Single-cell multi-omics reveals unique regulatory mechanisms that sustains unprecedented elongation rate of bone structure, the deer antler

Single-cell multi-omics reveals unique regulatory mechanisms that sustains unprecedented elongation rate of bone structure, the deer antler | 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 Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Biological Sciences - Article Single-cell multi-omics reveals unique regulatory mechanisms that sustains unprecedented elongation rate of bone structure, the deer antler Hengxing Ba, Shidian He, Hai-Xi Sun, Xin Wang, Zhang Hang, Guo Qianchi, and 14 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6919532/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Bone growth and regeneration remain clinically important problems in medicine, and understanding the mechanism of rapid bone growth is a key to new therapeutic approaches. Deer antlers represent the fastest-growing bone structure in mammals, undergoing regeneration through endochondral ossification and exhibiting extraordinary elongation rates of up to 2 cm per day, far exceeding human epiphyseal growth plate extension of approximately 2 cm annually. This research aimed to systematically map the cellular and molecular architecture of the antler growth center by integrating single-nucleus RNA sequencing (snRNA-seq), chromatin accessibility profiling (snATAC-seq), and spatial transcriptomics. Our analysis revealed that antler mesenchymal stem cells (AnSCs) drive the proliferation of antler progenitor cells (AnPCs) through paracrine signaling. These rapidly proliferating cells maintain genomic stability and evade oncogenic transformation, while displaying distinct molecular signatures that differentiate them from osteosarcoma. AnSC-derived cells also establish a vascularized niche that supports robust angiogenesis to meet the high metabolic demands essential for rapid antler elongation. Furthermore, antlers utilize a hybrid ossification strategy that combines classical endochondral ossification with the direct transdifferentiation of hypertrophic chondrocytes into osteoblasts via PHEX⁺ intermediates. These findings redefine the key principles of endochondral ossification and offer novel insights for the development of regenerative therapies. Biological sciences/Developmental biology/Angiogenesis Biological sciences/Developmental biology/Bone development Biological sciences/Developmental biology/Cell growth antler unprecedented bone elongation genomic stability vascularized niche PHEX⁺cells hybrid ossification transdifferentiation single-cell multi-omics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction Deer antlers, capable of elongating up to 2 cm per day, represent the fastest known form of skeletal regeneration in mammals 1 – 6 . This remarkable annual growth challenges conventional models of bone development, which rely on avascular growth plates and rigidly zoned chondrocyte maturation with growth at 2 cm per year during puberty 7 . Bone fractures can rapidly heal however critical sized defects, comminuted fractures, bone trauma and tumors pose challenges that medicine is yet to fully address. In contrast, antlers overcome these constraints through a vascularized cartilage matrix and a continuous gradient of cellular proliferation and differentiation producing bone 8 , 9 . Antlers also contain a specialized antler growth center (AGC), which integrates dense vascular networks, abundant progenitor populations, and a hybrid ossification mechanism 10 . Cranial neural crest cell (CCNC)-derived antler stem cells (AnSCs) have been identified as central to this regenerative process 6 , 11 – 13 , however their interactions with downstream progenitors and vascular niches remains poorly understood. Elucidating the molecular mechanisms that enable such rapid yet orderly growth is crucial, not only for understanding evolutionary adaptations but also for informing regenerative strategies in high-demand tissues. In addition, elucidating these interactions may help explain how antlers sustain extreme growth rates without progressing to oncogenic transformation 14 , in contrast to pathological conditions such as osteosarcoma. To address this, we applied integrated single-nucleus RNA sequencing (snRNA-seq), chromatin accessibility profiling (snATAC-seq), and spatial transcriptomics to map the cellular and molecular architecture of the AGC (Fig. 1 a). Our analysis revealed three key evolutionary innovations: 1) a stem–progenitor hierarchy maintained by AnSC-derived paracrine signals; 2) hypoxia-driven vascularization that fuels metabolic needs and supports appositional growth; 3) hybrid ossification involving both chondroclasia and the direct transdifferentiation of hypertrophic chondrocytes into osteoblasts via PHEX⁺ intermediates. Together, these features establish a "continuous flow" growth model that emphasizes speed without compromising genomic integrity. By bridging comparative biology with regenerative medicine, this study redefines the principles of endochondral ossification. It uncovers transcriptional and epigenetic programs that balance rapid osseous growth with structural stability, demonstrating nature’s capacity to optimize developmental mechanisms for high-performance tissue regeneration. These findings offer actionable insights for enhancing bone repair and engineering vascularized skeletal tissues. Results Integrated snRNA-seq and snATAC-seq analyses reveal the cellular landscape of the antler growth center To resolve the cellular landscape of the AGC at both the transcriptional and chromatin accessibility levels, snRNA-seq data was integrated from the five distinguishable AGC tissue layers of RM, PC, TZ, CA, and MC 6 (Fig. 1 a). This stratified framework provided a comprehensive view of the spatial and functional organization of the AGC. Following stringent quality control filtering (Figure S1 a, S1b and S1c), a total of 53,949 high-quality cells were retrieved for downstream analysis (Figure S1 d). Unsupervised clustering identified 10 distinct cell types (Fig. 1 b, 1 c, 1 d), each corresponding to a specific cluster (Figure S2a), as defined by established gene markers (Figure S2b) 12 , 15 , 16 . These included NT5E⁺, TWIST2⁺, SFRP2⁺ AnSCs; TNC⁺, TNN⁺ antler progenitor cells (AnPCs) and with MK167 proliferative AnPCs; COL2A1⁺ chondrocytes; COL10A1⁺ hypertrophic chondrocytes; ACP5⁺ chondroclasts; ACTA2⁺ mural cells; PECAM1⁺CD34⁺CDH5⁺ endothelial cells, CSF1R⁺ monocytes/macrophages, and TPSB2⁺ mast cells (Fig. 1 d). Enrichment analysis of differentially expressed genes within these clusters provided additional validation for the identified cell types (Figure S2c). Notably, AnPCs clearly comprised both proliferative and non-proliferative populations, suggesting that cellular proliferation is one of the key contributors to the rapid growth of antlers. After application of quality control thresholds to filter out low-quality nuclei (Figure S3a), unsupervised clustering of the snATAC-seq data identified the same 10 cell types based on the same marker genes, resulting in a total of 57,722 high-quality cells (Fig. 1 e, 1 f, 1 g and Figure S3b). The transcriptional activity and chromatin accessibility exhibited a high degree of concordance at cell type level, with an 87.6% consistency rate (Fig. 1 h), along with a complete (98.9%) agreement in their tissue-specific distributions. The proportional distributions of the ten cell types across the two omics datasets were also highly similar (Fig. 1 i). Data showed, AnSCs were exclusively localized in the RM, while both proliferative and non-proliferative AnPCs were predominantly found in the RM and PC. Chondrocytes were primarily distributed across the other four tissue layers (excluding RM), with hypertrophic chondrocytes specifically enriched in the CA and MC. Chondroclasts and monocytes/macrophages were largely confined to the MC. In contrast, vascular-associated cells, including mural cells and endothelial cells, were present across all five tissue layers with a predominance in RM. We also utilized previously generated bulk RNA-seq data 13 from the five tissue layers and performed deconvolution analysis, which revealed a consistent distribution of cell type proportions (Figure S2d). Furthermore, the global transcriptome and chromatin accessibility profiles of the identified cell types exhibited strong correlations (Fig. 1 j). Correlation analysis revealed three major groups: 1) AnSCs, AnPCs, and mural cells (Pearson correlation’s r = 0.87 ~ 0.97), 2) chondrocytes and hypertrophic chondrocytes (r = 0.93 ~ 0.99), and 3) all cells derived from peripheral blood (r = 0.74 ~ 0.99). Overall, our results demonstrate a highly consistent cellular composition, reflected at both the transcriptional and chromatin accessibility levels. Spatially resolved cellular architecture in the antler growth center We utilized Stereo-RNAseq to explore the spatial distribution of cell types within the AGC. After stringent quality control filtering (Figure S4a, S4b and S4c), a total of 57,311 cells were retrieved for the unsupervised clustering, which resided in 17 clusters (Figure S5a). These clusters were subsequently annotated into 11 cell types based on known marker genes (Fig. 2 a, 2 b; Figure S5b, S5c and S5d). All the previously identified cell types from snRNA-seq and snATAC-seq analyses were detected, except mast cells, likely due to their low abundance. Notably, these identified cells exhibited a spatially sequential organization: AnSCs were located in the distal-most region of the AGC, followed by proliferative and non-proliferative AnPCs, which correspond to RM cells in the inner RM layer and prechondroblasts in the PC layer, and then chondrocytes, hypertrophic chondrocytes and PHEX + cells in the CA and MC layers. This arrangement reflects the differentiation trajectory originating from the AnSCs. Chondroclasts, which accounted for 18% of the total cell population, were interspersed among hypertrophic chondrocytes in longitudinal columns. Monocytes/macrophages were distributed alongside with chondroclasts, supporting the notion that chondroclasts arise from differentiation of monocytes. Additional cell types, including PHEX⁺ cells and osteoblasts, were identified by Stereo-seq but were not detected in the snRNA-seq and snATAC-seq data. PHEX⁺ cells, accounting for approximately 5% of the total cell population, and were predominantly located in the vicinity of chondrocytes, particularly in the vicinity of specialized hypertrophic chondrocytes. The PHEX gene is known to encode a membrane-bound metalloproteinase that plays a critical role in phosphate metabolism and bone mineralization 17 , some hypertrophic chondrocytes may undergo further differentiation into PHEX⁺ cells with enhanced control of mineralization. Osteoblasts accounted for 13% of total cells and were predominantly localized at the proximal region of AGC, corresponding to the ossification zone below the MC, and are responsible for sustaining rapid bone formation. Consistent with the snRNA-seq data of the five tissue layers, mural cells and endothelial cells showed a continuous distribution