinations of specific cell subclusters and exhibited cellular continuity. Accordingly, these subclusters are well-organized into 4 independent cell communities (cell subclusters with common molecular features and a spatial mapping of specific cell lineage structures) that each separately featured by the expression of Sdc4 and Fmod (\u003cstrong\u003eFig. 5h\u003c/strong\u003e); Thy1 and Cav1 (\u003cstrong\u003eFig. 5i\u003c/strong\u003e); Mfap5 and Angptl1 (\u003cstrong\u003eFig. 5j\u003c/strong\u003e); and Has1 and Il6 (\u003cstrong\u003eFig. 5k\u003c/strong\u003e). These results strongly suggest that OLC subpopulations involved in the bone repair are actually of diverse origins and functionally different yet closely interlinked.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe dominant role switch of cell communities with distinct transcriptional signatures and spatial distributions represents bone regeneration or nonunion\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo better define the above 4 cell communities (CC) with unique characteristics, we first analyzed the differentially expressed genes (DEGs) with community-level specificity showing the segregation of these subpopulations (\u003cstrong\u003eFig. 6a, b\u003c/strong\u003e). GO enrichment analysis of the TOP 8 ranked GO terms based on the DEGs showed that CC1 and CC2 were differentially enriched with genes involved in \u0026ldquo;Respond to Reactive Oxygen Species/Chemical Stress\u0026rdquo; and \u0026ldquo;Basement Membrane\u0026rdquo;, respectively; while CC3 and CC4 were differentially enriched with genes involved in \u0026ldquo;Ossification\u0026rdquo; and \u0026ldquo;Cell-substrate Adhesion\u0026rdquo;, respectively (\u003cstrong\u003eFig. 6c\u003c/strong\u003e). By visualizing these differences with UMAP, we found that cell migration, chemotaxis, and inflammation-associated genes, such as Has1, Ackr3, Il6 and Procr, were enriched in CC1; when fibrosis-associated and immune-regulatory genes, such as Pdgfra, Mfap5, Dpp4 and Lrp1 were enriched in CC2 (\u003cstrong\u003eFig. 6d, e\u003c/strong\u003e); in contrast, skeletal stem/progenitor cells and osteochondral progenitors-associated genes, such as CD200, Clec11a, Sdc4, and Fmod were specifically enriched in the CC3; while osteoprogenitors-associated genes, such as Thy, Cav1, Tnn, and Postn were specifically enriched in the CC4 (\u003cstrong\u003eFig. 6f, g\u003c/strong\u003e). To this end, we have primarily identified these cell communities from OLCs that were also have spatially distinct distributions, and defined the cellular neighbors for each cell subcluster.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe further detected the CC distributions by immunostaining of regenerative tissues during the bone regeneration process. At the early stage of bone repair (2 weeks after surgery), representative CC3 maker (CD200) and CC4 marker (Thy1) were both located interstitially among OCO grafts in the central layer of the regenerated tissues (\u003cstrong\u003eFig. 6h\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Supplementary Fig. 8a, b\u003c/strong\u003e). Consistent with the gene expressions observed in the UMAP among the 4 communities, CD200 was located in the OCO inner surface, while Thy1 was found relatively in the OCO outer surface, at 4 weeks after bone injury (\u003cstrong\u003eFig. 6i\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Supplementary Fig. 8c, d\u003c/strong\u003e). In the peripheral layer of the regenerated tissues, except for the relatively few Pdgfra (a CC2 marker) distributions on the surface, abundant CD200 and Thy1 were actively expressed in most of the migrating cells at 2 weeks (\u003cstrong\u003eFig. 6j\u003c/strong\u003e), and clear and distinct spatial distributions were observed at 4 weeks (\u003cstrong\u003eFig. 6k\u003c/strong\u003e). Interestingly, enormous cells marked by the expression of the CC1 marker Has1 were found nearly throughout the whole fibrotic tissue layer in the untreated Defect group, as early as in the 2 weeks post-injury. Even after 4 weeks, the Has1-positive cells still widely resided on both sides of the fibrotic tissue, showing nonunion of the bone injury. In contrast, few Has1 distribution was detected in both the upper and lower surfaces of the OCO grafts in the peripheral tissue layer during the first 2 weeks, and Has1 was hardly detected after 4 weeks of injury (\u003cstrong\u003eFig. 6l\u003c/strong\u003e). We further compared the functional changes in the regulation of the cell communities caused by OCO implantation with those caused by untreated defects. Differential Gene Ontology (GO) analysis based on OCO/Defect DEGs (Differential Regulated Genes) at the CC level (\u003cstrong\u003eSupplementary Fig\u003c/strong\u003e\u003cstrong\u003e. 9a, b\u003c/strong\u003e) revealed that \u0026ldquo;Regulation of vasculature development\u0026rdquo;, \u0026ldquo;Regulation of angiogenesis\u0026rdquo;, and \u0026ldquo;rRNA metabolic process\u0026rdquo; were up-regulated in Cell Community 3 (\u003cstrong\u003eSupplementary Fig\u003c/strong\u003e\u003cstrong\u003e. 9c\u003c/strong\u003e); and \u0026ldquo;Response to hypoxia\u0026rdquo;, \u0026ldquo;Response to corticosteroid\u0026rdquo;, \u0026ldquo;Response to mechanical stimulus\u0026rdquo;, and \u0026ldquo;Response to transforming growth factor beta\u0026rdquo; were down-regulated in Cell Community 1 (\u003cstrong\u003eSupplementary Fig\u003c/strong\u003e\u003cstrong\u003e. 9d\u003c/strong\u003e). These results indicated that, in addition to the differences in spatial distribution, the cellular functions of specific cell communities also underwent changes after OCO implantation.\u003c/p\u003e\n\u003cp\u003eThe above findings suggested that the dominant role switch of cell communities (CC3 vs CC1) with distinct transcriptional signatures and spatial distributions could more precisely represent the state of bone regeneration or nonunion (OCO vs Defect) (\u003cstrong\u003eFig. 6m\u003c/strong\u003e). Together, these OLC communities not only display diverse organizations of transcriptionally different cell subpopulations, but also feature specific cell subpopulations that drive their own community functionality according to the cellular ecosystem. This findings support the concept that stem/progenitor cells may adopt bone developmental characteristics based on their own functions as well as the microenvironmental cells within each cell community.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe developmentally mimicking cell community that dominated with Krt8\u003csup\u003e+\u003c/sup\u003e SSCs was orchestrated by the pro-regenerative OC-like organoids.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo further test whether the OC-like organoids induced rapid bone regeneration by harnessing the endogenous skeletal stem/progenitor cells, we first focused on the differences in the cellular compositions and different modulations of the OLC subclusters. At 2 weeks post-injury, comparable cell numbers of all 9 subclusters were demonstrated among the 4 repair conditions (\u003cstrong\u003eFig. 7a, b\u003c/strong\u003e). Remarkably, the proportion of cells in subcluster 3 (Krt8\u003csup\u003e+\u003c/sup\u003e Skeletal Stem Cells, CC3 representative) exhibited significant increase in response to OCO implantation compared with that in the Defect, Vehicle, and Hybrid groups. In contrast, subcluster 0 (Has1\u003csup\u003e+\u003c/sup\u003e Migratory Fibroblasts, CC1 representative) accounted for a substantial proportion in the untreated Defect group compared to all the other groups (\u003cstrong\u003eFig. 7c\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Supplementary Fig. 7f-i\u003c/strong\u003e). These results strongly suggested that Krt8\u003csup\u003e+\u003c/sup\u003e SSCs and Has1\u003csup\u003e+\u003c/sup\u003e Migratory Fibroblasts were predominantly recruited and expanded by pro-regenerative and non-regenerative treatments, respectively. Based on such a marked contrast in the proportion of the osteo-lineage cells under the four repair conditions, we further found that the relative ratio of the two cell proportions (Krt8\u003csup\u003e+\u003c/sup\u003e SSCs / Has1\u003csup\u003e+\u003c/sup\u003e Migratory Fibroblasts) could more accurately reflect the disparity in the degree of bone regeneration degree among the groups (\u003cstrong\u003eFig. 7d\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Fig. 3i, j\u003c/strong\u003e). Apparently, distinct osteo-lineage cell compositions and their relative cell proportions are orchestrated in response to various strategies for bone injury repair.\u003c/p\u003e\n\u003cp\u003eTo understand the biological significance of the relative ratio of the two cell proportions, we next explored the cell characteristics and differential signatures of the above two OLC subclusters. As shown in our previous results, Krt8, a known marker for keratin filaments in epithelial cells, was identified to be OLCs-specific among all cell types (\u003cstrong\u003eSupplementary Fig\u003c/strong\u003e\u003cstrong\u003e. 10a, b\u003c/strong\u003e) and subcluster 3-specific by comparing subcluster 3 to the rest of the OLCs (\u003cstrong\u003eSupplementary Fig\u003c/strong\u003e\u003cstrong\u003e. 10c, d\u003c/strong\u003e). Moreover, pseudo-time analysis of OLCs showed that Krt8\u003csup\u003e+\u003c/sup\u003e SSCs were located in the upstream of the osteochondral lineage of the differentiation trajectory (\u003cstrong\u003eFig. 5g, h\u003c/strong\u003e and \u003cstrong\u003eSupplementary Fig\u003c/strong\u003e\u003cstrong\u003e. 10e, f\u003c/strong\u003e), indicating that despite demonstrating an epithelial signature, the Krt8\u003csup\u003e+\u003c/sup\u003e SSCs at the same time possess strong potential for osteochondral lineage differentiation, which supports our previous conclusion that OCO maturation promotes bone regeneration through an endochondral ossification pathway. As the progenitors of CC3, Krt8\u003csup\u003e+\u003c/sup\u003e SSCs thus perfectly mark the developmentally mimicking OLC community, which is predominantly recruited after the pro-regenerative OCO implantation. The top 30 GO terms of Krt8\u003csup\u003e+\u003c/sup\u003e SSCs showed bone repair-associated characteristics including \u0026ldquo;Ossification\u0026rdquo;, \u0026ldquo;External encapsulating structure\u0026rdquo;, \u0026ldquo;Regulation of vascularization\u0026rdquo;, and \u0026ldquo;Response to decreased oxygen levels\u0026rdquo; (\u003cstrong\u003eFig. 7e\u003c/strong\u003e). Meanwhile, Has1\u003csup\u003e+\u003c/sup\u003e Migratory Fibroblasts, which are mainly distributed at the end of the differentiation trajectory (\u003cstrong\u003eFig. 5g, j\u003c/strong\u003e and\u003cstrong\u003e\u0026nbsp;Supplementary Fig\u003c/strong\u003e\u003cstrong\u003e. 10e, f\u003c/strong\u003e), were active as the major cells of CC1 shown in the untreated bone injury (\u003cstrong\u003eFig. 7c\u003c/strong\u003e). The analysis of the top 30 GO terms in Has1\u003csup\u003e+\u003c/sup\u003e Migratory Fibroblasts showed that it was more likely to function as an injury-responsive OLC subcluster with a higher capability of migration, which was clearly involved in GO terms, such as \u0026ldquo;Positive regulation of cytokine production\u0026rdquo;, \u0026ldquo;Responsive to reactive oxygen species\u0026rdquo;, \u0026ldquo;Responsive to interleukin-1\u0026rdquo;, \u0026ldquo;Responsive to chemical stress\u0026rdquo;, and \u0026ldquo;Responsive to oxidative stress\u0026rdquo; (\u003cstrong\u003eFig. 7f\u003c/strong\u003e). These findings demonstrated that the implantation of bioactive materials leads to the dominance of the corresponding \u0026ldquo;effector cell subset\u0026rdquo; in the specific cell community, which in turn can accurately depict the underlying molecular characteristics of particular bone repair scenario.\u003c/p\u003e\n\u003cp\u003eIn consistent with the above results, verification by immunostaining showed that the presence of Krt8-positive OLCs (CC3 representative) as well as CD200 expression (CC3 marker) were greatly increased and distributed within the spherical OCO construct compared with those in the Defect, Vehicle, and Hybrid groups (\u003cstrong\u003eFig. 7g, h\u003c/strong\u003e). Interestingly, the Thy1-positive OLCs (CC4 representative) were more likely distributed at the interface of multiple OCO constructs, in which few of them were also krt8 positive at both 2 weeks and 4 weeks post-surgery (\u003cstrong\u003eSupplementary Fig\u003c/strong\u003e\u003cstrong\u003e. 8b, d\u003c/strong\u003e). In contrast, the Has1-positive OLCs (CC1 representative) were widely dispersed throughout the whole layers of repaired tissue in the Defect and Hybrid groups, while Has1 expression was significantly reduced in the treatment groups with either pure microspheres (Vehicle) or OC-like organoids (OCO) (\u003cstrong\u003eFig. 7h\u003c/strong\u003e). Up to this point, we have shown that microspheres-based repair strategies have structural advantages over pure bioactive ingredients for equalizing the ability of all OLCs to migrate into bone injuries. In the context of cellular compositions, these findings indicate that the relative ratio of Krt8\u003csup\u003e+\u0026nbsp;\u003c/sup\u003eSSCs/Has1\u003csup\u003e+\u003c/sup\u003e Migratory Fibroblasts could serve as a more precise indicator of rapid bone regeneration with the pro-regenerative OC-like organoids at single-cell resolution, as opposed to traditional studies that only focused on one individual cell subpopulation. Collectively, the specific osteo-lineage cell composition in a developmentally-mimicking cell community with Krt8\u003csup\u003e+\u003c/sup\u003e SSCs expansion was cultivated by the pro-regenerative OC-like organoids during rapid bone regeneration. The obvious expansion of Krt8\u003csup\u003e+\u003c/sup\u003e SSCs after OCO implantation prompted us to explore its contribution to the \u003cem\u003ein-situ\u003c/em\u003e OCO maturation for new bone formation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBone regenerative Krt8\u003csup\u003e+\u003c/sup\u003e SSCs share high similarities to those found in the ossification center during bone development.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn order to confirm the rationality of Krt8\u003csup\u003e+\u003c/sup\u003e SSCs in presenting the developmentally mimicking OLC community, we carefully analyzed the subsets of both regenerative and developmental tissues involved in new bone formation. With the unique subset specificity of Krt8 shown in our single-cell RNA-seq data (\u003cstrong\u003eSupplementary Fig\u003c/strong\u003e\u003cstrong\u003e. 10a-d\u003c/strong\u003e), we found that Msx1 perfectly labels both Krt8\u003csup\u003e+\u003c/sup\u003e SSCs and Has1\u003csup\u003e+\u003c/sup\u003e Migratory Fibroblasts by using Krt8\u003csup\u003e+\u003c/sup\u003e Msx1\u003csup\u003e+\u0026nbsp;\u003c/sup\u003edouble positive staining (\u003cstrong\u003eSupplementary Fig\u003c/strong\u003e\u003cstrong\u003e. 11a, b\u003c/strong\u003e), in which Krt8 and Msx1 together better enable the identification of the osteo-lineage cell subset with high potential for osteochondral differentiation\u003csup\u003e18\u003c/sup\u003e, as well as its participation in the endochondral ossification pathway, as shown by Mgp and Alpl expression (\u003cstrong\u003eSupplementary Fig\u003c/strong\u003e\u003cstrong\u003e. 11b\u003c/strong\u003e). To our surprise, immunostaining of the most active repair area was shown in the interstitial spaces (outer area) of OC-like organoids (inner area), in which the Krt8\u003csup\u003e+\u003c/sup\u003e Msx1\u003csup\u003e+\u003c/sup\u003e double positive cells were mainly located at the bone-forming front (\u003cstrong\u003eSupplementary Fig\u003c/strong\u003e\u003cstrong\u003e. 11c\u003c/strong\u003e). To obtain evidence associated with clinical relevance, we validated the emergence of Krt8\u003csup\u003e+\u003c/sup\u003e SSCs in human bone tissues at developmental stages by immunostaining assays. In human child phalangeal bone samples, an ossification center-like structure appeared in the anatomical fingertip and exhibited lacuna cell distribution between the bone-forming front and the immature bone epiphysis, which resembled the limb development of the secondary ossification center (\u003cstrong\u003eFig\u003c/strong\u003e\u003cstrong\u003e. 8a\u003c/strong\u003e). Krt8 expression was abundantly distributed in the tip side of the bone-forming front, whereas Krt8\u003csup\u003e+\u003c/sup\u003e Msx1\u003csup\u003e+\u0026nbsp;\u003c/sup\u003edouble positive cells were aggregated in the central region of the bone morphogenetic center (\u003cstrong\u003eFig\u003c/strong\u003e\u003cstrong\u003e. 8b\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Supplementary Fig. 11d\u003c/strong\u003e). To more directly confirm the strong consistency of the presence of Krt8-positive cells in bone development, we further investigated the OC-like structure and spatial location of Krt8\u003csup\u003e+\u003c/sup\u003e SSCs in the long bone area of the proximal phalanx during skeletal development (\u003cstrong\u003eFig\u003c/strong\u003e\u003cstrong\u003e. 8c\u003c/strong\u003e). Even though the mature bone epiphysis has already formed, a remarkable number of Krt8\u003csup\u003e+\u003c/sup\u003e cells have emerged on the lateral side of the growth plate. Krt8\u003csup\u003e+\u003c/sup\u003e cells that reside in the metaphysis side are also Msx1 positive, while those in the epiphysis side are closely inserted into the growth plate cartilage, strongly suggesting the contribution of Krt8\u003csup\u003e+\u003c/sup\u003e SSCs to osteochondral lineage cell differentiation during the development of the ossification center in long bones (\u003cstrong\u003eFig\u003c/strong\u003e\u003cstrong\u003e. 8d\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Supplementary Fig. 11e\u003c/strong\u003e). These findings provide novel insights into the pro-regenerative role of Krt8\u003csup\u003e+\u003c/sup\u003e SSCs in bone formation during both the bone healing and the OC developmental stages, in which active and rapid osteogenesis occurs through endochondral ossification.\u003c/p\u003e\n\u003cp\u003eBased on these results above, we utilized machine learning to compare the cross-species similarity between bone regeneration and development. We first merged publicly accessible scRNA-Seq data comprising three mouse bone development datasets\u003csup\u003e38-40\u003c/sup\u003e, two human embryonic bone development datasets\u003csup\u003e13,41\u003c/sup\u003e, and one rat bone regeneration dataset\u003csup\u003e18\u003c/sup\u003e (\u003cstrong\u003eFig. 8e\u003c/strong\u003e). After homologous gene conversion, 7536 genes were selected as intersected features. In order to select more precise features for model construction, we intersected these genes with cell markers of two distinct subclusters, and finally acquired 38 features (17 genes as featured markers of Krt8\u003csup\u003e+\u003c/sup\u003e Skeletal stem cells, and 21 genes as featured markers of Has1\u003csup\u003e+\u003c/sup\u003e Migratory Fibroblasts) (\u003cstrong\u003eFig. 8f-h\u003c/strong\u003e). We used our bone regeneration data in this study for model construction (80% for training and 20% for testing), Hist Gradient Boosting (HGBoost), Logistic Regression (LR), and Support Vector Machine (SVM) performed well among several models, with the HGBoost model achieving results with the highest accuracy (\u003cstrong\u003eFig. 8i, j, Supplementary Fig. 12b\u003c/strong\u003e). We also computed the feature importance of the HGBoost model, and found that specific markers such as Has1, Krt8, and Gsn ranked at the top (\u003cstrong\u003eSupplementary Fig. 12a\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eTo predict divergent OLC subpopulations in external datasets, we applied the HGBoost model (HGBoost) to test in both the bone development data and the bone injury repair data. The prediction results showed that Krt8\u003csup\u003e+\u003c/sup\u003e SSCs are involved in both developmental and injury repair processes, whereas the proportion of Krt8\u003csup\u003e+\u003c/sup\u003e SSCs was much higher in human embryonic long bone at developmental stage. Interestingly, we found Has1\u003csup\u003e+\u003c/sup\u003e migratory fibroblasts in the bone injury repair data but not in the human embryonic long bone data (\u003cstrong\u003eFig. 8k-l\u003c/strong\u003e). We also compared data from human embryonic calvarial bone with data from rat bone injury repair, and similar results were also found (\u003cstrong\u003eSupplementary Fig. 12c-d\u003c/strong\u003e). Thus, cross-species comparisons employing machine learning revealed high resemblance of the relative Krt8\u003csup\u003e+\u003c/sup\u003e SSCs/Has1\u003csup\u003e+\u003c/sup\u003e MFs composition in bone regeneration with that in public data of developmental bone tissues at embryonic stages.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this work, we demonstrated that multiple engineered ossification center-like organoids empowered with dual-modular neuropeptides by 3D printing achieved fast\u0026nbsp;bone regeneration in a “divide-and-conquer” way. Concurrent innervation, vascularization, and ossification with a cartilage-like tissue morphology were demonstrated during the in-situ fusion and maturation of OCO, which suggested that the rapid bone regeneration may occur partly through an endochondral ossification pathway. Our scRNA-seq study unveiled that distinct osteo-lineage cells organize into cell communities that have specific molecular functions and spatial distributions. Developmentally mimicking osteo-lineage cell compositions, which include specific expansion of Krt8\u003csup\u003e+\u003c/sup\u003e SSCs concomitant with simultaneous reduction of Has1\u003csup\u003e+\u003c/sup\u003e migratory fibroblasts, were orchestrated by the pro-regenerative OC-like organoids. Furthermore, cross-species comparisons employing machine learning revealed a remarkable resemblance of relative Krt8\u003csup\u003e+\u003c/sup\u003e SSCs after OCO implantation with those in public scRNA-seq datasets of developmental bone tissues. Our findings advocate a promising approach akin to “divide-and-conquer” utilizing engineered OC-like organoids for prompt and efficient regeneration of large-sized bone defects.\u003c/p\u003e\n\u003cp\u003eCurrent clinical treatments for large bone defects often necessitate extensive wound exposure, leading to increased risks of infection. Injectable biomaterials can be delivered to the injury site in a minimally invasive manner without compromising tissue integrity or the surrounding microenvironment\u003csup\u003e42\u003c/sup\u003e. The strategies of regulating the osteogenic differentiation of stem cells by biomineralized hydrogels\u003csup\u003e43\u003c/sup\u003e or \u003cem\u003ein vitro\u003c/em\u003e pre-cultures of osteogenic differentiation\u003csup\u003e44\u003c/sup\u003e have difficulty achieving rapid bone healing and structural reconstruction of critical-sized bone defects, which usually leads to the formation of immature bone tissue. Well-known treatments for bone injury include the absorption of recombinant human BMP-2 onto collagen sponge carriers\u003csup\u003e45,46\u003c/sup\u003e or loading into microspheres\u003csup\u003e47\u003c/sup\u003e to promote the osteogenic differentiation of endogenous stem cells. The therapeutic effect is largely restricted by ectopic bone formation, osteoclast activation and soft tissue inflammation triggered by high doses and the need for additional mechanical stimulation for functional bone reconstruction\u003csup\u003e4,48\u003c/sup\u003e. As vascularization is crucial for early-stage nutrient supply and MSC recruitment, as well as the subsequent matrix remodeling, the combined delivery of vascular endothelial growth factor (VEGF) with BMPs by coupled activation of angiogenesis and osteogenesis\u003csup\u003e34,35\u003c/sup\u003e synergistically enhances bone regeneration. However, difficulties associated with their limited bioactivity (short half-life) \u003cem\u003ein vivo,\u003c/em\u003e and the risk of supraphysiological dosage-induced abnormal vascularization and heterotopic ossification\u003csup\u003e49\u003c/sup\u003e have impeded their further clinical application.