from the distal to the proximal regions of the AGC. RUNX2, a key transcription factor that regulates both early chondrocyte and osteoblast differentiation and skeletal mineralization 18 – 20 . It was broadly expressed across the AnSC-derived lineages, including AnPCs, chondrocytes, and hypertrophic chondrocytes, with particularly high expression levels observed in AnPCs and hypertrophic chondrocytes (Fig. 2 c and 2 d). Chromatin accessibility analysis revealed strong RUNX2 activity specifically in the RM region, encompassing both AnSCs and AnPCs (Fig. 2 e), underscoring its essential role in AnSC differentiation. MCAM, encodes a critical adhesion factor for angiogenesis and is primarily expressed in vascular cells 21 , 22 . It exhibited high expression levels in both endothelial and mural cells in the AGC (Fig. 2 g and 2 h), accompanied by marked chromatin accessibility at its genomic locus (Fig. 2 i). Immunohistochemical staining further confirmed the consistent expression patterns of both RUNX2 and MCAM at the protein-level (Fig. 2 f and 2 j). In the AGC, AnPCs further differentiate into chondrocytes and, as expected, express early chondrogenic marker genes, including SOX9, SOX5, SOX6, ACAN and COL9A1 (Fig. 2 k), as well as COL2A1 (Fig. 2 b), which also is typically expressed in chondroprogenitor cells 23 . Interestingly, in addition to RUNX2, AnPCs also express osteogenic marker genes such as SP7, IBSP, SPP1, COL1A1, and CDH11, indicating that AnPCs possess both chondrogenic and osteogenic features and differentiation potential. Bank and Newbrey (1982) reported the presence of RM cells, prechondroblasts (proliferative capacity), and chondroblasts in the distal AGC (Fig. 2 l). In the present study, we redefined the corresponding cell populations as AnSCs, proliferative AnPCs, and non-proliferative AnPCs. Furthermore, within the previously defined chondrocyte region, we delineated three distinct cell types: chondrocytes, hypertrophic chondrocytes and PHEX⁺ cells (mineralized chondrocytes). The observed spatial distribution of these cell types also aligned well with the established tissue layer architecture described in our previous studies 6 . Overall, our comprehensive spatial mapping of the cellular architecture in the AGC not only confirms and further refines the finding of previously published histological structure, but also provides novel insights into the mechanism underlying rapid formation of bone in antlers at the cellular level. Signaling from AnSCs promotes proliferation of AnPCs We found that AnPCs consisted of distal proliferative a subset (Fig. 3 a and 3 b), suggesting the rapid addition of AnPCs-derived tissues, at least in part, to the unprecedented growth rate of antlers. Previous studies have demonstrated that IGF1 strongly stimulates AnSCs in vitro 24 . In the present study, IGF1 was specifically expressed in AnSCs, while its receptor, IGF1R, was highly expressed in proliferative AnPCs (Fig. 3 c). A similar expression pattern was observed for PTHLH and its receptor PTH1R, which are also known to regulate chondrocyte proliferation and differentiation 25 . These findings suggest that AnSCs promote AnPC proliferation through paracrine signaling (Fig. 3 d). To further explore potential signaling interactions, we employed CellChat to investigate cell–cell communication between AnSCs and proliferative AnPCs, as well as among proliferative AnPCs themselves. Both snRNA-seq and Stereo-seq data revealed significantly stronger signals outgoing from the AnSCs to proliferative AnPCs than in the reverse direction (Fig. 3 e). Several active ligand–receptor (LR) pairs mediating intercellular communication between AnSCs and proliferative AnPCs were identified, including WNT5A/11-FZD, THBS4–SDC1/CD47, THBS1–SDC1, PDGFA–PDGFRA, PTN–SDC1, PTN–SDC2, IGF1–IGF1R, and FGF16–FGFR2 (Fig. 3 f). Notably, MDK, a ligand uniquely expressed by AnSCs, interacted with multiple receptors located on the proliferative AnPCs, including SDC1, SDC2, NCL and LRP1. MDK has high homology with PTN and is well known to promote cell proliferation and inhibit apoptosis, particularly in rapidly growing and regenerating tissues 26 – 29 . In addition to the AnSC–AnPC signaling, we also observed significant signaling activities that occurred among the proliferative AnPCs themselves, mediated by LR pairs such as TNC–SDC1, TNN–SDC1, PTN–SDC1, PTN–SDC2, PTN–NCL, FN1–ITGA4/ITGB1, and FN1–SDC1. These LR interactions were further spatially mapped in the AGC (Fig. 3 g). To validate these signaling interactions experimentally, we established an in vitro co-culture system for AnSCs and AnPCs (Fig. 3 h). The results demonstrated that putative substances secreted by AnSCs significantly promoted AnPC proliferation in a time-dependent manner (Fig. 3 i). Together, these findings highlight the critical role of AnSC-derived molecular signaling in driving the rapid proliferation of AnPCs, in turn promoting elongation of antlers. Enhanced genomic stability and multipotent potential of proliferative AnPCs The unprecedented growth rate of antlers imposes considerable stress on cellular and genomic integrity. To investigate how the AGC copes with this stress, we assessed the activity (AUCell scores) of several key pathways in proliferative AnPCs. The results revealed that proliferative AnPCs exhibited significantly higher activity in DNA damage response ( p < 2.22 × 10 –16 ) and DNA repair pathways ( p < 2.22 × 10 –16 ) compared to their non-proliferative counterparts (Fig. 4 a). Interestingly, although statistically significant, only minor differences were observed in apoptotic signaling ( p < 2.22 × 10 − 7 ) and in DNA damage response signal transduction via p53 signaling ( p < 2.75 × 10 − 8 ) between the proliferative and non-proliferative AnPCs. Chromatin accessibility analysis showed consistent results (Fig. 4 a) and demonstrated that proliferative AnPCs possess an enhanced capacity to maintain genomic stability without excessive activation of apoptotic pathways, supporting the “fast yet stable” growth characteristic of antlers. Moreover, the spatial expression patterns of apoptotic factors (e.g., CASP9, ANXA5, ANXA6, BAK1, BID and TP53) in proliferative and non-proliferative AnPCs, relative to other AnSC-derived cell types (Fig. 4 b and 4 c), may represent a safeguard mechanism. This apoptotic preparedness likely enables the rapid elimination of damaged cells during intense proliferation and differentiation 30 , thereby preserving tissue integrity and ensuring genomic stability during the process of rapid growth. Next, we integrated transcriptome and chromatin accessibility to identify active transcription factors (TFs) in the proliferative AnPCs. This analysis revealed that TWIST2, MEOX2, EGR2, PRRX2, and ZNF37A were the top five most active TFs in the proliferative AnPCs (Fig. 4 d). Previous studies have shown that TWIST2 and PRRX2 are signature genes of CNCC-derived ectomesenchyme 12 , and that AnSCs originate from CNCCs 4 , 31 , 32 . Given that TWIST2 and PRRX2 are associated with high cellular plasticity 33 , we hypothesized that proliferative AnPCs may retain transcriptional signatures linked to multipotent differentiation potential. To further investigate this, we performed DAVID Gene Ontology (GO) enrichment analysis using these five TFs and their target genes. The results showed that enrichment in biological processes (BPs) was associated with embryo development (MEOX2), blood vessel development (MEOX2 and EGR2), neurogenesis (EGR2), and osteoblast differentiation (TWIST2) (Fig. 4 e), further supporting the notion that proliferative AnPCs possess multipotent differentiation potential. In contrast, GO enrichment analysis of non-proliferative AnPCs did not reveal these terms (Figure S6a). We also applied CytoTRACE2 to calculate scores reflecting the pluripotency of proliferative and non-proliferative AnPCs. The results showed that proliferative AnPCs exhibited higher scores compared to non-proliferative AnPCs (Figure S6b). Thus, the proliferative AnPCs have a unique phenotype associated with genomic and apoptotic protection and have TFs suggestive they retain multipotent differentiation potential. PHEX⁺ cells contribute to antler cartilage mineralization via hedgehog signaling As PHEX⁺ cells can contribute to approximately 5% of the total cell population they were investigated further. Spatial analysis of this transmembrane peptidase involved in mineralization revealed that PHEX⁺ cells were predominantly localized around hypertrophic chondrocytes (Fig. 5 a). Genes highly expressed in PHEX⁺ cells were significantly enriched in BPs related to bone mineralization ( p = 8.16 × 10 − 7 ) (Fig. 5 b and S7a). Moreover, the spatial expression patterns of bone mineralization-related genes (e.g., LGR4, FITM5, ASPN, IBSP, LOX, BGLAP, MMP9 and MMP13) showed strong concordance with the distribution of PHEX⁺ cells (Fig. 5 c). To further investigate the molecular signals driving bone mineralization in the AGC, we applied CellChat to infer intercellular communication among chondrocytes, hypertrophic chondrocytes and PHEX⁺ cells. The results revealed that chondrocytes and hypertrophic chondrocytes acted as the main signaling sources, transmitting significantly strong signals to PHEX⁺ cells (Fig. 5 d). Among the identified pathways (Figure S7b), Hedgehog (HH) signaling emerged as a key signaling axis (Fig. 5 e). HH signaling is a crucial regulator in cell fate determination and tissue patterning, particularly in cartilage mineralization 34 . In the HH pathway, the ligand IHH was highly expressed in chondrocytes and hypertrophic chondrocytes, whereas the receptors PTCH1 and the transducer SMO were predominantly expressed in PHEX⁺ cells (Fig. 5 f and Figure S7c). Notably, HHIP, a soluble inhibitory factor that binds extracellular IHH and limits its interaction with PTCH1, was also highly expressed in PHEX⁺ cells. This indicates a feedback mechanism within PHEX⁺ cells to fine-tune HH signaling and maintain mineralization homeostasis. Additionally, interaction between IHH and its receptors PTCH1/SMO stimulates vascular invasion into the hypertrophic zone, activates osteoblast differentiation, and promotes bone matrix deposition 35 . High expression of MMP9 and MMP13 in PHEX⁺ cells further facilitates vascular invasion by degrading the extracellular matrix, creating conditions favorable for osteoblast migration and bone matrix deposition 36 . To further validate the identity and signaling involvement of PHEX⁺ cells, we performed unsupervised clustering analysis based on snRNA-seq data, which revealed four distinct chondrocyte clusters (C0–C3, Figure S7d). Among these clusters, C3 displayed markedly higher expression of PHEX, identifying it as the PHEX⁺ cell population (Figure S7e and S7f). This cluster was predominantly localized in the MC layer (Figure S7g). Consistent with their role in mineralization, C3 cells also showed elevated expression of IBSP, IFITM5, MMP9 and MMP13 (Figure S7h). The expression profiles of IHH, PTCH1, HHIP, and SMO across the four clusters closely mirrored the spatial expression pattern (Figure S7i), for example, IHH was highly expressed in C0, while