\u003cem\u003e\u003csup\u003e50,51\u003c/sup\u003e\u003c/em\u003eNovel strategies for delivering growth factors (GFs) both with sustainable release and at low dosages to enhance bone regeneration are still under extensive investigation. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBone injury healing is regulated by the sequential signals that promote reinnervation, revascularization and ossification, indicating that reinnervation precedes angiogenesis and ossification during the formation of the bone-callus\u003csup\u003e22\u003c/sup\u003e. Coincidentally, neuropeptides secreted by peripheral nerves were shown to directly regulate bone growth and metabolism, and the corresponding receptors of these neuropeptides have been confirmed to be expressed in osteoblasts and osteoclasts\u003csup\u003e52,53\u003c/sup\u003e. Histological detection has also revealed the existence of neuropeptides from sensory, sympathetic and glutaminergic types in bone tissues\u0026nbsp;\u003csup\u003e54\u003c/sup\u003e. The identified role of magnesium in promoting CGRP-mediated osteogenesis has suggested its therapeutic implications in the repair of osteoporotic bone fracture or other bone diseases\u003csup\u003e55\u003c/sup\u003e. More recently, neuro-immune interactions have shown to be beneficial for non-healing tissue, and delivery of an engineered version of CGRP accelerated wound healing and promoted muscle regeneration\u003csup\u003e56\u003c/sup\u003e. Interestingly, SP was proved to be an injury-inducible factor that acts early in the wound healing process to induce the mobilization of CD29\u003csup\u003e+\u003c/sup\u003e stromal-like cells, which also occurs at uninjured sites after intravenous injection. In addition, research has shown no specific binding of biotin-conjugated SP to other tissues, such as liver, dental pulp, skin, spleen and kidney, but binding to the trabecular bone of the femur has been observed\u003csup\u003e57\u003c/sup\u003e. Higher doses of BMP-2 (e.g., greater than 150 μg/mL) even induce osteolysis\u003csup\u003e58\u003c/sup\u003e, while the lowest dose of BMP-2 that can induce bone regeneration is 2-10 μg/mL\u003csup\u003e59\u003c/sup\u003e. Our results in both 2D and 3D culture revealed that CGRP synergistically enhances osteogenic differentiation and cell survival in combination with BMP-2 at a physiological dosage. This finding is crucial for the fabrication of “morphogenetic center” units in OC-like organoid assemblies.\u003c/p\u003e\n\u003cp\u003eBone repair often progresses from the wound edge to the center\u003csup\u003e60\u003c/sup\u003e, and limited recruitment of osteo-lineage cells with osteogenic potential in critical-sized bone defects is a significant factor leading to bone nonunion\u003csup\u003e61\u003c/sup\u003e. This limitation hinders complete bone formation within the wound in a short timeframe, leading to scattered bone healing and fracture nonunion. One possible explanation for this phenomenon is the inability of stem/progenitor cells to migrate over long distances and the absence of a pro-osteogenic microenvironment\u003csup\u003e24\u003c/sup\u003e. Therefore, engineered osteogenic units are required as spatial bridges to provide a uniform adhering surface for stem cell attachment and an optimal microenvironment for endogenous new bone formation. Given the lack of in-situ osteogenic capacity upon improvement of the cell migration distance in large-sized bone defects, the development of sophisticated tissue-engineered grafts with spatiotemporal release of GFs at physiologically relevant dosages and biophysical support for spatially continuous bone regeneration is highly important. Stem cell encapsulation technology for tissue regeneration has been extensively studied and utilized as an effective way to provide a biomimetic microenvironment for cell preservation and tissue regeneration\u003csup\u003e62\u003c/sup\u003e. Modular microtissues, such as injectable osteogenic microtissues with stem cell incorporation by emulsification\u003csup\u003e63\u003c/sup\u003e, and the simultaneous delivery of osteogenic and vascularized cells that using cryogel building blocks, could potentiate the osteogenic differentiation of MSCs\u003csup\u003e64\u003c/sup\u003e. Furthermore, with improved mechanical properties, the opening of porous core-shell microtissue with demineralized bone matrix particles core modified with BMP-2 and a gelatin shell with BMSCs on the surface provides with another available method for the repair of large bone defects\u003csup\u003e65\u003c/sup\u003e. Together, these studies have shown the potential of the use of microtissue loaded with stem cells or together with GFs for bone regeneration. In the present study, “all-in-one” bioprinting enables efficient engineering of an “ossification center”-like cell construct with one shot of bone morphogenetic factors. Our results have also revealed the structural advantages of hydrogel microsphere-based repair strategies over delivery of pure bioactive GFs delivery for equalizing the migratory ability of all osteo-lineage cells to get into bone injuries (\u003cstrong\u003eFig. 3i, j\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Fig. 7d, g, h\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eOwing to their specific cell origin and genetic similarity, organoids offer low immunogenicity and are a sustainable alternative tissue source for stem cell transplantation\u003csup\u003e66\u003c/sup\u003e. Recently, the concept of bone organoids has been proposed, including woven bone organoids\u003csup\u003e67\u003c/sup\u003e, callus organoids\u003csup\u003e68\u003c/sup\u003e, trabecular bone organoids\u003csup\u003e69\u003c/sup\u003e, and humanized ossicles\u003csup\u003e70\u003c/sup\u003e etc., which can be utilized to investigate the osteogenic differentiation mechanism and promote bone tissue regeneration. Bone organoids are generated through stem/progenitor cell-driven differentiation, resulting in three-dimensional micro-bone tissues with biomimetic spatial characteristics, self-renewal, and self-assembly capabilities\u003csup\u003e71\u003c/sup\u003e. Matrigel and synthetic hydrogels have been employed to facilitate the assembly of bone organoids in order to achieve oxygen supply, metabolic gradient regulation, and intercellular connectivity\u003csup\u003e7\u003c/sup\u003e. Injectable bone microtissues formed by emulsification and incorporation of stem cells can effectively provide necessary cells for repair\u003csup\u003e63\u003c/sup\u003e. Delivery of frozen gel microspheres containing osteogenic and angiogenic cells can enhance osteogenic differentiation\u003csup\u003e64\u003c/sup\u003e. Furthermore, demineralized bone matrix particles modified with BMP-2 contribute to the repair of highly mechanically deficient bone defects\u003csup\u003e42\u003c/sup\u003e. These studies demonstrate that an “all-in-one” strategy involving stem cells and growth factors serves as an alternative for constructing bone organoids by completing the osteogenesis in advance. However, it is still challenging to replicate the developing bone for building bone organoids solely through mimicking mature bone structure and composition; moreover, there are limitations on cell survival after\u003cem\u003e\u0026nbsp;in vivo\u003c/em\u003e transplantation. There has been increasing utilization of cellular structures designed according to developmental biology principles for repairing extensive tissue defects\u003csup\u003e9\u003c/sup\u003e. Along with bone morphogenetic factors, stem cells play crucial roles in generating cartilage intermediates during the bone healing process; they facilitate the recruitment of blood vessels and osteoblasts while promoting endogenous new bone formation via the endochondral ossification pathway \u003cem\u003ein vivo\u003c/em\u003e\u003csup\u003e6,8\u003c/sup\u003e. Nevertheless, ongoing studies on these cell constructs are still in the early stages and are characterized by \u003cem\u003ein vitro\u003c/em\u003e assembly of scaffold-free elements. Further exploration of the temporal regulation of sequential innervation, vascularization, and ossification \u003cem\u003ein vivo\u003c/em\u003e following cell transplantation is needed. In this study, multiple engineered ossification center-like organoids (OCOs) were fabricated as fundamental building blocks for reconstructing bone defects via a “divide-and-conquer” strategy (\u003cstrong\u003eFig. 3i, j\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Supplementary Fig. 5b, c\u003c/strong\u003e). In addition, the in-situ fusion and maturation of OCO promote bone repair with tissue morphological changes that are characterized by innervation, vascularization and ossification (\u003cstrong\u003eFig. 4h, j, l\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eGiven the common challenges of cellular complexity and limited specificity for cell sub-clustering in scRNA-seq data analysis of regenerative tissue samples, it is imperative to explore “cell community” perspectives that could substantiate the rationality of subcluster annotations\u003csup\u003e72\u003c/sup\u003e. In tumor cells with extensive heterogeneity and cardiac cell types with complex morphological structures, systematic dissection of single-cell ecosystems helps provide a much more comprehensive understanding of cell heterogeneity with common markers and spatial mapping of their neighboring cell subpopulations\u003csup\u003e72,73\u003c/sup\u003e. Our data revealed two osteo-lineage cell subsets that express one common marker, distinct differentiation origins, distinct cellular community features, and spatial proximities with other cell subsets, suggesting their differential roles in the formation of new bone tissue. Differences in the relative proportions of cell subsets from scRNA-seq data can provide valuable insights into biological significance, which could be statistically challenging owing to the inherent noise in single-cell data and inter-sample variability. Alok K. Maity et al. suggested that leveraging cell attribution information when defining cell communities can effectively reduce noise in single-cell data, eliminating the necessity for batch correction and facilitating the retrieval of cell states for subsequent differential abundance comparisons\u003csup\u003e74\u003c/sup\u003e. Besides, cell composition shifts in relation to the specific healthy or pathological conditions when inferring differentially enriched cell communities. In our study, cell communities were well identified with cell attributions, as specific OLC subset within cell communities are predominantly representative of their corresponding regenerative approaches (\u003cstrong\u003eFig. 6\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Fig. 7\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eSkeletal stem cells (SSCs) subpopulations have been extensively investigated in the past decade with the application of scRNA-seq and lineage-tracing techniques. In postnatal skeletogenesis and impaired osteogenesis, SSCs residing in the bone marrow and growth plate that labeled with Lepr\u003csup\u003e75\u003c/sup\u003e, Gremlin 1\u003csup\u003e76\u003c/sup\u003e, or Gli1\u003csup\u003e77\u003c/sup\u003e, have been identified. Subsequently, SSCs subpopulations from long bone and bone marrow tissues have been identified by a combination of cell surface markers of CD45\u003csup\u003e−\u003c/sup\u003eTer119\u003csup\u003e−\u003c/sup\u003eAlphaV\u003csup\u003e+\u003c/sup\u003eThy\u003csup\u003e−\u003c/sup\u003e6C3\u003csup\u003e−\u003c/sup\u003eCD105\u003csup\u003e−\u003c/sup\u003eCD200\u003csup\u003e+\u003c/sup\u003e in mice\u003csup\u003e78\u003c/sup\u003e and PDPN\u003csup\u003e+\u003c/sup\u003eCD146\u003csup\u003e−\u003c/sup\u003eCD73\u003csup\u003e+\u003c/sup\u003eCD164\u003csup\u003e+\u003c/sup\u003e in human\u003csup\u003e79\u003c/sup\u003e, respectively. In addition, periosteum-derived Ctsk\u003csup\u003e+\u003c/sup\u003e SSCs that are found in both the long bones and calvarium undergo intramembranous bone formation at baseline and contribute to the endochondral ossification for cortical bone repair\u003csup\u003e80\u003c/sup\u003e. In our previous bone repair study, Msx1\u003csup\u003e+\u003c/sup\u003e SSCs were identified as sources of osteochondral progenitors and osteoprogenitors for full-thickness reconstruction during skull bone regeneration\u003csup\u003e18\u003c/sup\u003e. The above definitions for SSCs greatly differ from our identification on Krt8\u003csup\u003e+\u003c/sup\u003e skeletal stem/progenitor cells in the rat calvarium (\u003cstrong\u003eFig\u003c/strong\u003e\u003cstrong\u003e. 5f\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Fig. 6f\u003c/strong\u003e), in which the common expression of CD200 indicates a more closer similarity to the cells from mice\u003csup\u003e78\u003c/sup\u003e. Interestingly, the Krt8\u003csup\u003e+\u003c/sup\u003e SSCs identified in this study were also Msx1-positive (\u003cstrong\u003eFig\u003c/strong\u003e\u003cstrong\u003e. 8b, d\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Supplementary Fig. 10c\u003c/strong\u003e), indicating that Krt8\u003csup\u003e+\u003c/sup\u003e Msx1\u003csup\u003e+\u003c/sup\u003e SSCs better represent cells with higher potential for chondrogenic and osteogenic differentiation both during both bone injury healing and bone development.