PTCH1 was enriched in C3. Collectively, these findings highlight the pivotal role of PHEX⁺ cells in driving antler cartilage mineralization, orchestrated by HH signaling derived from surrounding chondrocytes and hypertrophic chondrocytes. PHEX⁺ cells act as pivotal transitional intermediates in the direct transdifferentiation of hypertrophic chondrocytes into osteoblasts. As a bony appendage, the antler growth is contingent on osteoblast origin, differentiation, and maturation to sustain its rapid and extensive expansion for the tissue. Interestingly, enrichment analysis revealed that genes (e.g., IBSP, BGLAP, SPP1, SP7, LOX, and COL1A1) being highly expressed in PHEX⁺ cells are involved in osteoblast differentiation ( p = 6.83 × 10 –13 ; Fig. 5 b). This result indicates that PHEX + cells exhibit osteoblast-like transcriptional feature. Previous studies have reported that, during endochondral ossification, a subset of hypertrophic chondrocytes can avoid apoptosis and instead directly transdifferentiate into osteoblasts 37 , 38 . Moreover, RUNX2 has been shown to inhibit terminal hypertrophic chondrocyte apoptosis and promote their transdifferentiation into osteoblasts 39 , 40 . In the AGC, we observed strong RUNX2 expression not only in AnPCs but also in hypertrophic chondrocytes (Fig. 2 c, 2 d, 2 e and 2 f). Importantly, PHEX⁺ cells were predominantly localized in the vicinity of hypertrophic chondrocytes (Fig. 5 a). These findings suggest that antler hypertrophic chondrocytes may give rise to osteoblasts via an intermediate PHEX⁺ cell type. Since deer are not yet established as transgenic model organisms, genetic tools such as PHEX-Cre for lineage tracing are currently unavailable. As a result, directly testing this hypothesis by tracking the fate of PHEX ⁺ cells remains technically challenging. To circumvent this limitation, we leveraged the spatially resolved cellular architecture of the AGC. Specifically, we identified two subtypes from osteoblasts based on their proximity to PHEX⁺ cells (Fig. 5 g): 1) OB-D, located exclusively adjacent to PHEX⁺ cells; and 2) OB-P, situated in regions lacking direct contact with PHEX⁺ cells. We hypothesized that OB-D cells are derived from hypertrophic chondrocytes through direct transdifferentiation, while OB-P cells are peripheral blood or bone marrow mesenchymal stem cells (BMSCs) through osteogenic differentiation. Dimension reduction analysis confirmed that OB-D and OB-P represent two distinct osteoblast populations (Fig. 5 h). Both subtypes expressed canonical osteoblast markers (Fig. 5 i), including IBSP, MGP, BGLAP, OMD, SP7, and SPP1. Among them SP7, a critical transcription factor regulating BMSC commitment to the osteoblast lineage 41 , BGLAP, a marker of mature osteoblasts 42 , and OMD, involved in osteoblast function and matrix mineralization 43 , were relatively highly expressed in OB-P cells. In contrast, MGP, which modulates the balance between cartilage and bone mineralization 44 , and IBSP, which facilitates osteoblast adhesion and mineralization 45 , were more highly expressed in OB-D cells. These results support our proposed origins of the two subtypes. Furthermore, differentiation flow along the pseudotime trajectory illustrated a developmental progression from chondrocytes and hypertrophic chondrocytes, through PHEX⁺ cells, toward OB-D cells (Fig. 5 j, 5 k and Figure S8a, S8b and S8c). Notably, two trajectory endpoints were observed: one at hypertrophic chondrocytes and the other at OB-D cells. This suggests that a small subset of hypertrophic chondrocytes transdifferentiated into OB-D cells via the PHEX⁺ intermediate type. Along this pseudotime trajectory, the expression of osteoblast-related genes (e.g., IBSP, SPP1, BGLAP and SP7) progressively increased, whereas the hypertrophic chondrocyte marker COL10A1 was downregulated during the transdifferentiation process (Fig. 5 l). To further validate whether OB-D cells are formed via transdifferentiation, we employed an in vivo model using nude mice for creation of xenogeneic antlers that consist of avascularized cartilage via transplantation of antler stem cell tissue (AP) 46 . On day 30 of AP transplantation, osteoblasts began to emerge in the center of the large avascularized cartilage nodules without visible signs of osteogenesis via chondroclasia, and by day 60, osteoblasts in the trabeculae and woven bone were found to be broadly distributed (Fig. 5 m). In parallel, we also generated xenogeneic antlers that consists of vascularized cartilage using the RFP-expressing nude mice through the same procedure but the skull periosteum on the transplantation site thoroughly scraped. Our previous studies have shown that the vascular endothelial cells in xenogeneic antlers are primarily derived from the RFP-expressing host mice 47 . Immunofluorescent analysis of osteoblast marker genes, BGLAP and SP7, revealed two distinct regions of expression (Fig. 5 n): one surrounding blood vessels (OB-P), indicating osteoblasts differentiated from perivascular BMSCs; and another in the avascularized cartilage/bone nodules (OB-D), implicating that the osteoblasts are transdifferentiated from hypertrophic chondrocytes or PHEX⁺ cells. These findings strongly support that a subset of osteoblasts are derived via transdifferentiation from chondrocyte-lineage cells. Antler exhibits a distinct transcriptional profiling from osteosarcoma despite their rapid growth To further characterize the transcriptional profiling of AnSC-derived cells, we compared their transcriptomic signatures with those of other cartilage- and bone-associated tissues, including normal intervertebral discs (ID), embryonic long bones (EL), and pathological osteosarcoma (Os) (Fig. 6 a). We first leveraged publicly available single-cell RNA-seq datasets from ID, EL, and Os to extract cartilage/bone-related cell populations (Figure S9a, S9b and S9c). Subsequently, we employed the scPred tool to compare the transcriptional profiles of AnSC-derived cells with those of these non-antler tissues (Fig. 6 b). Our scPred analysis revealed that AnSCs exhibited striking similarity in expression profiling to stromal mesenchymal stem cells (MSCs) in Os, with an average classification probability exceeding 88% (Fig. 6 c). Notably, MSCs represent normal cells and not of tumor origin. However, they have been shown to promote tumor cell proliferation and metastasis in various types of cancer, including osteosarcoma 48 . AnPCs showed partial similarity to osteoprogenitors from EL, while mature AGC chondrocytes demonstrated greater similarity to chondrocytes from both EL and ID, and only limited resemblance to Os chondroblasts. These findings highlight the unique, non-pathological nature of rapid antler growth, emphasizing that the developmental trajectory and transcriptomic profiling of AnSC-derived cells are fundamentally distinct from those of tumorigenic processes, despite superficial similarities in growth dynamics. AnSC-derived cells provide a vascularized microenvironment rapid for cartilage growth and ossification Unlike cartilage in other parts of the body, antler cartilage is highly vascularized. We hypothesized that AnSC-derived cells contribute to the formation of a microenvironment that promotes vascular development (Fig. 7 a). To investigate how the AGC microenvironment influences vascular development, we assessed the AUC scores of several key GO BPs along the developmental trajectory of AnSC-derived cells (Figure S8d). Our analysis revealed that Hypoxia inducible factor 1alpha signaling (Pearson correlation, r = 0.84), regulation of cellular response to hypoxia (r = 0.89), and regulation of blood vessel endothelial cell migration (r = 0.84) predominantly occurred in the proximal region of the AGC, which is primarily composed of hypertrophic chondrocytes and likely creates a hypoxic microenvironment (Figure S10a). Under these conditions, HIF1A induces the expression of VEGFA, which in turn promotes angiogenesis, as indicated by enrichment in GO BPs such as angiogenesis involved in wound healing (r = 0.84). Consistently, the expression levels of both HIF1A and VEGFA were significantly upregulated in this region (Fig. 7 b). In contrast, BPs such as branching involved in blood vessel morphogenesis (r = 0.77) and blood vessel maturation (r = 0.82) were mainly observed in the distal region of the AGC (Figure S10a). Genes associated with vascular maturation and stabilization (e.g., PDGFB, ANGPT1, and ANGPT2) were highly expressed in the distal AGC (Fig. 7 b). Immunofluorescence staining further confirmed these findings, showing that mural cells were predominantly localized in the distal AGC (Fig. 7 c). These results indicate that the AGC provides distinct vascular microenvironments along its distal-to-proximal axis. Vascular cells, including mural and endothelial cells, were primarily mitotically active in the RM and PC layers (Figure S10b), suggesting that these cells are activated by signals from the AGC microenvironment. To further explore this, we applied CytoTRACE2 to calculate scores reflecting the differentiation potential of mural and endothelial cells. Based on CytoTRACE2 scores, we defined two cellular states (Fig. 7 d): an activated state (CytoTRACE2 score > 0.3) and a quiescent state (CytoTRACE2 score < 0.3). We then used CellChat to compare the cell–cell signaling networks between these two states. In endothelial cells, LR pairs associated with proliferation and differentiation, such as PTN–NCL, MDK–NCL, MDK–ITGA6_ITGB1, and WNT5A–MCAM, were significantly upregulated in the activated state compared to the quiescent state (Fig. 7 e). Conversely, BMP2/4/6–BMPR1A_BMPR2 signaling was notably downregulated in the activated state. Previous studies have shown that BMP signaling is associated with vascular calcification 49 , consistent with its primary activity during chondrocyte hypertrophy. Interestingly, VEGFA signaling exhibited receptor-specific regulation: VEGFA–KDR signaling was upregulated, whereas VEGFA–FLT1 signaling was downregulated in the activated state. This aligns with the distinct functional roles of the two receptors, KDR promotes vasculogenesis 50 , while FLT1 acts as a decoy receptor that inhibits angiogenic signaling 51 . In mural cells, we observed a similar upregulation of PTN, MDK, and WNT5A signaling, along with additional signals such as TNC, TNN, and FN1 (Fig. 7 e), all known to facilitate antler regeneration 15 , 52 , 53 . In contrast, platelet-derived growth factor LR signaling (PDGFA/D–PDGFRB), critical for recruiting mural cells to nascent vessels and promoting vascular maturation and stability 54 , was downregulated in the activated state. This downregulation likely represents a functional shift, even as mural cells undergo reactivation. Overall, these findings demonstrate that AnSC-derived cells establish a microenvironment that favors vascular formation and development through spatially distinct mechanisms, and they reveal unique LR interactions that drive this process. Enriched TF network drives vascular niche cell activation Next, we performed a joint analysis of transcriptional and chromatin accessibility profiles to identify differentially