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eStrikingly, scRNA-seq data analysis of the alveolar regeneration process showed that airway and alveolar stem cells converge in a Krt8 \u003csup\u003e+\u003c/sup\u003e transitional stem cell state, which has been demonstrated to not undergo proliferative expansion and likely undergo normal homeostatic turnover\u003csup\u003e81\u003c/sup\u003e. Since our data do not include any epithelial cells, nor do the tissues acquired in single cell suspension, we discovered an epithelial feature of skeletal stem cells that forms a unique cell niche that peaks during the osteogenic phase of bone tissue repair (\u003cstrong\u003eFig. 7e, g\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Supplementary Fig. 10b, c\u003c/strong\u003e). Although the epithelial–mesenchymal transitions (EMTs) are considered as the characteristics of cell plasticity in embryonic development and cancer metastases, the importance of EMTs or the dual epithelial–mesenchymal properties of osteoprogenitors in tissue repair has become increasingly recognized\u003csup\u003e82-84\u003c/sup\u003e. These highly innovative studies suggest that the activation of transient or partial EMTs process, on the other hand, could confer phenotypic and functional plasticity on cells, which could enhance extraordinary tissue repair following damage\u003csup\u003e82\u003c/sup\u003e. This provides new insight into the mechanisms underlying tissue healing and has the potential to pave the way for novel therapeutics targeting cell plasticity to promote regeneration of currently irreversible tissue damage.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eHuman bone marrow-derived stem cells and human child phalangeal bone tissues.\u0026nbsp;\u003c/strong\u003eHuman BM aspirates were obtained in The Second Affiliated Hospital of Zhejiang University School of Medicine (Zhejiang University, China) with written informed consent from orthopedic individuals with femoral fracture (v 1.3, 2016.8.3). All samples were obtained and used according to standard guidelines approved by the Ethics Committee of the Second Affiliated Hospital of Zhejiang University School of Medicine (Ethics number: 2016-033). Human child phalangeal bone tissues were obtained in Children's Hospital of Zhejiang University School of Medicine (Zhejiang University, China) with written informed consent from individuals after polydactyly resection surgery. Samples were obtained and used according to standard guidelines approved by the ethics committee of Children's Hospital of Zhejiang University (Ethics number: 2020-IRB-007). The bone marrow samples were processed for the following experiments, after the filters used to trap bone spicules and cell aggregates, and were carefully and aseptically washed with cold PBS several times. BMSCs were isolated using a Percoll gradient (Gibco) and cultured in Low-Glucose Dulbecco’s Modified Eagle’s Medium (DMEM; Gibco) containing 10% fetal bovine serum (FBS; Gibco), 1% penicillin/streptomycin (P/S) and incubated at 37 °C, 5% CO2. The culture medium was replaced every other day until about 90% confluence was achieved. Passage 3-5 of the BMSCs were utilized for the \u003cem\u003ein vitro\u003c/em\u003e experiments either in 2D or 3D culture model in this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRat bone marrow-derived stem cells isolation and culture.\u0026nbsp;\u003c/strong\u003eThe bone marrow from the Sprague Dawley rat (2 weeks old) femora and tibia was flushed in freshly prepared culture medium with L-DMEM medium (DMEM; Gibco) containing 10% fetal bovine serum (FBS; Gibco), and 1% penicillin/streptomycin (P/S; Thermo Fisher Scientific), incubated at 37°C in a humidified atmosphere containing 5% CO\u003csub\u003e2\u003c/sub\u003e. After centrifuged, the bone marrow was resuspended in growth medium and then seeded in 2D or 3D culture system, and samples were analyzed in certain time points. The culture medium was refreshed every 2 days until about 90% confluence was achieved. Passage 3-6 of the BMSCs was utilized for the \u003cem\u003ein vitro\u003c/em\u003e experiments in this study.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCell viability assay.\u0026nbsp;\u003c/strong\u003e1 X 10\u003csup\u003e5\u003c/sup\u003e cells/well were suspended in DMEM media supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. And the cells and media were incubated for 1, 3, 5 days in 2D culture or 1, 4, 7 days in 3D microsphere-based culture, and the CCK-8 kit (Dojindo, Japan) was chosen to determine cell viability.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOsteogenic differentiation assays.\u0026nbsp;\u003c/strong\u003eFor osteogenic differentiation, the cells were plated at confluence in osteogenic medium containing α-DMEM with 10% FBS supplemented with 0.1μM dexamethasone, 0.2 mM L-ascorbic acid, and 10 mM glycerol 2-phosphate disodium salt hydrate (Sigma, St. Louis, MO). For comparisons among different neuropeptides, distinct neuropeptides VIP (10\u003csup\u003e-7\u003c/sup\u003e M), NYP (10\u003csup\u003e-8\u003c/sup\u003e M), and CGRP (10\u003csup\u003e-8\u003c/sup\u003e M) were supplemented in the osteogenic medium. For comparisons among different treatments, CGRP (10\u003csup\u003e-8\u003c/sup\u003e M), BMP-2 (0.5μg/mL), CGRP (10\u003csup\u003e-8\u003c/sup\u003e M) combined with BMP-2 (0.5μg/mL) (CGRP + BMP-2), were supplemented in the osteogenic medium. The medium was changed every 2 days during 1 or 2 weeks.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRT-qPCR of osteogenic genes.\u0026nbsp;\u003c/strong\u003eOsteogenic gene expressions of BMSCs were measured on day 7 and 14 by a Real-time quantitative reverse transcription-polymerase chain reaction (RT-qPCR) system, and four groups were set as above-mentioned. Angiogenic gene VEGF expression of BMSCs was measured by RT-qPCR at day 3 with the treatment of SP (10\u003csup\u003e-7\u0026nbsp;\u003c/sup\u003eM). RNA was extracted from cells and reverse transcribed into complementary DNA (cDNA) using Trizol reagent and PrimeScript RT reagent kit (Takara, Tokyo, Japan) according to manufacturer's instructions. Then diluted cDNA was mixed with SYBR Premix Ex TaqTM (Takara, Tokyo, Japan), forward and reverse primers and RNase free water to perform RT-qPCR. Osteogenic differentiation markers including ALPL and RUNX2 were evaluated, with GAPDH used as housekeeping gene. Relative expression level for each gene (fold change) to that of blank control was calculated. Primer sequences used in this study were listed in \u003cstrong\u003eSupplementary Table S1\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAlkaline Phosphatase (ALP) activity.\u0026nbsp;\u003c/strong\u003eBMSCs were seeded in 24-well plate at a density of 2×10\u003csup\u003e4\u003c/sup\u003e cells/well, and 24 h post-seeding the medium was changed to experimental medium. Cells were briefly washed with PBS and fixed for 10 min with 2% paraformaldehyde in PBS (Sigma). Cells were washed twice with PBS and incubated for 20 min at room temperature with 50 mg/mL Naphthol AS-MX phosphate, 0.5% N, N-Dimethylformamide, and 0.6 mg/mL Fast Red Violet LB in 0.1 M Tris-HCl, pH 8.9.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAlizarin Red Staining.\u0026nbsp;\u003c/strong\u003eCells were washed with cold PBS and fixed with 70% ethanol for 15 min on ice. Cells were then washed with distilled water and stained with 2% alizarin red solution for 5 min. Cells were subsequently washed thoroughly with distilled water and air dried before microscopic visualization.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSynthesis of GelMA.\u0026nbsp;\u003c/strong\u003eGelMA (Methylacrylated Gelatin) was synthesized according to a previous study\u003csup\u003e18\u003c/sup\u003e. Briefly, type A gelatin (Sigma-Aldrich) was dissolved in PBS at 50°C and stirred to make a 10% w/v homogeneous solution. And a 0.1 mL methacrylic anhydride (MA) (Sigma-Aldrich) per gram of gelatin was added to homogeneous gelation solution at a rate of 0.5 mL per minute with a continuous stirring. The mixed solution was allowed to react at 50°C for 3 h with stirring. Then, the GelMA solution was poured into 8–14 kDa cutoff dialysis tubing (VWR Scientific USA) and dialyzed against deionized water for 6 days at 50°C to remove untreated MA and other byproducts. The deionized water was replaced every 1-2 days. The resulted GelMA solution was frozen overnight or longer at -80°C and lyophilized and stored at -20°C for further use.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSynthesis of NB and HA-NB.\u0026nbsp;\u003c/strong\u003eMethyl 4-(4-(hydroxymethyl)-2-methoxy-5-nitrophenoxy) butanoate (mNB) was synthesized based on a previous research\u003csup\u003e85\u003c/sup\u003e. mNB (0.5g, 1.8mmol) and ethylenediamine (1.1mL, 2 mmol, Sigma Aldrich) were dissolved in methanol. Then, the mixture was refluxed overnight until the starting individual components were not detectable by thin layer chromatography (TLC). The solvent was evaporated under vacuum after the reaction was complete. The crude precipitate was then dissolved in methanol and re-precipitated three times using ethyl acetate. The filter cake was dried at 30°C for 12h under vacuum until NB appeared as a light-yellow powder (0.4 g, 1.2 mmol, 66.7%). HA-NB was synthesized according to a published report\u003csup\u003e85\u003c/sup\u003e. Briefly, HA (408 mg, 1 mmol of disaccharide unit, Dongyuan Biotech, Zhenjiang) was dissolved in 50 mL deionized water at room temperature and NB (224 mg, 0.69 mmol) was added followed by HOBt (153 mg, 1 mmol, Sigma-Aldrich). The pH of the mixture was adjusted to pH 4.5, and the 1-(3-Dimethylaminopropyl)-3-ethylcarbodimide hydrochloride (EDC) (200 mg, 1.04, Sigma Aldrich) was added to the mixed solution and then stirred at room temperature for 48h. Then, the solution was loaded into dialysis tubing (Molecular Weight (MW) cutoff 3500, Spectrum®) and dialyzed against diluted HCl (pH 3.5) containing 0.1M NaCl for two days, then dialyzed against deionized water for 2 days. The solution was lyophilized and HA-NB was obtained in powder form.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSynthesis and preparation of the photo-initiator.