enriched TFs between activated and quiescent states in both endothelial and mural cells. The results revealed that TF enrichment levels in activated endothelial and mural cells were positively correlated with CytoTRACE2 scores, indicating that increased TF enrichment was associated with enhanced cellular differentiation (Fig. 7 f). Top enriched TFs in activated endothelial cells included E2F7, E2F8, CENPN, ETV4, and THAP11, whereas those enriched in activated mural cells included PRRX1, CEBPA, GTF3A, ZBTB32, and YBX1. Most of these TFs are known to be involved in cell cycle regulation. PRRX1, a marker of mesenchymal cells, has been shown to promote cell activation, including increased migratory and invasive capacities 55 . KEGG enrichment analysis of these TFs and their target genes further revealed that, in the activated state, they are primarily involved in cell cycle progression, genomic stability and developmental signaling pathways (Fig. 7 g and S10c). In contrast, TFs highly enriched in quiescent endothelial and mural cells, along with their targets, showed no enrichment in these categories but were instead associated with inflammatory and immune-related pathways, such as the TNF signaling pathway (Figure S10d). These findings suggest that the activation of vascular niche cells triggers distinct transcriptional programs that promote both proliferative and differentiation capacity, potentially contributing to their rapid growth. Activated vascular cells in AGC exhibit distinct metabolic signatures compared to other counterparts in normal and pathological tissues To further characterize antler vascular cells in the activated state, we compared their transcriptomic profiles with those of other cartilage/bone-associated tissues, including normal ID and pathological Os (Fig. 8 a). Notably, the proportion of mitotically active vascular cells was significantly higher in the AGC compared to both ID and Os (Fig. 8 b). Similarly, cells in the activated state, defined as those with CytoTRACE2 scores greater than 0.3 in mural and endothelial populations, were further analyzed (Fig. 8 c). We identified the differentially expressed genes (DEGs) in activated AGC vascular cells relative to those in ID and Os (Fig. 8 d). Interestingly, the number of overlapping DEGs was markedly higher in the co-upregulated group (corresponding to ID_down & Os_down; 1,371 for endothelial cells and 1,425 for mural cells) and the co-downregulated group (ID_up & Os_up; 2,648 and 2,429, respectively) than in the discordant groups (ID_up & Os_down or ID_down & Os_up; 44 and 48). These findings suggest that activated vascular cells in the AGC exhibit a transcriptional program distinct from both ID and Os. KEGG enrichment analysis of the co-upregulated and co-downregulated genes further revealed an upregulation of the citrate cycle (TCA cycle) pathway ( p = 6.17 × 10⁻⁴ for endothelial cells and p = 2.94 × 10⁻⁶ for mural cells) and a downregulation of oxidative phosphorylation ( p = 1.57 × 10⁻¹⁶ and p = 4.34 × 10⁻²², respectively) (Fig. 8 d and Figure S11a and S11b). This metabolic shift suggests that activated antler vascular cells may prioritize anabolic biosynthetic processes over mitochondrial oxidative metabolism, a pattern commonly observed in rapid growth tissues. The thermogenic uncoupling seen is these rapidly growing cells is also fascinating. This results is consistent with our previous finding that antler cells depend on the TCA cycle to obtain their energy 14 . Additionally, the pronounced downregulation of the ribosome pathway ( p = 2.82 × 10⁻¹⁵ and p = 4.80 × 10⁻¹⁵) may reflect a shift from global protein synthesis to selective translational control, ensuring the efficient production of key regulatory proteins needed to support rapid and energy-efficient growth. Discussion The unprecedented elongation rate of deer antlers, which can reach up to 2 cm per day, represents the fastest known skeletal growth in vertebrates. By integrating single-cell multi-omics, this study reveals how antlers overcome the metabolic and structural constraints that typically limit somatic bone growth. We identify three key evolutionary innovations: expansive progenitor proliferation, vascularized cartilage, and hybrid ossification. Collectively, these mechanisms enable antlers to achieve exceptional growth velocity while maintaining tissue stability. Antler growth challenges the conventional model of endochondral ossification by redefining spatial tissue organization, regulatory dynamics, and cell fate decisions. In contrast to somatic growth plates, which maintain strict zonal segregation (resting, proliferative, hypertrophic) to accommodate mechanical loading, the AGC functions as a continuous system optimized for maximal elongation. Within the AGC, overlapping zones of proliferation and differentiation allow AnPCs to undergo sustained divisions (2 cm/day) 2 , far exceeding those observed in growth plates (2 cm/year) 7 , before committing to hypertrophy. This continuous flow architecture is regulated by two levels of control: systemic endocrine cues, such as IGF1 surges that coincide with peak antler growth 56 – 59 , and localized TF cues of autocrine or paracrine signals and receptor activation, including AnSC-derived IGF1, PTHLH, and MDK identified in this study. This evolutionary adaptation reflects a clear trade-off. Antlers sacrifice zonal precision, which is critical for structural integrity under mechanical stress, in order to support rapid and extensive tissue expansion. A striking example is the vascularization of the AGC cartilage, which contrasts sharply with the avascular structure of somatic growth plates. Vascular channels permeate the cartilage matrix, ensuring continuous delivery of oxygen and nutrients. This enables chondrocytes to reach larger sizes and persist in greater numbers than their counterparts in somatic plates. The presence of vasculature also accelerates ossification, as osteoblasts can reach mineralization fronts through preformed vascular conduits, bypassing the delay caused by metaphyseal vascular invasion 60 . These adaptations effectively resolve the metabolic bottlenecks that typically constrain hypertrophic expansion in avascular cartilage. A central feature of antler growth is the amplification and maintenance of a proliferative reservoir. This stem–progenitor hierarchy is sustained by AnSC-derived ligands, including IGF1, PTHLH, and MDK, which collectively ensure the continuous generation of proliferating AnPCs. These AnPCs exhibit elevated activity of cell cycle regulators and enhanced DNA repair capabilities, enabling rapid yet genomically stable cell division. In contrast, somatic growth plates restrict chondrocyte proliferation to preserve mechanical integrity, prioritizing quality control over growth rate 10 , 61 . Notably, the IHH–PTHLH feedback loop, a conserved mechanism in skeletal development 62 – 64 , is markedly amplified in antlers. Chondrocyte-derived IHH maintains PTHLH expression in AnSCs, thereby sustaining the proliferative pool of AnPCs. In parallel, IGF1–IGF1R signaling works synergistically with mitogenic pathways such as WNT, FGF, and PDGF to drive cellular hyperplasia. This multilayered regulatory architecture ensures that antler development remains precisely controlled through mechanisms such as regulated apoptosis, effectively preventing the uncontrolled proliferation characteristic of pathological conditions like osteosarcoma. These findings align with our transcriptomic comparison, which demonstrated that AnPCs share greater molecular similarity with osteoprogenitors located in embryonic long bone, while exhibiting marked differences from osteosarcoma cells. In addition, key regulatory networks such as TWIST2, EGR2, and MEOX2 were identified, suggesting that proliferating AnPCs possess multipotent potential. In our previous xenogeneic antler implantation experiments using fluorescent nude mice, we observed that stem–progenitor cells were capable of differentiating into mural-like cells 47 . Antlers resolve a fundamental biological paradox: sustaining centimeter-scale daily growth in a tissue where nutrient supply is typically constrained by diffusion across millimeter-scale distances 65 . This is achieved through the vascularization of the cartilage matrix, which represents a striking deviation from conventional growth plate biology. In the present study, we identified extensive ligand and receptor interactions between AnSC-derived microenvironmental cells and activated vascular cells. These interactions are closely associated with cell proliferation and differentiation and are accompanied by enriched TFs and targets involved in cell cycle progression and developmental signaling pathways. Angiogenesis within the AGC gives rise to a dense vascular network that permeates the antler cartilage 8 . This network not only delivers oxygen and nutrients but also provides a structural scaffold for appositional growth, which is the principal mechanism of antler elongation. It divides the cartilage into longitudinal trabeculae lined with perivascular osteoblasts. Compared to the somatic growth plate, this vascular strategy in AGC cartilage confers three key advantages. First, it offers enhanced metabolic support, as hypertrophic chondrocytes maintain direct contact with vasculature, thereby preventing hypoxia-induced apoptosis. Second, it provides greater spatial flexibility, enabling new cartilage to be added more efficiently in an appositional manner to existing trabeculae. Third, it increases ossification efficiency, as osteoblasts can immediately colonize mineralization fronts via vascular routes, bypassing the delayed metaphyseal invasion that occurs in growth plates 60 . In contrast to growth plates, where angiogenic factors such as VEGF are restricted to hypertrophic zones, antlers exhibit continuous expression of these factors throughout the cartilage matrix, creating a sustained pro-angiogenic environment. For example, proximally, angiogenesis is driven by hypoxia via HIF1α and VEGFA, while distally, vascular maintenance and maturation are promoted by PDGFB and ANGPT1. This spatially coordinated regulatory system ensures that vascular expansion keeps pace with the rapid tissue growth of the antler. Osteogenesis in the somatic growth plate is primarily achieved through chondroclasia 66 . In contrast, our study shows that antler ossification incorporates an additional mechanism: the direct transdifferentiation of hypertrophic chondrocytes into osteoblasts, a process that is absent in somatic growth plates. These cells downregulate chondrogenic markers such as COL10A1 and upregulate osteogenic markers including IBSP and BGLAP. PHEX + cells eventually differentiate into mature osteoblasts. At the molecular level, this transdifferentiation may be driven by sustained RUNX2 expression, which help reduce the lag between cartilage resorption and bone formation 67 . Importantly, IHH produced by chondrocytes-derived cells activates PTCH1/SMO signaling in PHEX + cells, thereby promoting vascular invasion and osteoblast activation. Simultaneously, matrix remodeling is facilitated by MMP9 and MMP13 secretion, which