\u0026nbsp;\u003c/strong\u003eThe photo-initiator lithium phenyl-2,4,6- trimethylbenzoylphosphinate (LAP) was synthesized based on our previous research\u003csup\u003e86,87\u003c/sup\u003e. In brief, Dimethyl phenylphosphonite (Ourchem) was reacted with 2,4,6-trimethylbenzoyl chloride (Sigma-Aldrich) via a Michaelis–Arbuzov reaction. At room temperature and under argon gas, 3.2 g (0.018 mol) of 2,4,6-trimethylbenzoyl chloride was added dropwise to an equimolar amount of continuously stirred dimethyl phenylphosphonite (3.0 g). The reaction mixture was stirred for 18h whereupon a four-fold excess of lithium bromide (Aladdin, 6.1g) in 100mL of 2-butanone (Sinopharm Chemical Reagent) was added to the reaction mixture from the previous step, which was then heated to 50°C, a solid precipitate had formed after 10 minutes. Then, the mixture was cooled to ambient temperature and allowed to rest for 4h, and then filtered. The filtrate was washed and filtered 3 times with 2-butanone to remove unreacted lithium bromide, and excess solvent was removed by vacuum.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFabrication of GelMA/HA-NB microcarriers.\u0026nbsp;\u003c/strong\u003eFor precursors of GelMA/HA-NB hydrogels, the freeze-dried GelMA foams and HA-NB foams were dissolved in PBS solution at 40 °C to a final concentration of 5% GelMA and 1.25% HA-NB. For precursors of GelMA/LAP hydrogels, the freeze-dried GelMA foams were dissolved in PBS solution at 40 °C and then added to the photo-initiator LAP to a final concentration of 5% GelMA and 0.1% LAP. In this study, GelMA/HA-NB microcarriers were fabricated by DLP-based 3D bioprinting method. In biocompatibility comparisons, the GelMA/HA-NB microcarriers (2×10\u003csup\u003e6\u003c/sup\u003e cells/mL of the hydrogel) with diameter of 100μm, 200μm, 400μm that encapsulating BMSCs were fabricated. In cell-loaded experiments, the GelMA/HA-NB microcarriers with a diameter of 400μm that encapsulating BMSCs were fabricated (2×10\u003csup\u003e6\u003c/sup\u003e cells/mL of the hydrogel) with no proteins, single BMP-2, or BMP-2 combined with CGRP.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLive/dead staining assay.\u003c/strong\u003e 1\u0026nbsp;×\u0026nbsp;10\u003csup\u003e5\u003c/sup\u003e cells were added to per GelMA/HA-NB/LAP to evaluate the cytotoxicity of GelMA/HA microspheres. And the cells were cultured in the microspheres in DMEM media supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin at 37 ◦C and incubated at 5% CO2 for 1, 4, 7, and 14 days. Images were captured by a laser scanning confocal microscope (OLYMPUS IX83-FV3000, Japan) for live/dead assay.\u0026nbsp;After 1, 4 and 7 days of culture, 100μL microcarriers with diameter of 200μm and 400μm were collected and digested with type I collagenase and cells were counted. Growth factors used for Live/dead staining assay including\u0026nbsp;BMP-2 (1μg/mL) and CGRP (10\u003csup\u003e-7\u003c/sup\u003e M).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eELISA tests of growth factor release.\u0026nbsp;\u003c/strong\u003eThe microcarriers containing BMP-2 (5μg/ml) was mixed with GelMA hydrogel containing VEGF (5μg/mL) in equal proportion, and the mixture of 200μL was absorbed and cross-linked to cure under UV light for 1 min. The mixture was then immersed in PBS and placed in an incubator at 37°C. PBS was absorbed at regular intervals after soaking, and the release of VEGF and BMP-2 from hydrogel compound structure was detected by Human BMP-2 ELISA kit (NeoBioscience)\u0026nbsp;and Human VEGF ELISA kit (NeoBioscience), respectively.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAnimal and surgical procedures.\u0026nbsp;\u003c/strong\u003eSD rats (~ 250 g, 8~10 weeks old, Male) were used in this study. All surgical procedures were performed under 4% isoflurane anesthesia. Surgical sites were sterilized using iodine solution after hair removal using a clipper.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn subcutaneous transplantation model, an incision with the length of 1.5 cm was made in the mediodorsal skin and a lateral subcutaneous pocket of SD rats was prepared. For biocompatibility assessment, GelMA and GelMA/HA-NB hydrogel (cylindrical flake: 8 mm in diameter and 1mm in thickness) were implanted under sterile conditions. At 1, 2, and 4 weeks, the rats were sacrificed and the samples were processed for histological analysis. For ectopic ossification experiments, the integrated delivery systems were divided into three groups: (1) GelMA hydrogel + GelMA/HA-NB Ms; (2) GelMA hydrogel (SP) + GelMA/HA-NB Ms; (3) GelMA hydrogel (SP) + GelMA/HA-NB Ms (BMSCs), and mixed according to the volume ratio of 1:2. The mixture of 200μL was absorbed and cross-linked and cured by UV light for 1 min, and then implanted subcutaneously into rats. At 3 weeks, the rats were sacrificed and the samples were processed for histological analysis.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn calvarial bone defect model, a cranial defect with a diameter of 5 mm was created in the center of the calvarium using dental trephine. The rats were divided into four groups: (1) Defect control (Defect), (2) Microsphere in hydrogel (Vehicle), (3) Mixed MSCs and Neurotrophins (Hybrid), (3) Ossification center organoid (OCO). Growth factors used for bone injury repair \u003cem\u003ein vivo\u003c/em\u003e including\u0026nbsp;BMP-2 (5μg/mL), CGRP (10\u003csup\u003e-7\u003c/sup\u003e M), SP (10\u003csup\u003e-7\u0026nbsp;\u003c/sup\u003eM). At 2 weeks after surgery, the rats were sacrificed and the regenerative tissues in the defect area were harvested and digested for single-cell collection. At 4 weeks and 8 weeks after surgery, the rats were sacrificed and the calvarium were harvested for further histological assessments, respectively. The rats used in this study were fed in separated cages in a temperature, humidity-controlled (~25°C, 50–80%) and 12 h light/dark cycle room. All animals were treated according to standard guidelines approved by the Zhejiang University Ethics Committee (Ethical NO. ZJU20210114).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMicro CT analysis and paraffin embedding.\u0026nbsp;\u003c/strong\u003eAfter the defect surgery for 4 and 8 weeks, rats (n = 5, per group, per time point) were euthanatized and the calvarial specimens were harvested and fixed overnight with 4% paraformaldehyde at 4 °C. The fixed samples were scanned using micro-CT (U-CT-XUHR, MILabs) at 4μm resolution. The three-dimensional (3D) structures of calvarium were reconstructed through MILabs-Rec interface, and analyzed by IMALYTICS Preclinical 2.1. software of the micro-CT. A cylinder space representing the region of interest (ROI) was designated to evaluate both bone and tissue volume for calculation of bone volume/tissue volume (BV/TV) and Bone density (BMD). Samples were then decalcified with 0.5 M EDTA for 8 weeks and then subjected to paraffin embedding.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHistological analysis.\u003c/strong\u003e The harvested tissue specimens were fixed and then decalcified in 10% ethylenediamine tetra-acetic acid (EDTA, pH = 7.4) solution for two months\u0026nbsp;at room temperature, and then dehydrated through graded alcohol series and embedded in paraffin. Sections of the central segment were cut into 10μm thick slices using a rotary microtome (Leica, Hamburg, Germany). Hematoxylin and Eosin (H\u0026amp;E) staining, Safranin-O staining, and Masson’s Trichrome staining were performed on paraffin sections according to standard protocols, and observed using bright-field microscopy.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImmunofluorescent staining.\u0026nbsp;\u003c/strong\u003eCells cultured either on dishes or in microspheres were fixed with 4% paraformaldehyde (PFA) for 20 min and then permeabilized with 0.03% Triton X-100 for 10 min at room temperature. After washing with PBS for 3 times, the samples were incubated with blocking solution (1% bovine serum albumin) for 30 min at room temperature to prevent nonspecific binding. The primary antibodies were diluted 200 or 500-fold with blocking solution and added to the cell cultures at 4 °C overnight. Tissues were fixed with 4% PFA and the paraffin-embedded samples were cut into 10-μm-thick sections. The tissue samples were deparaffinized and the antigens were activated by heating the slides in 10 mM citrate buffer (pH 6.0) at 65 °C overnight. After treating the sections with 0.3% H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e in MeOH, the samples were incubated with blocking solution (5% bovine serum albumin and 0.03% Triton X-100 in PBS).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe samples were treated with primary antibodies at 4 °C overnight, including Ki67 (1:250 dilution, ab16667, Abcam, USA), γh2AX (1:250 dilution, ab22551, Abcam, USA), RUNX2 (1:200 dilution, ab76956, Abcam, UK), Collagen I (1:100 dilution, ab260043, Abcam, UK), OCN (1:50 dilution, MAB1419, R\u0026amp;D Systems, USA), NGFR (1:50 dilution, NBP2-67296, NOVUS, USA), CGRP (1:100 dilution, ab81887, Abcam, UK), beta III Tubulin (1:500 dilution, ab18207, Abcam, UK), Collagen X (1:100 dilution, 14-9771-82, Invitrogen, USA), CD31 (1:100 dilution, ab222783, Abcam, UK), CD200 (1:100 dilution, AF2724, R\u0026amp;D Systems, USA), Thy1 (1:200 dilution, ab181469, Abcam, UK), PDGFR\u0026nbsp;alpha (1:250 dilution, ab203491, Abcam, UK), KRT8 (1:100 dilution, ab53280, Abcam, UK), HAS1 (1:250 dilution, PA5-95599, Invitrogen, USA), MSX1 (1:100 dilution, ab93287, Abcam, UK). After incubation with primary antibody, samples were then incubated with Alexa Fluor® secondary antibodies (G-Rabbit Alexa Fluor® 488, A11008; Goat anti mouse Alexa Fluor® 488, A11001; G-Rabbit Alexa Fluor® 546, A21430-f; Donkey-Mouse Alexa Fluor® 405, ab175658) (diluted 1:500) for 1 h at room temperature. After incubation, the nuclei were stained with 0.1μg/mL\u0026nbsp;DAPI (Invitrogen, USA). After staining, the samples were observed using a confocal microscope (OLYMPUS IX83-FV1000 and FV3000-OSR, Japan).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImmunohistochemistry staining.\u0026nbsp;\u003c/strong\u003eFor immune-histochemical staining, sections were prepared, and followed by antigen retrieval by heating the slides in 10 mM citrate buffer (pH 6.0) at 65 °C overnight, inactivation of endogenous peroxidase by hydrogen peroxide with 0.3% H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e in MeOH, 5% BSA blocking solution, and was incubated with primary antibodies against OCN (1:50 dilution, MAB1419, R\u0026amp;D Systems, USA), at 4 °C overnight. Then sections were incubated with anti-Mouse secondary antibody conjugated with 1:1000 HRP (Beyotime Institute of Biotechnology). The stained specimens were photographed digitally and viewed under the Digital Slide Scanners.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSingle-cell RNA Sequencing and data analysis.\u0026nbsp;\u003c/strong\u003eAt 2 weeks after surgery, the rats (male, n=6 biological rats, per group) were sacrificed.\u0026nbsp;The regenerated tissues in calvarial defect area were minced with razor blades and washed several times by 4 °C PBS, and then digested by enzyme mixture (type I collagenase 0.1% and type II collagenase 0.1%, incubated at 37°C for 40 min) before filtered through a 70μm nylon mesh to obtain single cell suspension. Subsequently, Single-cell suspensions (2×10\u003csup\u003e5\u003c/sup\u003e cells/mL) with PBS (HyClone) were loaded onto microwell chip using the Singleron Matrix® Single Cell Processing System. Next, the scRNA-seq libraries were constructed according to the protocol of the GEXSCOPE single-cell RNA library kits (Singleron). Individual libraries were diluted to 4 nM and pooled for sequencing. Finally, the pools were sequenced on the Illumina NovaSeq 6000 system with 150 bp paired-end reads.