prepares the environment for new bone deposition 68 . This form of hybrid ossification, integrating both chondroclasia and direct transdifferentiation, illustrates how antlers achieve rapid bone formation without compromising structural fidelity. Antlers and somatic growth plates exemplify divergent evolutionary strategies. Antlers exhibit spatially and temporally expanded proliferative zones to maximize growth rate, unburdened by mechanical loading except for gravity, as they grow from the antler tip. In contrast, somatic growth plates are optimized for structural precision to withstand mechanical stress, such as body weight. Antlers incorporate vasculature early to support their intense metabolic demands, whereas growth plates maintain avascularity to preserve zonal organization. During ossification, antlers repurpose apoptotic signals into cues for chondrocyte transdifferentiation, while growth plates enforce terminal differentiation and lineage cessation. These distinctions parallel industrial paradigms: growth plates function as assembly lines with compartmentalized processes, whereas antlers operate like continuous flow systems. By minimizing transitional delays, antlers achieve exceptional growth speed and scale, while growth plates prioritize mechanical fidelity and structural precision. The case of antler elongation illustrates how extreme growth rates can evolve through modifications of conserved developmental programs. Classical constraints such as limited proliferation, avascularity, and obligatory chondrocyte apoptosis are relaxed to permit extensive growth, vascular integration, and cell fate plasticity. These adaptations likely evolved in response to selective pressures for rapid regeneration within a brief seasonal window, during which antlers must reach the size that is proportional to the deer body-size (around 1.5 meters in length and up to 30 kilograms in weight, within approximately 90 days in large deer species, such as wapti) 5 , 11 , 69 , 70 . The translational potential of these findings is substantial, with implications for bone grafting and fracture repair. The concept of vascularized cartilage may advance tissue engineering for metabolically demanding structures such as articular cartilage. Moreover, leveraging transdifferentiation pathways could accelerate bone healing or enable therapeutic reprogramming of chondrocytes into osteoblasts for the treatment of osteoarthritis. Future research should investigate the feasibility of replicating these mechanisms in human tissues, particularly in contexts requiring rapid tissue regeneration. Methods Ethical approval The experimental protocols involving deer and mice were approved by the Ethics Committee of Changchun Sci-Tech University (Permit Number: CKARI202109). We have complied with all relevant ethical regulations for animal use. Samples collection The experimental protocols involving deer and mice were approved by the Ethics Committee of Changchun Sci-Tech University (Permit Number: CKARI201911), and all relevant ethical manipulations for animal use were strictly followed. For sampling AGC tissues used in snRNA-seq and snATAC-seq, tissues were collected from three healthy 2-year-old sika deer approximately 30 days of growth after casting of their previous hard antlers. The distal 8 cm portion of the AGC was excised and sectioned sagittally along the longitudinal axis. Five distinguishable tissue layers of the AGC were immediately identified, dissected and processed as previously described 6 . For Stereo-seq analysis, a relevant area of 1 cm × 2 cm was selected, which is the maximum size that can be accommodated by the technique. To fit this size, a first growing antler (approximately 10 days old) from a healthy 1-year-old sika deer was used, allowing the 1 cm × 2 cm section to encompass all five tissue layers along with the underlying bone tissue. Following layer dissection, AGC tissues were embedded in Tissue-Tek OCT Compound (Sakura Finetek), respectively, snap-frozen in liquid-nitrogen–precooled isopentane, and stored at −80 °C prior to Stereo-seq. Cryosections were prepared at a thickness of 10 μm using a cryostat (Leica CM3050 S, Leica Biosystems). Cell dissociation from AGC tissues Each tissue sample was cut into pieces smaller than 1 mm³, transferred to a 50-ml tube, and digested in DMEM (Sigma-Aldrich) containing 100 μg/ml collagenase type I and II (Invitrogen) at 37 °C for 60–100 minutes with gentle shaking. Digestion was terminated by adding 10% fetal bovine serum (Gibco) once more than 50,000 cells were released. The suspension was filtered through a 70-μm strainer and centrifuged at 500 g for 5 minutes at 4 °C, and treated with 1× RBC lysis buffer (Beyotime) to remove red blood cells. Cells were washed with PBS and live cells were counted using an AO/PI staining kit (Beyotime). snRNA-seq and snATAC-seq library preparation and sequencing Single-cell RNA sequencing libraries were prepared following established protocols using the DNBelab C Series Single-Cell Library Prep Set (MGI, catalog #1000021082). The procedure involved several key steps: First, single-cell suspensions were processed through droplet-based encapsulation to partition individual cells. After emulsion breakage, mRNA-captured beads were collected for downstream processing. Subsequent steps included reverse transcription of captured mRNA, followed by cDNA amplification and purification. For single-cell ATAC-seq library construction, we employed the DNBelab C Series Single-Cell ATAC Library Prep Set (MGI, catalog #1000021878). Following the manufacturer's recommendations, we prepared one library per 10,000 viable cells. The library preparation workflow involved several critical steps: First, transposed single-nucleus suspensions were partitioned using droplet-based microfluidics. The encapsulated nuclei then underwent preamplification before emulsion disruption. Subsequent processing included DNA amplification and purification to generate barcoded sequencing libraries. For both RNA and ATAC sequencing libraries, we implemented a standardized quality control pipeline. Library quantification was performed using the Qubit ssDNA Assay Kit (Thermo Fisher Scientific, Q10212) to ensure accurate DNA concentration measurements. This rigorous quality assessment was applied uniformly across all library types prior to sequencing. For both RNA-seq and ATAC-seq analyses, sequencing was performed on DIPSEQ T1 platform at China National GeneBank (CNGB). RNA-seq libraries were sequenced on the DIPSEQ T1 platform (China National GeneBank) using a paired-end 130 bp format, consisting of Read 1 (30 bp; containing 10 bp cell barcode 1, 10 bp cell barcode 2, and 10 bp unique molecular identifier) and Read 2 (100 bp transcript sequence), with an additional 10 bp sample index. ATAC-seq libraries were processed on the BGISEQ-500 platform with 50 bp paired-end reads. Both sequencing approaches implemented stringent quality control measures throughout the experimental workflow to ensure data reliability. snRNA-seq data pre-processing First, PISA (v.1.1) 71 was used to extract bead barcodes and unique molecular identifier (UMI) sequences from raw sequencing data. Next, processed reads were aligned to the reference genome 72 using STAR (v.2.5.3) 73 . To ensure data quality, beads with UMI counts below the set threshold were filtered out, and those representing the same cell were merged. Finally, gene expression profiles for each barcode were quantified using PISA. snATAC-seq data pre-processing Raw sequencing reads were processed using PISA, and then aligned to the genome 72 using BWA (v.0.7.15) 74 to generate BAM files. The BAM files were subsequently processed using bap2 (v.0.6.2) 75 to generate fragment files for each snATAC-seq library. Computational analysis of snRNA-seq Data Gene-barcode expression count matrices were loaded into R (v.4.4.1) and converted to Seurat objects for downstream analysis using the Seurat package (v.4.4.0) 76 . Poor-quality cells, defined as those detecting fewer than 200 genes, and genes detected in fewer than 3 cells, were filtered out. Potential doublets were identified and excluded using DoubletFinder (v.2.0.3) 77 to minimize the impact of technical artifacts on the analysis. Cells with a high mitochondrial gene percentage (>5%) were also excluded, as these may indicate apoptosis or cell lysis. After quality control, a total of 56,254 high-quality barcodes were retained for downstream analysis. Log-normalization was conducted using the ‘NormalizeData’ function, which scaled each cell’s total read count to 10,000. The top 2,000 highly variable genes (HVGs) were identified using the ‘FindVariableFeatures’ function with the ‘vst’ method. Gene expression for each cell was then standardized using the ‘ScaleData’ function. Dimensionality reduction was performed using principal component analysis (PCA). Batch effects across samples were corrected using the Harmony (v.1.2.0) 78 algorithm. The Harmony-generated embeddings were used for clustering by constructing a shared nearest-neighbor (SNN) graph. The clusters identified by the Louvain algorithm were evaluated for the expression patterns of well-known cell-type marker genes. The ‘CellCycleScoring’ function from Seurat was used to assign a cell-cycle score to each cell and predict its phase (G2M, S, or G1). Computational analysis of snATAC-seq Data For the analysis of single-nucleus chromatin accessibility, downstream processing was conducted using the ArchR package (v.1.0.2) 79 , which generated an Arrow file for each sample. All Arrow files were then integrated into an ArchRProject object for further analysis. Genome and gene annotations required for the ArchR workflow were constructed using the ‘createGenomeAnnotation’ and ‘createGeneAnnotation’ functions, respectively. Mitochondrial-derived fragments were excluded, and quality metrics were calculated for each nucleus. Nuclei were excluded if their transcription start site (TSS) enrichment score was less than 1, if their unique fragment count was below 1,000, or if their nucleosome signal score was exceeded 2. Potential doublets for each sample were identified using the ‘addDoubletScores’ function with the following parameters: nTrials = 20, k = 10, knnMethod = ‘UMAP’, dimsToUse = 1:30, and were subsequently removed using the ‘filterDoublets’ function with ‘filterRatio’ set to 1. After quality control, 103,278 nuclei passed the filtration. To assess chromatin accessibility at the gene level, the ‘addGeneScoreMatrix’ function was used to generate gene scores. Dimensionality reduction was performed using iterative latent semantic indexing (LSI) on the TileMatrix via the ‘addIterativeLSI’ function. The top 30 LSI components were selected for batch effect correction across samples using the Harmony algorithm. Cell clusters were then identified by applying the ‘addClusters’ function to the Harmony embeddings, with the following parameters: resolution = 1.2, dimsToUse = 1:30, maxClusters = 50. Activated genes for each cluster were identified using the ‘getMarkerFeatures’ function based on gene scores with the following parameters: useMatrix = ‘GeneScoreMatrix’, bias = c(‘TSSEnrichment’, ‘log10(nFrags)’), and testMethod = ‘wilcoxon’. Cell type annotation for the snATAC-seq clusters was performed based on the same markers used in the snRNA-seq analysis. For visualization, Uniform Manifold Approximation and Projection (UMAP) was performed using the ‘addUMAP’ function. To integrate the snATAC-seq and snRNA-seq datasets with high precision, we performed a constrained integration using the ‘addGeneIntegrationMatrix’ function. Cell types were categorized into four groups: group 1 (AnSCs, AnPCs, proliferative AnPCs, and mural cells), group 2 (chondrocytes), group 3 (endothelial cells, monocytes/macrophages, mast cells, and chondroclasts), and group 4 (hypertrophic chondrocytes). Within each group, cells were further stratified by tissue layers (RM, PC, TZ, CA, and MC) to account for potential biological heterogeneity. The snATAC-seq datasets were then integrated with their corresponding snRNA-seq datasets. During the integration, each cell in the snATAC-seq dataset was assigned a predicted RNA cell barcode, generating an integrated gene expression profile, stored in the ‘GeneIntegrationMatrix’. The ‘addGroupCoverages’ function was used to create pseudo-bulk replicates at the cell type scale. Next, the ‘addReproduciblePeakSet’ function was used to call reproducible and fixed-width peaks (501 bp) within each cell type using MACS2 80 . Peaks were annotated based on their relative position to the nearest gene, and classified into four types: promoter, distal, exonic, and intronic. TF binding motifs for humans were downloaded from the CIS-BP (v.2.0.0) 81 database. Peaks containing TF binding sites were annotated using the ‘addMotifAnnotations’ function. The activity of each TF motif enriched in individual cells was calculated using the ‘addDeviationsMatrix’ function, and the results were stored in the ‘MotifMatrix’. Stereo-seq chip preparation Tissue sections were placed on the Stereo-seq chip (BGI) and incubated at 37°C for 3 minutes. The sections were then fixed in methanol at -20°C for 30 minutes before Stereo-seq preparation. The tissue sections were washed with 0.1x SSC buffer (Thermo) containing RNase inhibitor (NEB), then permeabilized with 0.1% pepsin (Sigma) in 0.01 M HCl at 37°C for 5 minutes. RNA was captured by DNBs and reverse transcribed overnight at 42°C using SuperScript II (Invitrogen) with appropriate reagents. After reverse transcription, tissue sections were washed and digested with Tissue Removal buffer at 55°C for 10 minutes. cDNA was released and purified using VAHTS™ DNA Clean Beads (0.8×). The cDNA was amplified using KAPA HiFi HotStart ReadyMix (Roche) with cDNA-PCR primer. PCR conditions were as follows: 95°C for 5 minutes, 15 cycles of 98°C for 20 seconds, 58°C for 20 seconds, 72°C for 3 minutes, and a final extension at 72°C for 5 minutes. Stereo-seq library preparation and sequencing The concentration of amplified cDNA was measured using the Qubit™ dsDNA Assay Kit (Thermo Fisher). A total of 20 ng of DNA was fragmented using in-house Tn5 transposase at 55 °C for 10 minutes. The reaction was stopped by adding 0.02% SDS and gently mixed at 37 °C for 5 minutes. For library amplification, 25 μl of the fragmentation product was combined with 1× KAPA HiFi HotStart ReadyMix, 0.3 mM Stereo-seq-Library-F primer, and 0.3 mM Stereo-seq-Library-R primer, and nuclease-free water to a total volume of 100 μl. The PCR conditions were: 95 °C for 5 minutes; 13 cycles of 98 °C for 20 seconds, 58 °C for 20 seconds, and 72 °C for 30 seconds; and a final extension at 72 °C for 5 minutes. The final PCR products were purified using AMPure XP beads at bead-to-sample ratios of 0.63× and 0.153×, used for DNA nanoball (DNB) generation, and sequenced on the MGI DNBSEQ-Tx platform (Novogene). Stereo-seq data processing and annotation The FASTQ files generated by the sequencer were processed using the publicly available software suite SAW (v.7.1), accessible at https://github.com/STOmics/SAW. We followed the standard workflow provided by SAW to generate spatial gene expression matrices in GEF format. The raw spatial expression matrix was converted into bins of 100 × 100 DNBs, each representing approximately one cell. This resulted in the detection of a median of 5,430 mRNA molecules and 1,166 genes per spot. The spatial transcriptomics profile was analyzed using the Seurat framework. Initially, normalization and variance stabilization of molecular counts were performed with the SCTransform v2 method, introducing the "SCT" array. PCA was then conducted, and the first 30 principal components were selected for cluster identification using the Louvain algorithm, with a resolution parameter of 0.935. Gene set activity scoring To assess the activity of specific gene sets in the snRNA-seq dataset, we employed the AUCell (v.1.24.0) 82 package to calculate the area under the curve (AUC) score for each cell. For the snATAC-seq dataset, gene set activity scores were calculated using the ‘addModuleScore’ function in the ArchR package. Cell-Cell communication inference To investigate key cell-cell communication signals associated with the proliferation of Proliferative AnPCs, we used the CellChat (v.2.1.2) 83 toolkit to construct cell–cell communication networks based on the snRNA-seq and spatial transcriptomics datasets independently. Communication probabilities between cell types were computed using the ‘computeCommunProb’ function. To visualize key signaling interactions, the netVisual_bubble() function was applied to display significant LR pairs mediating communication from 1) AnSCs to Proliferative AnPCs and 2) Proliferative AnPCs to themselves (autocrine signaling). To explore the dynamics of cell–cell communication in the vascular microenvironment provided by antler stem cell-derived cells, we applied CellChat to construct intercellular signaling networks based on the snRNA-seq dataset. We compared communication probabilities between activated and quiescent subpopulations within endothelial and mural cells across all significant LR pairs. The delta probability was computed and visualized as a dot plot. Identification of enriched TFs in proliferative and non-proliferative AnPCs To identify TFs with potential regulatory roles in proliferative versus non-proliferative AnPCs, we first performed differential analysis of TF motif activity and TF expression using the ‘getMarkerFeatures’ function. Specifically, TF motif activity was assessed by setting useMatrix = ‘MotifMatrix’, while TF expression was evaluated by setting useMatrix = ‘GeneIntegrationMatrix’. TFs were considered enriched in proliferative AnPCs if they exhibited a log 2 fold change (log 2 FC) in expression ≥ 1 and an AUC score ≥ 0.6. Conversely, TFs enriched in non-proliferative AnPCs were identified if their log 2 FC ≤ −1 and AUC ≤ 0.4. Next, we sought to identify downstream target genes of the enriched TFs by integrating chromatin accessibility and gene expression data. Gene expression imputation was performed using the ‘imputeMatrix’ function with the Markov Affinity-based Graph Imputation of Cells (MAGIC) algorithm. Differentially accessible peaks between proliferative and non-proliferative AnPCs were identified using the ‘getMarkerFeatures’ function (with useMatrix = "GeneIntegrationMatrix"), applying thresholds of log 2 FC ≥ 1 and p < 0.05 for proliferative AnPCs, and log 2 FC ≤ −1 and p < 0.05 for non-proliferative AnPCs. A TF was considered functionally linked to a downstream gene if the corresponding motif was able to bind to peaks (located in the promoter or distal regulatory regions) that were differentially accessible in the relevant AnPC population, and if the TF-target gene pair exhibited a significant positive correlation in gene expression (Pearson’s r > 0.3, p < 0.05). To gain insights into the biological functions associated with the enriched TFs in proliferative AnPCs and their potential targets, we performed functional annotation using the DAVID (v.2023q4) database 84 . Finally, we used Cytoscape (v.3.9.0) 85 to construct a network visualizing TF-target regulatory relationships and pathway-gene annotations. Trajectory analysis Cellular trajectories were independently inferred using Monocle3 (v1.3.4) 86 and StaVia (Via 2.0) 87 . Monocle3 applied principal graph learning on the UMAP embedding to order cells along lineage progression, while StaVia used higher-order lazy-teleporting random walks with memory to capture both local transitions and global topology. Fine-grained vector fields generated by StaVia were visualized using the ‘via_streamplot’ function to depict the flow of cellular state transitions. Finally, Pearson correlation was calculated between the pseudotime values inferred by Monocle3 and StaVia across all shared cells. Processing of published datasets The scRNA-seq expression profiles of human adult intervertebral discs (ID) were obtained from the Gene Expression Omnibus (GEO) database (accession number: GSE160756) 88 . Cells with fewer than 300 or more than 4,000 detected genes, fewer than 2,000 or more than 20,000 UMIs, and apoptotic or lysed cells (defined by high mitochondrial gene percentage > 5% or high ribosomal gene percentage > 40%) were removed. Normalization and identification of highly variable features were performed. Dimensionality reduction was performed with PCA, followed by batch effect correction using Harmony based on the first 50 principal components. UMAP visualization was performed, and clustering was conducted with a resolution of 0.3. Cell type annotation was performed using markers from the original publication. The scRNA-seq datasets of human embryonic long bones (EL) were downloaded from the GEO database (accession number: GSE143753) 89 . Low-quality cells (fewer than 300 or more than 6,000 detected genes, fewer than 2,000 or more than 40,000 UMIs) and apoptotic or lysed cells (mitochondrial gene percentage > 5% or ribosomal gene percentage > 60%) were excluded. Normalization and identification of highly variable features were conducted, followed by dimensionality reduction using PCA and batch effect correction with Harmony based on the first 50 principal components. UMAP was used for visualization, and clustering analysis was performed at a resolution of 0.13. Cell type annotation was performed using markers from the original publication. The scRNA-seq datasets of osteosarcoma (Os) were acquired from the GEO database (accession number: GSE152048) 90 . Poor-quality nuclei (fewer than 300 or more than 8,000 detected genes, fewer than 2,000 or more than 50,000 UMIs) and apoptotic or lysing cells (high mitochondrial gene percentage > 10% or ribosomal gene percentage > 40%) were removed. Normalization and identification of highly variable features were performed. Dimensionality reduction was done using PCA, followed by batch effect correction using Harmony based on the first 50 principal components. UMAP visualization and clustering analysis were performed with a resolution of 0.15. Cell type annotation was based on markers from the original publication. Comparison of cell types in published datasets with AnSC-derived cell types We utilized ScPred V1.9.2 91 which employs all principal components as gene