\u003c/p\u003e\n\u003cp\u003eFor single-cell clustering and annotation,after dimensionality reduction, we performed clustering using FindClusters function offered by Seurat. Genes expressed specifically in each clusters were calculated by Seurat FindAllMarkers function. To identify the biological cell type of each cluster, we performed SingleR\u003csup\u003e88\u003c/sup\u003e analysis, combined with conventional markers of some known cell types. For differentiation trajectory analysis, to map differentiation in bone regeneration , we performed pseudotime analysis with R package Monocle\u003csup\u003e89\u003c/sup\u003e (version 2.20.0). Specifically, we computed trajectory of osteo-lineage cells and compared different states of cells using BEAM function provided by Monocle. RNA velocity was performed with scvelo python package (version 0.1.25).For gene functional annotation analysis,GO enrichment analysis was performed for markers of single-cell clusters using clusterProfiler\u003csup\u003e90\u003c/sup\u003e package. The enriched GO terms were filtered by setting pvalueCutoff to 0.01.For single-cell regulatory network analysis,the analysis of single-cell gene regulatory network was performed using the SCENIC\u003csup\u003e91\u003c/sup\u003e package followed by the standard pipeline. Dot plot shows the cell-type specific regulons with top Regulon Specificity Score (RSS)\u0026nbsp;\u003csup\u003e92\u003c/sup\u003e and their average expression (Z) in the cell subtype.Homologous gene conversion among human, rat and mouse genes was performed by R package babelgene (version 22.9). Machine learning and model optimization in this study was performed using sklearn python package (version 1.0.2).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical analysis.\u0026nbsp;\u003c/strong\u003eValues are expressed as mean ± SD or mean ± SEM unless otherwise indicated in the figure legends. The significance between two groups was analyzed using two-tailed Student’s t-tests. For multiple comparisons, one-way analysis of variance (ANOVA) with Tukey’s post hoc test was used. Statistical analysis was performed using the Graphpad software. P \u0026lt; 0.05 was considered to be significant; *P \u0026lt; 0.05, **P \u0026lt; 0.01, ***P \u0026lt; 0.001, ****P \u0026lt; 0.0001.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe single cell RNA-sequencing data generated in this study have been deposited in the Genome Sequence Archive (GSA) database with accession number CRA018756, which is now publicly available can be accessed from the following link [https://bigd.big.ac.cn/gsa/browse/CRA018756]. All other relevant data supporting the key findings of this study are available within the article and its Supplementary Information files or from the corresponding author upon reasonable request.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the National Key Research and Development Program of China (2023YFB3183000), and NSFC grants (NO. T2121004, 92268203, 82301016). The authors thank Weiliang Shen (from The Second Affiliated Hospital of Zhejiang University) for providing the human bone marrow samples from trauma subjects in traffic accidents or discarded tissue during the operation of fractures. We thank Wei Yin and Junli Xuan (Core Facilities, Zhejiang University School of Medicine) for their assistance with Confocal laser scanning microscope.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eXianzhu Zhang and Hongwei Ouyang designed the project, performed experiments and wrote the whole manuscript; Xianzhu Zhang, Wei Jiang, Xinyu Wu, Xiaohui Zou, and Hongwei Ouyang helped revised the manuscript; Xianzhu Zhang, Xinyu Wu, and Yi Zhang performed the \u003cem\u003ein vitro\u003c/em\u003e cell culture and osteogenic induction assays; Xinyu Wu and Chang Xie, Renjie Liang, and Liying Li completed the hydrogel and microsphere preparation; Xianzhu Zhang, Xinyu Wu, Tao Zhang, Wei Sun, Jingchun Ye, and Youzhi Cai performed the animal experiments and the histological analysis;\u0026nbsp;Chang Xie, Yi Zhang,\u0026nbsp;and Yuqing Gu, and Liying Li\u0026nbsp;helped with the\u0026nbsp;immunofluorescent and immunohistochemical staining of regenerative tissue sections; Xianzhu Zhang, Wei Jiang, and Zihao Hu performed single-cell sequencing and data analysis, and machine learning-based data analysis; Xiaozhao Wang, Wei Wei, and Yi Hong helped with the material preparation and DLP printing; Shufang Zhang,\u0026nbsp;Xiaohui Zou,\u0026nbsp;Yihe Hu\u0026nbsp;and\u0026nbsp;Hongwei Ouyang\u0026nbsp;helped with the results analysis and discussion.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003cstrong\u003e\u003cbr\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e"},{"header":" References","content":"\u003col\u003e\n \u003cli\u003eZura, R.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Epidemiology of Fracture Nonunion in 18 Human Bones. \u003cem\u003eJAMA Surg\u003c/em\u003e \u003cstrong\u003e151\u003c/strong\u003e, e162775, doi:10.1001/jamasurg.2016.2775 (2016).\u003c/li\u003e\n \u003cli\u003eWang, L.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Development of a centrally vascularized tissue engineering bone graft with the unique core-shell composite structure for large femoral bone defect treatment. \u003cem\u003eBiomaterials\u003c/em\u003e \u003cstrong\u003e175\u003c/strong\u003e, 44-60, doi:10.1016/j.biomaterials.2018.05.017 (2018).\u003c/li\u003e\n \u003cli\u003eChen, S. X.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Biomaterials with structural hierarchy and controlled 3D nanotopography guide endogenous bone regeneration. \u003cem\u003eScience Advances\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, doi:ARTN eabg3089 /sciadv.abg3089 (2021).\u003c/li\u003e\n \u003cli\u003eHo-Shui-Ling, A.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Bone regeneration strategies: Engineered scaffolds, bioactive molecules and stem cells current stage and future perspectives. \u003cem\u003eBiomaterials\u003c/em\u003e \u003cstrong\u003e180\u003c/strong\u003e, 143-162, doi:10.1016/j.biomaterials.2018.07.017 (2018).\u003c/li\u003e\n \u003cli\u003eBianco, P.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e The meaning, the sense and the significance: translating the science of mesenchymal stem cells into medicine. \u003cem\u003eNature Medicine\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 35-42, doi:10.1038/nm.3028 (2013).\u003c/li\u003e\n \u003cli\u003eMcDermott, A. M.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Recapitulating bone development through engineered mesenchymal condensations and mechanical cues for tissue regeneration. \u003cem\u003eScience Translational Medicine\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, doi:ARTN eaav7756/scitranslmed.aav7756 (2019).\u003c/li\u003e\n \u003cli\u003eDai, K.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Construction of developmentally inspired periosteum-like tissue for bone regeneration. \u003cem\u003eBone Research\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, doi:ARTN 1 /s41413-021-00166-w (2022).\u003c/li\u003e\n \u003cli\u003eHerberg, S.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Combinatorial morphogenetic and mechanical cues to mimic bone development for defect repair. \u003cem\u003eScience Advances\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, doi:ARTN eaax2476 /sciadv.aax2476 (2019).\u003c/li\u003e\n \u003cli\u003eNilsson Hall, G.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Developmentally Engineered Callus Organoid Bioassemblies Exhibit Predictive In Vivo Long Bone Healing. \u003cem\u003eAdv Sci (Weinh)\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 1902295, doi:10.1002/advs.201902295 (2020).\u003c/li\u003e\n \u003cli\u003eHasani-Sadrabadi, M. M.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e An engineered cell-laden adhesive hydrogel promotes craniofacial bone tissue regeneration in rats. \u003cem\u003eSci Transl Med\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, doi:10.1126/scitranslmed.aay6853 (2020).\u003c/li\u003e\n \u003cli\u003ePetersen, A.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e A biomaterial with a channel-like pore architecture induces endochondral healing of bone defects. \u003cem\u003eNature Communications\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, doi:ARTN 4430 /s41467-018-06504-7 (2018).\u003c/li\u003e\n \u003cli\u003eMoroni, L.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Biofabrication strategies for 3D in vitro models and regenerative medicine. \u003cem\u003eNature Reviews Materials\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 21-37, doi:10.1038/s41578-018-0006-y (2018).\u003c/li\u003e\n \u003cli\u003eXie, C.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e High-efficient engineering of osteo-callus organoids for rapid bone regeneration within one month. \u003cem\u003eBiomaterials\u003c/em\u003e \u003cstrong\u003e288\u003c/strong\u003e, doi:ARTN 121741 /j.biomaterials.2022.121741 (2022).\u003c/li\u003e\n \u003cli\u003eHe, Q. L.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e \u0026quot;All-in-One\u0026quot; Gel System for Whole Procedure of Stem-Cell Amplification and Tissue Engineering. \u003cem\u003eSmall\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, doi:ARTN 1906539 /smll.201906539 (2020).\u003c/li\u003e\n \u003cli\u003eSubbiah, R.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Osteogenic/Angiogenic Dual Growth Factor Delivery Microcapsules for Regeneration of Vascularized Bone Tissue. \u003cem\u003eAdvanced Healthcare Materials\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 1982-1992, doi:10.1002/adhm.201500341 (2015).\u003c/li\u003e\n \u003cli\u003eKim, D. S.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Advanced PLGA hybrid scaffold with a bioactive PDRN/BMP2 nanocomplex for angiogenesis and bone regeneration using human fetal MSCs. \u003cem\u003eSci Adv\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, eabj1083, doi:10.1126/sciadv.abj1083 (2021).\u003c/li\u003e\n \u003cli\u003eFreeman, F. E.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 3D bioprinting spatiotemporally defined patterns of growth factors to tightly control tissue regeneration. \u003cem\u003eSci Adv\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, eabb5093, doi:10.1126/sciadv.abb5093 (2020).\u003c/li\u003e\n \u003cli\u003eZhang, X.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Msx1(+) stem cells recruited by bioactive tissue engineering graft for bone regeneration. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 5211, doi:10.1038/s41467-022-32868-y (2022).\u003c/li\u003e\n \u003cli\u003eTomlinson, R. E.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e NGF-TrkA Signaling by Sensory Nerves Coordinates the Vascularization and Ossification of Developing Endochondral Bone. \u003cem\u003eCell Reports\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 2723-2735, doi:10.1016/j.celrep.2016.08.002 (2016).\u003c/li\u003e\n \u003cli\u003eLee, S. Y.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e NGF-TrkA signaling dictates neural ingrowth and aberrant osteochondral differentiation after soft tissue trauma. \u003cem\u003eNature Communications\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, doi:ARTN 4939 /s41467-021-25143-z (2021).\u003c/li\u003e\n \u003cli\u003eLi, Z.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Fracture repair requires TrkA signaling by skeletal sensory nerves. \u003cem\u003eJournal of Clinical Investigation\u003c/em\u003e \u003cstrong\u003e129\u003c/strong\u003e, 5137-5150, doi:10.1172/Jci128428 (2019).\u003c/li\u003e\n \u003cli\u003eLi, Z.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Fracture repair requires TrkA signaling by skeletal sensory nerves. \u003cem\u003eJ Clin Invest\u003c/em\u003e \u003cstrong\u003e129\u003c/strong\u003e, 5137-5150, doi:10.1172/JCI128428 (2019).\u003c/li\u003e\n \u003cli\u003eMeyers, C. A.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e A Neurotrophic Mechanism Directs Sensory Nerve Transit in Cranial Bone. \u003cem\u003eCell Rep\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 107696, doi:10.1016/j.celrep.2020.107696 (2020).\u003c/li\u003e\n \u003cli\u003eXu, J.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e NGF-p75 signaling coordinates skeletal cell migration during bone repair. \u003cem\u003eSci Adv\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, eabl5716, doi:10.1126/sciadv.abl5716 (2022).\u003c/li\u003e\n \u003cli\u003eLi, W. T.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Bioprinted Constructs that Mimic the Ossification Center Microenvironment for Targeted Innervation in Bone Regeneration. \u003cem\u003eAdv Funct Mater\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, doi:ARTN 2109871 /adfm.202109871 (2022).\u003c/li\u003e\n \u003cli\u003eZhang, Y. F.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Implant-derived magnesium induces local neuronal production of CGRP to improve bone-fracture healing in rats. \u003cem\u003eNature Medicine\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 1160-1169, doi:10.1038/nm.4162 (2016).\u003c/li\u003e\n \u003cli\u003eLu, Y. Z.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e CGRP sensory neurons promote tissue healing via neutrophils and macrophages. \u003cem\u003eNature\u003c/em\u003e, doi:10.1038/s41586-024-07237-y (2024).\u003c/li\u003e\n \u003cli\u003eHong, H. S.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e A new role of substance P as an injury-inducible messenger for mobilization of CD29 stromal-like cells. \u003cem\u003eNature Medicine\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 425-435, doi:10.1038/nm.1909 (2009).\u003c/li\u003e\n \u003cli\u003eZhang, Y.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Neuronal Induction of Bone-Fat Imbalance through Osteocyte Neuropeptide Y. \u003cem\u003eAdv Sci\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, doi:ARTN 2100808/advs.202100808 (2021).\u003c/li\u003e\n \u003cli\u003eShi, L.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Vasoactive Intestinal Peptide Stimulates Bone Marrow-Mesenchymal Stem Cells Osteogenesis Differentiation by Activating Wnt/\u0026beta;-Catenin Signaling Pathway and Promotes Rat Skull Defect Repair. \u003cem\u003eStem Cells and Development\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 655-666, doi:10.1089/scd.2019.0148 (2020).\u003c/li\u003e\n \u003cli\u003eYu, H. P.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Application of the neuropeptide NPVF to enhance angiogenesis and osteogenesis in bone regeneration. \u003cem\u003eCommun Biol\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, doi:ARTN 197 /s42003-023-04567-x (2023).\u003c/li\u003e\n \u003cli\u003eWang, L. P.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Mice lacking substance P have normal bone modeling but diminished bone formation, increased resorption, and accelerated osteopenia with aging. \u003cem\u003eBone\u003c/em\u003e \u003cstrong\u003e144\u003c/strong\u003e, doi:ARTN 115806 /j.bone.2020.115806 (2021).\u003c/li\u003e\n \u003cli\u003eLiu, L. J., Dana, R. \u0026amp; Yin, J. Sensory neurons directly promote angiogenesis in response to inflammation via substance P signaling. \u003cem\u003eFaseb J\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 6229-6243, doi:10.1096/fj.201903236R (2020).\u003c/li\u003e\n \u003cli\u003eTang, W.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Enhancement and orchestration of osteogenesis and angiogenesis by a dual-modular design of growth factors delivery scaffolds and 26SCS decoration (vol 232, 119645, 2020). \u003cem\u003eBiomaterials\u003c/em\u003e \u003cstrong\u003e268\u003c/strong\u003e, doi:ARTN 120616 /j.biomaterials.2020.120616 (2021).\u003c/li\u003e\n \u003cli\u003eLee, S. S.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Sequential growth factor releasing double cryogel system for enhanced bone regeneration. \u003cem\u003eBiomaterials\u003c/em\u003e \u003cstrong\u003e257\u003c/strong\u003e, doi:ARTN 120223 /j.biomaterials.2020.120223 (2020).\u003c/li\u003e\n \u003cli\u003eFreeman, F. E.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 3D bioprinting spatiotemporally defined patterns of growth factors to tightly control tissue regeneration. \u003cem\u003eScience Advances\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, doi:ARTN eabb5093 /sciadv.abb5093 (2020).\u003c/li\u003e\n \u003cli\u003eWang, Q., Liu, K. L., Yang, L., Wang, H. H. \u0026amp; Yang, J. BoneClear: whole-tissue immunolabeling of the intact mouse bones for 3D imaging of neural anatomy and pathology. \u003cem\u003eCell Research\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 870-872, doi:10.1038/s41422-019-0217-9 (2019).\u003c/li\u003e\n \u003cli\u003eKelly, N. H., Huynh, N. P. T. \u0026amp; Guilak, F. Single cell RNA-sequencing reveals cellular heterogeneity and trajectories of lineage specification during murine embryonic limb development. \u003cem\u003eMatrix Biol\u003c/em\u003e \u003cstrong\u003e89\u003c/strong\u003e, 1-10, doi:10.1016/j.matbio.2019.12.004 (2020).\u003c/li\u003e\n \u003cli\u003eLi, J.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Systematic Reconstruction of Molecular Cascades Regulating GP Development Using Single-Cell RNA-Seq. \u003cem\u003eCell Rep\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 1467-1480, doi:10.1016/j.celrep.2016.04.043 (2016).\u003c/li\u003e\n \u003cli\u003eSivaraj, K. K.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Regional specialization and fate specification of bone stromal cells in skeletal development. \u003cem\u003eCell Rep\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 109352, doi:10.1016/j.celrep.2021.109352 (2021).\u003c/li\u003e\n \u003cli\u003eHe, J.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Dissecting human embryonic skeletal stem cell ontogeny by single-cell transcriptomic and functional analyses. \u003cem\u003eCell Res\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 742-757, doi:10.1038/s41422-021-00467-z (2021).\u003c/li\u003e\n \u003cli\u003eLuo, C.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Biomimetic open porous structured core-shell microtissue with enhanced mechanical properties for bottom-up bone tissue engineering. \u003cem\u003eTheranostics\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 4663-4677, doi:10.7150/thno.34464 (2019).\u003c/li\u003e\n \u003cli\u003eCui, Z. K.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Microporous methacrylated glycol chitosan-montmorillonite nanocomposite hydrogel for bone tissue engineering. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 3523, doi:10.1038/s41467-019-11511-3 (2019).\u003c/li\u003e\n \u003cli\u003eHarvestine, J. N.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Osteogenic preconditioning in perfusion bioreactors improves vascularization and bone formation by human bone marrow aspirates. \u003cem\u003eSci Adv\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, eaay2387, doi:10.1126/sciadv.aay2387 (2020).\u003c/li\u003e\n \u003cli\u003eDai, K.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Generation of rhBMP-2-induced juvenile ossicles in aged mice. \u003cem\u003eBiomaterials\u003c/em\u003e \u003cstrong\u003e258\u003c/strong\u003e, 120284, doi:10.1016/j.biomaterials.2020.120284 (2020).\u003c/li\u003e\n \u003cli\u003eHerberg, S.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Combinatorial morphogenetic and mechanical cues to mimic bone development for defect repair. \u003cem\u003eSci Adv\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, eaax2476, doi:10.1126/sciadv.aax2476 (2019).\u003c/li\u003e\n \u003cli\u003eZhou, P.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Loading BMP-2 on nanostructured hydroxyapatite microspheres for rapid bone regeneration. \u003cem\u003eInt J Nanomedicine\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 4083-4092, doi:10.2147/IJN.S158280 (2018).\u003c/li\u003e\n \u003cli\u003eBoerckel, J. D.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Effects of protein dose and delivery system on BMP-mediated bone regeneration. \u003cem\u003eBiomaterials\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, 5241-5251, doi:10.1016/j.biomaterials.2011.03.063 (2011).\u003c/li\u003e\n \u003cli\u003eGeuze, R. E.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e A Differential Effect of Bone Morphogenetic Protein-2 and Vascular Endothelial Growth Factor Release Timing on Osteogenesis at Ectopic and Orthotopic Sites in a Large-Animal Model. \u003cem\u003eTissue Eng Pt A\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 2052-2062, doi:10.1089/ten.tea.2011.0560 (2012).\u003c/li\u003e\n \u003cli\u003eLee, K., Silva, E. A. \u0026amp; Mooney, D. J. Growth factor delivery-based tissue engineering: general approaches and a review of recent developments. \u003cem\u003eJ R Soc Interface\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 153-170, doi:10.1098/rsif.2010.0223 (2011).\u003c/li\u003e\n \u003cli\u003eYan, H. J.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Synthetic design of growth factor sequestering extracellular matrix mimetic hydrogel for promoting in vivo bone formation. \u003cem\u003eBiomaterials\u003c/em\u003e \u003cstrong\u003e161\u003c/strong\u003e, 190-202, doi:10.1016/j.biomaterials.2018.01.041 (2018).\u003c/li\u003e\n \u003cli\u003eMuschter, D.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Sensory Neuropeptides and their Receptors Participate in Mechano-Regulation of Murine Macrophages. \u003cem\u003eInt J Mol Sci\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, doi:ARTN 503 /ijms20030503 (2019).\u003c/li\u003e\n \u003cli\u003eKonttinen, Y. T., Imai, S. \u0026amp; Suda, A. Neuropeptides and the puzzle of bone remodeling - State of the art. \u003cem\u003eActa Orthop Scand\u003c/em\u003e \u003cstrong\u003e67\u003c/strong\u003e, 632-639, doi:Doi 10.3109/17453679608997772 (1996).\u003c/li\u003e\n \u003cli\u003eElefteriou, F. Neuronal signaling and the regulation of bone remodeling. \u003cem\u003eCell Mol Life Sci\u003c/em\u003e \u003cstrong\u003e62\u003c/strong\u003e, 2339-2349, doi:10.1007/s00018-005-5175-3 (2005).\u003c/li\u003e\n \u003cli\u003eZhang, Y.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Implant-derived magnesium induces local neuronal production of CGRP to improve bone-fracture healing in rats. \u003cem\u003eNat Med\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 1160-1169, doi:10.1038/nm.4162 (2016).\u003c/li\u003e\n \u003cli\u003eLu, Y. Z.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e CGRP sensory neurons promote tissue healing via neutrophils and macrophages. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e628\u003c/strong\u003e, 604-611, doi:10.1038/s41586-024-07237-y (2024).\u003c/li\u003e\n \u003cli\u003eHong, H. S.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e A new role of substance P as an injury-inducible messenger for mobilization of CD29(+) stromal-like cells. \u003cem\u003eNature Medicine\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 425-435, doi:10.1038/nm.1909 (2009).\u003c/li\u003e\n \u003cli\u003eKim, R. Y.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e The effect of dose on rhBMP-2 signaling, delivered via collagen sponge, on osteoclast activation and in vivo bone resorption. \u003cem\u003eBiomaterials\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 1869-1881, doi:10.1016/j.biomaterials.2013.11.029 (2014).\u003c/li\u003e\n \u003cli\u003eZara, J. N.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e High Doses of Bone Morphogenetic Protein 2 Induce Structurally Abnormal Bone and Inflammation In Vivo. \u003cem\u003eTissue Eng Pt A\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 1389-1399, doi:10.1089/ten.tea.2010.0555 (2011).\u003c/li\u003e\n \u003cli\u003eBixel, M. G.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Angiogenesis is uncoupled from osteogenesis during calvarial bone regeneration. \u003cem\u003eNature Communications\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, doi:ARTN 4575 /s41467-024-48579-5 (2024).\u003c/li\u003e\n \u003cli\u003eLu, G.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e An instantly fixable and self-adaptive scaffold for skull regeneration by autologous stem cell recruitment and angiogenesis. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 2499, doi:10.1038/s41467-022-30243-5 (2022).\u003c/li\u003e\n \u003cli\u003eKim, H.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Mesenchymal stem cell 3D encapsulation technologies for biomimetic microenvironment in tissue regeneration. \u003cem\u003eStem Cell Res Ther\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, doi:ARTN 51 /s13287-018-1130-8 (2019).\u003c/li\u003e\n \u003cli\u003eAnnamalai, R. T.