feature space, to train the classifiers using AnSC-derived cells as references. By default, prediction models use a support vector machine with a radial kernel. Subsequently, we predicted the similar probability of cell types in ID, EL and Os by comparing them with the reference cell types. Definition of activated and quiescent states of vascular cells We used the CytoTRACE2 (v.1.0.0) 92 algorithm to characterize the cellular differentiation potential of endothelial and mural cells based on gene expression profiles, with higher values indicating a more progenitor-like, transcriptionally active state. These cells were then stratified into activated and quiescent states using a fixed threshold of 0.3, determined from the score distribution across regions. Cells with scores above the threshold were classified as activated state, while those below were assigned to the quiescent state. Identification of enriched TFs in activated and quiescent states of both endothelial and mural cells To identify enriched TFs in activated and quiescent states of both endothelial and mural cells, we first integrated the snATAC-seq and snRNA-seq datasets using the ‘addGeneIntegrationMatrix’ function in the ArchR package. Each cell in the snATAC-seq dataset was assigned the label of its most transcriptionally similar neighbor in the snRNA-seq dataset, thereby classifying endothelial and mural cells in the snATAC-seq dataset into activated and quiescent states. The enriched TFs were identifined through integrated transcriptional and chromatin accessibility analyses. For each TF, differential gene expression between activated and quiescent states was assessed using the ‘FindMarkers’ function in the Seurat package, while differential chromatin accessibility was calculated using the ‘getMarkerFeatures’ function in the ArchR package (with >useMatrix = "GeneScoreMatrix"). TFs showing positive log₂FC in both RNA expression and ATAC accessibility were designated as activated-state-enriched TFs, whereas those with negative log₂FC in both modalities were considered enriched in the quiescent state. Downstream target gene prediction for enriched TFs was performed as described in the “Identification of enriched TFs in proliferative and non-proliferative AnPCs” section. Dual-comparison differential expression analysis for AGC vascular cells To investigate the transcriptomic uniqueness of vascular cells within AGC, we performed differential gene expression analyses comparing antler endothelial and mural cells with their corresponding cell types in human adult intervertebral discs (ID) and osteosarcomas (Os). Independent comparisons for each cell type were performed using the ‘FindMarkers’ function in Seurat, with default parameters. Genes with an average log₂FC > 0.25 and p < 0.05 were considered differentially expressed. The genes were categorized into four groups: 1) Upregulated in AGC relative to both ID and Os (double-positive); 2) Downregulated in AGC relative to both ID and Os (double-negative); 3) Upregulated compared to Os but downregulated compared to ID; 4) Upregulated compared to ID but downregulated compared to Os. To interpret the biological significance of genes specifically upregulated or downregulated in antler-derived vascular cells, we performed KEGG enrichment analysis for double-positive and double-negative genes sets separately using the DAVID database. Pathways with p < 0.05 were considered significantly enriched. To infer the directionality of biological changes, Z-scores were calculated for each enriched term as follows: , where and refer to the number of upregulated and downregulated genes, respectively, and N is the total number of genes annotated to that term 93 . Co-culture of AnSCs and AnPCs and assessment of AnPC viability The THY1⁺RXFP2⁺ AnSCs 94 and TNN⁺TNC⁺ AnPCs 53 were isolated from AGC as previously described. All cells were cultured in Dulbecco's Modified Eagle Medium (DMEM; Gibco) supplemented with 100 U/mL penicillin, 100 μg/mL streptomycin, and 10% fetal bovine serum (FBS; HyClone), in a humidified incubator at 37 °C with 5% CO₂. For the co-culture assay, AnPCs were seeded in the lower chamber, while AnSCs were seeded in the upper chamber of transwell inserts with 0.4 μm pores. The co-culture was maintained for one week under standard culture conditions (37 °C, 5% CO₂). Following co-culture, AnPCs were subjected to cell viability analysis using the Cell Counting Kit-8 (CCK-8, Cat. No. K1018, Apexbio). Viability was determined by measuring absorbance and comparing it to that of untreated control cells. Immunohistochemistry Paraffin-embedded AGC tissue sections (2-3 mm in thickness) were de-parafnized and rehydrated. Endogenous peroxidase was quenched with 3% H 2 O 2 for 5 min. Antigen retrieval was performed by boiling in 10 mM sodium citrate buffer (pH 6.0) for 10 min. The non-specifc binding sites were blocked in PBS plus 10% normal goat serum for 30 min and then incubated with primary antibodies: RUNX2, (1:500, cat. No. 20700-1-AP, Proteintech) and MCAM (1:200, A13927, ABclonal) at 37 °C for 2h. After rinsing with PBS, sections were incubated with secondary antibody for 30 min. After rinsing in PBS, all sections were stained with DAB chromogen reaction solution. Creation of ectopic antlers either with vascularized or with avascularized cartilage on the nude mice foreheads A 9-month-old male sika deer calf was selected prior to pedicle initiation (March) for antler stem cell tissue (antlerogenic periosteum, AP) collection and transplantation into nude mice, aiming to generate two types of ectopic antlers: one comprising vascularized cartilage (VC-antler) in red fluorescent protein (RFP)-expressing nude mice, and the other comprising avascular cartilage (AC-antler) in conventional nude mice. The AP from one antler was divided into six pieces for VC-antler induction, and the AP from the contralateral side was similarly divided for AC-antler formation. Detailed surgical procedures were described in our previous studies 47,95,96 . Briefly, to induce VC-antlers, the periosteum at the implantation site on each mouse skull was completely removed prior to subcutaneous implantation of the AP. In contrast, to induce AC-antlers, the periosteum was left intact. In the absence of host periosteum, the implanted AP fused more effectively with the underlying periosteum-free skull, leading to the formation of larger ectopic antlers with vascularized cartilage. When the host periosteum was preserved, fusion was inhibited, resulting in smaller ectopic antlers with avascular cartilage. Immunofluorescence staining Frozen tissue samples were cryosectioned at a thickness of 10 μm, thawed, and washed with PBS. To block non-specific binding, sections were incubated for 1 hour in PBS containing 10% goat serum and 0.3% Triton X-100. Primary antibodies (CD31, 1:1000, ab281583, Abcam; α-Smooth Muscle Actin, 1:500, ab7817, Abcam; BGLAP, 1:200, A20800, ABclonal; SP7, 1:500, ab209484, Abcam) were diluted in PBS and applied to the sections overnight at 4 °C. After washing three times with PBS, sections were incubated with secondary antibodies for 1 hour at room temperature. Sections were then washed with PBS for 3 minutes and counterstained with DAPI (1:500) for 5 minutes. Fluorescence images were acquired using an EVOS M5000 microscope (Thermo Fisher). Statistical and Reproducibility The statistical methods used are indicated in the figure legends. All calculations and visualizations were performed in R. No statistical methods were used to predetermine the sample size, and the experiment was not randomized. Declarations Acknowledgement This project was supported by National Natural Science Foundation of China (Grant No.: U23A20523, 32470892 and 32370899), Natural Science Foundation of Jilin Province (Grant No.: 20240602030RC), Shenzhen Science and Technology Program (Grant No. RCJC20221008092804002) and Jilin Merit Aid Study Abroad Programs (2024). Competing interests The authors declare that they have no competing interests. References Goss R (1963) In: Sognnaes RF (ed) Mechanism of Hard Tissue Destruction. American Association for the Advancement of Science, pp 339–369 Goss RJ (1983) Deer Antlers. Regeneration, Function and Evolution. Academic Li C, Yang F, Sheppard A (2009) Adult stem cells and mammalian epimorphic regeneration-insights from studying annual renewal of deer antlers. Curr Stem Cell Res Therapy 4:237–251. 10.2174/157488809789057446 Kierdorf U, Kierdorf H, Szuwart T (2007) Deer antler regeneration: cells, concepts, and controversies. J Morphol 268:726–738. 10.1002/jmor.10546 Price J, Allen S (2004) Exploring the mechanisms regulating regeneration of deer antlers. Philos Trans R Soc Lond B Biol Sci 359:809–822. 10.1098/rstb.2004.1471 Li C, Clark DE, Lord EA, Stanton JA, Suttie JM (2002) Sampling technique to discriminate the different tissue layers of growing antler tips for gene discovery. Anat Rec 268:125–130 Sissons HA, Kember NF (1977) Longitudinal bone growth of the human femur. Postgrad Med J 53:433–437. 10.1136/pgmj.53.622.433 Clark DE, Lord EA, Suttie JM (2006) Expression of VEGF and pleiotrophin in deer antler. The anatomical record. Part A, Discoveries in molecular, cellular, and evolutionary biology 288, 1281–1293. 10.1002/ar.a.20393 Clark DE, Li C, Wang W, Martin SK, Suttie JM (2006) Vascular localization and proliferation in the growing tip of the deer antler. The anatomical record. Part A, Discoveries in molecular, cellular, and evolutionary biology 288, 973–981. 10.1002/ar.a.20364 Banks JWN (1983) In: Banks RD (ed) W. J. in Antler development in Cervidae. Caesar Kleburg Wildlife Research Institute, pp 279–306 Landete-Castillejos T et al (2019) Antlers - Evolution, development, structure, composition, and biomechanics of an outstanding type of bone. Bone 128:115046. 10.1016/j.bone.2019.115046 Ba H et al (2024) Insights into the molecular characteristics of embryonic cranial neural crest cells and their derived mesenchymal cell pools. Commun Biology 7:1347. 10.1038/s42003-024-07056-x Ba H, Wang D, Yau TO, Shang Y, Li C (2019) Transcriptomic analysis of different tissue layers in antler growth Center in Sika Deer (Cervus nippon). BMC Genomics 20:173. 10.1186/s12864-019-5560-1 Li C et al (2023) Deer antlers: the fastest growing tissue with least cancer occurrence. Cell Death Differ 30:2452–2461. 10.1038/s41418-023-01231-z Ba H et al (2022) Single-cell transcriptome reveals core cell populations and androgen-RXFP2 axis involved in deer antler full regeneration. Cell regeneration (London England) 11:43. 10.1186/s13619-022-00153-4 Wang D et al (2019) Deer antler stem cells are a novel type of cells that sustain full regeneration of a mammalian organ-deer antler. Cell Death Dis 10:443. 10.1038/s41419-019-1686-y Addison WN, Nakano Y, Loisel T, Crine P, McKee MD (2008) MEPE-ASARM peptides control extracellular matrix mineralization by binding to hydroxyapatite: an inhibition regulated by PHEX cleavage of ASARM. J bone mineral research: official J Am Soc Bone Mineral Res 23:1638–1649. 10.1359/jbmr.080601 Zhang X et al (2017) Analysis of novel RUNX2 mutations in Chinese patients with cleidocranial dysplasia. PLoS ONE 12:e0181653. 