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Injectable osteogenic microtissues containing mesenchymal stromal cells conformally fill and repair critical-size defects. \u003cem\u003eBiomaterials\u003c/em\u003e \u003cstrong\u003e208\u003c/strong\u003e, 32-44, doi:10.1016/j.biomaterials.2019.04.001 (2019).\u003c/li\u003e\n \u003cli\u003eYuan, Z.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Injectable GelMA Cryogel Microspheres for Modularized Cell Delivery and Potential Vascularized Bone Regeneration. \u003cem\u003eSmall\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, e2006596, doi:10.1002/smll.202006596 (2021).\u003c/li\u003e\n \u003cli\u003eLuo, C.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Biomimetic open porous structured core-shell microtissue with enhanced mechanical properties for bottom-up bone tissue engineering. \u003cem\u003eTheranostics\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 4663-4677, doi:10.7150/thno.34464 (2019).\u003c/li\u003e\n \u003cli\u003eChen, S., Chen, X., Geng, Z. \u0026amp; Su, J. The horizon of bone organoid: A perspective on construction and application. \u003cem\u003eBioact Mater\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 15-25, doi:10.1016/j.bioactmat.2022.01.048 (2022).\u003c/li\u003e\n \u003cli\u003eAkiva, A.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e An Organoid for Woven Bone. \u003cem\u003eAdv Funct Mater\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, doi:ARTN 2010524 /adfm.202010524 (2021).\u003c/li\u003e\n \u003cli\u003eHall, G. N.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Developmentally Engineered Callus Organoid Bioassemblies Exhibit Predictive In Vivo Long Bone Healing. \u003cem\u003eAdv Sci\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, doi:ARTN 1902295 /advs.201902295 (2020).\u003c/li\u003e\n \u003cli\u003ePark, Y.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Trabecular bone organoid model for studying the regulation of localized bone remodeling. \u003cem\u003eSci Adv\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, doi:10.1126/sciadv.abd6495 (2021).\u003c/li\u003e\n \u003cli\u003eGrigoryan, A.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Engineering human mini-bones for the standardized modeling of healthy hematopoiesis, leukemia, and solid tumor metastasis. \u003cem\u003eSci Transl Med\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, eabm6391, doi:10.1126/scitranslmed.abm6391 (2022).\u003c/li\u003e\n \u003cli\u003eWang, J.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Engineering Large-Scale Self-Mineralizing Bone Organoids with Bone Matrix-Inspired Hydroxyapatite Hybrid Bioinks. \u003cem\u003eAdv Mater\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, e2309875, doi:10.1002/adma.202309875 (2024).\u003c/li\u003e\n \u003cli\u003eFarah, E. N.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Spatially organized cellular communities form the developing human heart. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e627\u003c/strong\u003e, 854-864, doi:10.1038/s41586-024-07171-z (2024).\u003c/li\u003e\n \u003cli\u003eKang, J.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Systematic dissection of tumor-normal single-cell ecosystems across a thousand tumors of 30 cancer types. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 4067, doi:10.1038/s41467-024-48310-4 (2024).\u003c/li\u003e\n \u003cli\u003eMaity, A. K. \u0026amp; Teschendorff, A. E. Cell-attribute aware community detection improves differential abundance testing from single-cell RNA-Seq data. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 3244, doi:10.1038/s41467-023-39017-z (2023).\u003c/li\u003e\n \u003cli\u003eShen, B.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e A mechanosensitive peri-arteriolar niche for osteogenesis and lymphopoiesis. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e591\u003c/strong\u003e, 438-444, doi:10.1038/s41586-021-03298-5 (2021).\u003c/li\u003e\n \u003cli\u003eWorthley, D. L.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Gremlin 1 identifies a skeletal stem cell with bone, cartilage, and reticular stromal potential. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e160\u003c/strong\u003e, 269-284, doi:10.1016/j.cell.2014.11.042 (2015).\u003c/li\u003e\n \u003cli\u003eShi, Y.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Gli1 identifies osteogenic progenitors for bone formation and fracture repair. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 2043, doi:10.1038/s41467-017-02171-2 (2017).\u003c/li\u003e\n \u003cli\u003eChan, C. K.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Identification and specification of the mouse skeletal stem cell. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e160\u003c/strong\u003e, 285-298, doi:10.1016/j.cell.2014.12.002 (2015).\u003c/li\u003e\n \u003cli\u003eChan, C. K. F.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Identification of the Human Skeletal Stem Cell. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e175\u003c/strong\u003e, 43-56 e21, doi:10.1016/j.cell.2018.07.029 (2018).\u003c/li\u003e\n \u003cli\u003eDebnath, S.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Discovery of a periosteal stem cell mediating intramembranous bone formation. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e562\u003c/strong\u003e, 133-139, doi:10.1038/s41586-018-0554-8 (2018).\u003c/li\u003e\n \u003cli\u003eStrunz, M.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Alveolar regeneration through a Krt8+transitional stem cell state that persists in human lung fibrosis. \u003cem\u003eNature Communications\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, doi:ARTN 3559 /s41467-020-17358-3 (2020).\u003c/li\u003e\n \u003cli\u003eYoussef, K. K. \u0026amp; Nieto, M. A. Epithelial-mesenchymal transition in tissue repair and degeneration. \u003cem\u003eNat Rev Mol Cell Biol\u003c/em\u003e, doi:10.1038/s41580-024-00733-z (2024).\u003c/li\u003e\n \u003cli\u003eWeng, Y.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e A novel lineage of osteoprogenitor cells with dual epithelial and mesenchymal properties govern maxillofacial bone homeostasis and regeneration after MSFL. \u003cem\u003eCell Res\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, 814-830, doi:10.1038/s41422-022-00687-x (2022).\u003c/li\u003e\n \u003cli\u003eStorer, M. A.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Acquisition of a Unique Mesenchymal Precursor-like Blastema State Underlies Successful Adult Mammalian Digit Tip Regeneration. \u003cem\u003eDev Cell\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, 509-524 e509, doi:10.1016/j.devcel.2019.12.004 (2020).\u003c/li\u003e\n \u003cli\u003eYang, Y. L.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Tissue-Integratable and Biocompatible Photogelation by the Imine Crosslinking Reaction. \u003cem\u003eAdv Mater\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 2724-2730, doi:10.1002/adma.201505336 (2016).\u003c/li\u003e\n \u003cli\u003eHong, Y.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e A strongly adhesive hemostatic hydrogel for the repair of arterial and heart bleeds. \u003cem\u003eNature Communications\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, doi:ARTN 2060 /s41467-019-10004-7 (2019).\u003c/li\u003e\n \u003cli\u003eZhou, F. F.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Rapid printing of bio-inspired 3D tissue constructs for skin regeneration. \u003cem\u003eBiomaterials\u003c/em\u003e \u003cstrong\u003e258\u003c/strong\u003e, doi:ARTN 120287/j.biomaterials.2020.120287 (2020).\u003c/li\u003e\n \u003cli\u003eAran, D.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Reference-based analysis of lung single-cell sequencing reveals a transitional profibrotic macrophage. \u003cem\u003eNat Immunol\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 163-+, doi:10.1038/s41590-018-0276-y (2019).\u003c/li\u003e\n \u003cli\u003eQiu, X. J.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Reversed graph embedding resolves complex single-cell trajectories. \u003cem\u003eNat Methods\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 979-+, doi:10.1038/Nmeth.4402 (2017).\u003c/li\u003e\n \u003cli\u003eYu, G. C., Wang, L. G., Han, Y. Y. \u0026amp; He, Q. Y. clusterProfiler: an R Package for Comparing Biological Themes Among Gene Clusters. \u003cem\u003eOmics\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 284-287, doi:10.1089/omi.2011.0118 (2012).\u003c/li\u003e\n \u003cli\u003eAibar, S.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e SCENIC: single-cell regulatory network inference and clustering. \u003cem\u003eNat Methods\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 1083-+, doi:10.1038/Nmeth.4463 (2017).\u003c/li\u003e\n \u003cli\u003eSuo, S. B.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Revealing the Critical Regulators of Cell Identity in the Mouse Cell Atlas. \u003cem\u003eCell Reports\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 1436-+, doi:10.1016/j.celrep.2018.10.045 (2018).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5248946/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5248946/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eCurrent approaches for bone repair are often focused on local delivery of growth factors that are aimed at coupled angiogenesis and osteogenesis. However, delayed revascularization and regeneration of severe bone defects are still challenging. In this study, we engineered an ossification center-like organoid (OCO) that consist of inner-core bone morphogenetic and neurotrophic spheroid generated via MSCs-loaded 3D printing, alongside the interstitially distributed outer-shell proangiogenic neurotrophic phase. Our results demonstrate that collective \u0026nbsp;implantation of OCOs achieved rapid bone bridging with successive OC-like bone ossicles formation across the bone defect in a “divide-and-conquer” way. Single-cell RNA sequencing analysis unveiled a developmentally mimicking stem cell community that dominated with Krt8\u003csup\u003e+\u003c/sup\u003e skeletal stem cells (SSCs) was uniquely recruited by the pro-regenerative in-situ organoid fusion and maturation. Particularly noteworthy is the specific expansion of Krt8\u003csup\u003e+\u003c/sup\u003e SSCs concomitant with the simultaneous reduction of Has1\u003csup\u003e+\u003c/sup\u003e migratory fibroblasts (MFs) 2 weeks post-OCO implantation. Furthermore, cross-species comparisons employing machine learning revealed high resemblance of relative Krt8\u003csup\u003e+\u003c/sup\u003e SSCs/Has1\u003csup\u003e+\u003c/sup\u003e MFs composition in bone regeneration with that in public data from developmental bone tissues. Our findings advocate an approach akin to “divide-and-conquer” utilizing engineered OC-like organoids for prompt regeneration of large-sized bone defects.\u003c/p\u003e","manuscriptTitle":"Divide-and-conquer strategy with engineered “ossification center” organoids for rapid bone healing via recruiting developmental cell community","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-11-27 14:00:35","doi":"10.21203/rs.3.rs-5248946/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"31f3ace9-aba1-4937-b366-3e4113ff648f","owner":[],"postedDate":"November 27th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":40292223,"name":"Biological sciences/Biotechnology/Regenerative medicine"},{"id":40292224,"name":"Biological sciences/Biotechnology/Tissue engineering"},{"id":40292225,"name":"Physical sciences/Engineering/Biomedical engineering"}],"tags":[],"updatedAt":"2025-07-05T07:06:56+00:00","versionOfRecord":{"articleIdentity":"rs-5248946","link":"https://doi.org/10.1038/s41467-025-61619-y","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2025-07-04 04:00:00","publishedOnDateReadable":"July 4th, 2025"},"versionCreatedAt":"2024-11-27 14:00:35","video":"","vorDoi":"10.1038/s41467-025-61619-y","vorDoiUrl":"https://doi.org/10.1038/s41467-025-61619-y","workflowStages":[]},"version":"v1","identity":"rs-5248946","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5248946","identity":"rs-5248946","version":["v1"]},"buildId":"cTy_lsJlmDsVRNrSptgXS","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}