10.1371/journal.pone.0181653 Jung YJ, Bae HS, Ryoo HM, Baek SH (2018) A novel RUNX2 mutation in exon 8, G462X, in a patient with Cleidocranial Dysplasia. J Cell Biochem 119:1152–1162. 10.1002/jcb.26283 Zeng L et al (2017) Functional analysis of novel RUNX2 mutations in cleidocranial dysplasia. Mutagenesis 32:437–443. 10.1093/mutage/gex012 Shih IM, Elder DE, Speicher D, Johnson JP, Herlyn M (1994) Isolation and functional characterization of the A32 melanoma-associated antigen. Cancer Res 54:2514–2520 Bardin N et al (1996) S-Endo 1, a pan-endothelial monoclonal antibody recognizing a novel human endothelial antigen. Tissue Antigens 48:531–539. 10.1111/j.1399-0039.1996.tb02666.x McAlinden A (2014) Alternative splicing of type II procollagen: IIB or not IIB? Connect Tissue Res 55:165–176. 10.3109/03008207.2014.908860 Li C, Pearson A, McMahon C (2013) Morphogenetic mechanisms in the cyclic regeneration of hair follicles and deer antlers from stem cells. BioMed research international 643601. 10.1155/2013/643601 (2013) Vortkamp A et al (1998) Recapitulation of signals regulating embryonic bone formation during postnatal growth and in fracture repair. Mech Dev 71:65–76. 10.1016/s0925-4773(97)00203-7 Xing H, Wang Q, Ma Y, Han R, Li H (2024) The significance of MDK growth factor in the antler development of sika deer (Cervus nippon): An in-depth analysis. Gene Expr Patterns 119388. 10.1016/j.gep.2024.119388 Dong Z, Li C, Coates D (2021) PTN-PTPRZ signalling is involved in deer antler stem cell regulation during tissue regeneration. J Cell Physiol 236:3752–3769. 10.1002/jcp.30115 Ishimoto H et al (2006) Midkine, a heparin-binding growth factor, selectively stimulates proliferation of definitive zone cells of the human fetal adrenal gland. J Clin Endocrinol Metab 91:4050–4056. 10.1210/jc.2006-1139 Sinha S et al (2022) Fibroblast inflammatory priming determines regenerative versus fibrotic skin repair in reindeer. Cell 185:4717–4736e4725. 10.1016/j.cell.2022.11.004 Colitti M, Allen SP, Price JS (2005) Programmed cell death in the regenerating deer antler. J Anat 207:339–351. 10.1111/j.1469-7580.2005.00464.x Dong Z, Coates D (2021) Bioactive Molecular Discovery Using Deer Antlers as a Model of Mammalian Regeneration. J Proteome Res 20:2167–2181. 10.1021/acs.jproteome.1c00003 Calamari ZT, Flynn JJ (2024) Gene expression supports a single origin of horns and antlers in hoofed mammals. Commun Biol 7:509. 10.1038/s42003-024-06134-4 Soldatov R et al (2019) Spatiotemporal structure of cell fate decisions in murine neural crest. Sci (New York N Y) 364. 10.1126/science.aas9536 Bechtold TE et al (2016) Osteophyte formation and matrix mineralization in a TMJ osteoarthritis mouse model are associated with ectopic hedgehog signaling. Matrix biology: J Int Soc Matrix Biology 52–54:339–354. 10.1016/j.matbio.2016.03.001 Colnot Cl et al (2005) Indian hedgehog synchronizes skeletal angiogenesis and perichondrial maturation with cartilage development. Development 132, 1057–1067. 10.1242/dev.01649%J Development Ortega N, Wang K, Ferrara N, Werb Z, Vu TH (2010) Complementary interplay between matrix metalloproteinase-9, vascular endothelial growth factor and osteoclast function drives endochondral bone formation. Disease Models & Mechanisms 3, 224–235. 10.1242/dmm.004226%J Disease Models & Mechanisms Wolff LI, Hartmann C (2019) A Second Career for Chondrocytes-Transformation into Osteoblasts. Curr Osteoporos Rep 17:129–137. 10.1007/s11914-019-00511-3 Zhou X et al (2014) Chondrocytes transdifferentiate into osteoblasts in endochondral bone during development, postnatal growth and fracture healing in mice. PLoS Genet 10:e1004820. 10.1371/journal.pgen.1004820 Komori T (2022) Whole Aspect of Runx2 Functions in Skeletal Development. Int J Mol Sci 23. 10.3390/ijms23105776 Qin X et al (2020) Runx2 is essential for the transdifferentiation of chondrocytes into osteoblasts. PLoS Genet 16:e1009169. 10.1371/journal.pgen.1009169 Sinha KM, Zhou X (2013) Genetic and molecular control of osterix in skeletal formation. J Cell Biochem 114:975–984. 10.1002/jcb.24439 Hojo H, Ohba S (2020) Gene regulatory landscape in osteoblast differentiation. Bone 137:115458. https://doi.org/10.1016/j.bone.2020.115458 Rehn AP, Cerny R, Sugars RV, Kaukua N, Wendel M (2008) Osteoadherin is upregulated by mature osteoblasts and enhances their in vitro differentiation and mineralization. Calcif Tissue Int 82:454–464. 10.1007/s00223-008-9138-1 Luo G et al (1997) Spontaneous calcification of arteries and cartilage in mice lacking matrix GLA protein. Nature 386:78–81. 10.1038/386078a0 Bianco P, Fisher LW, Young MF, Termine JD, Robey PG (1991) Expression of bone sialoprotein (BSP) in developing human tissues. Calcif Tissue Int 49:421–426. 10.1007/bf02555854 Li C et al (2009) Development of a nude mouse model for the study of antlerogenesis–mechanism of tissue interactions and ossification pathway. J Exp Zool B Mol Dev Evol 312:118–135. 10.1002/jez.b.21252 Wang D et al (2022) Chimeric blood vessels sustained development of the xenogeneic antler: a unique model for xenogeneic organ generation. Life Med 2. 10.1093/lifemedi/lnac021 Zheng Y, Wang G, Chen R, Hua Y, Cai Z (2018) Mesenchymal stem cells in the osteosarcoma microenvironment: their biological properties, influence on tumor growth, and therapeutic implications. Stem Cell Res Ther 9:22. 10.1186/s13287-018-0780-x Niu Z et al (2022) Vascular Calcification: New Insights Into BMP Type I Receptor A. Front Pharmacol 13:887253. 10.3389/fphar.2022.887253 Waltenberger J, Mayr U, Pentz S, Hombach V (1996) Functional upregulation of the vascular endothelial growth factor receptor KDR by hypoxia. Circulation 94:1647–1654. 10.1161/01.cir.94.7.1647 Fong GH, Rossant J, Gertsenstein M, Breitman ML (1995) Role of the Flt-1 receptor tyrosine kinase in regulating the assembly of vascular endothelium. Nature 376:66–70. 10.1038/376066a0 Wang D, Landete-Castillejos T (2023) Stem cells drive antler regeneration. Sci (New York N Y) 379:757–758. 10.1126/science.adg9968 Qin T et al (2023) A population of stem cells with strong regenerative potential discovered in deer antlers. Science 379:840–847. 10.1126/science.add0488 Hellberg C, Ostman A, Heldin CH (2010) PDGF and vessel maturation. Recent results cancer Res Fortschr der Krebsforschung Progres dans les recherches sur le cancer 180:103–114. 10.1007/978-3-540-78281-0_7 Ocaña OH et al (2012) Metastatic Colonization Requires the Repression of the Epithelial-Mesenchymal Transition Inducer Prrx1. Cancer Cell 22:709–724. 10.1016/j.ccr.2012.10.012 Suttie JM et al (1985) Insulin-like growth factor 1 (IGF-1) antler-stimulating hormone? Endocrinology 116, 846–848. 10.1210/endo-116-2-846 Li C, Littlejohn RP, Suttie JM (1999) Effects of insulin-like growth factor 1 and testosterone on the proliferation of antlerogenic cells in vitro. J Exp Zool 284:82–90 Sadighi M, Haines SR, Skottner A, Harris AJ, Suttie JM (1994) Effects of insulin-like growth factor-I (IGF-I) and IGF-II on the growth of antler cells in vitro. J Endocrinol 143:461–469. 10.1677/joe.0.1430461 Suttie JM, Fennessy PF, Gluckman PD, Corson ID (1988) Elevated plasma IGF 1 levels in stags prevented from growing antlers. Endocrinology 122:3005–3007. 10.1210/endo-122-6-3005 Wilsman NJ, Van Sickle DC (1970) The relationship of cartilage canals to the initial osteogenesis of secondary centers of ossification. 168:381–391. https://doi.org/10.1002/ar.1091680305 Brighton CT (1984) The growth plate. Qld Gov Min J 15:571–595 Vortkamp A et al (1996) Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Sci (New York N Y) 273:613–622. 10.1126/science.273.5275.613 St-Jacques B, Hammerschmidt M, McMahon AP (1999) Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes Dev 13:2072–2086. 10.1101/gad.13.16.2072 Faucheux C et al (2004) Recapitulation of the parathyroid hormone-related peptide-Indian hedgehog pathway in the regenerating deer antler. Dev Dyn 231:88–97 Stockwell R (1979) Biology of Cartilage Cells. (Cambridge University Magne D et al (2005) Cartilage formation in growth plate and arteries: from physiology to pathology. BioEssays: news reviews Mol Cell Dev biology 27:708–716. 10.1002/bies.20254 Wuelling M, Vortkamp A (2011) Chondrocyte proliferation and differentiation. Endocr Dev 21:1–11. 10.1159/000328081 Yang MC, Nakamura M, Kageyama Y, Igari Y, Sasano Y (2023) Age-Related Gene and Protein Expression in Mouse Mandibular Condyle Analyzed by Cap Analysis of Gene Expression and Immunohistochemistry. Gerontology 69:1295–1306. 10.1159/000533921 Goss RJ (1970) Problems of antlerogesis. Clin Orthop Relat Res 69:227–238 Li C et al (2024) Bone metabolism associated with annual antler regeneration: a deer insight into osteoporosis reversal. Biol Direct 19:123. 10.1186/s13062-024-00561-3 Shi Q, Liu S, Kristiansen K, Liu L (2022) The FASTQ + format and PISA. Bioinformatics 38:4639–4642. 10.1093/bioinformatics/btac562 Ba H et al (2020) Chromosome-level genome assembly of Tarim red deer, Cervus elaphus yarkandensis. Sci data 7:187. 10.1038/s41597-020-0537-0 Dobin A et al (2013) STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29:15–21. 10.1093/bioinformatics/bts635 Li H (2013) Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. arXiv: Genomics Lareau CA et al (2019) Droplet-based combinatorial indexing for massive-scale single-cell chromatin accessibility. Nat Biotechnol 37:916–924. 10.1038/s41587-019-0147-6 Hao Y et al (2021) Integrated analysis of multimodal single-cell data. Cell 184:3573–3587e3529. 10.1016/j.cell.2021.04.048 McGinnis CS, Murrow LM, Gartner ZJ, DoubletFinder (2019) Doublet Detection in Single-Cell RNA Sequencing Data Using Artificial Nearest Neighbors. Cell Syst 8:329–337e324. 10.1016/j.cels.2019.03.003 Korsunsky I et al (2019) Fast, sensitive and accurate integration of single-cell data with Harmony. Nat Methods 16:1289–1296. 10.1038/s41592-019-0619-0 Granja JM et al (2021) ArchR is a scalable software package for integrative single-cell chromatin accessibility analysis. Nat Genet 53:403–411. 10.1038/s41588-021-00790-6 Zhang Y et al (2008) Model-based analysis of ChIP-Seq (MACS). Genome biology 9, R137. 10.1186/gb-2008-9-9-r137 Weirauch MT et al (2014) Determination and inference of eukaryotic transcription factor sequence specificity. Cell 158:1431–1443. 10.1016/j.cell.2014.08.009 Aibar S et al (2017) SCENIC: single-cell regulatory network inference and clustering. Nat Methods 14:1083–1086. 10.1038/nmeth.4463 Jin S, Plikus MV, Nie Q (2025) CellChat for systematic analysis of cell-cell communication from single-cell transcriptomics. Nat Protoc 20:180–219. 10.1038/s41596-024-01045-4 Sherman BT, Panzade G, Imamichi T, Chang W (2024) DAVID Ortholog: an integrative tool to enhance functional analysis through orthologs. Bioinformatics 40. 10.1093/bioinformatics/btae615 Shannon P et al (2003) Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 13:2498–2504. 10.1101/gr.1239303 Cao J et al (2019) The single-cell transcriptional landscape of mammalian organogenesis. Nature 566:496–502. 10.1038/s41586-019-0969-x Stassen SV et al (2024) StaVia: spatially and temporally aware cartography with higher-order random walks for cell atlases. Genome Biol 25:224. 10.1186/s13059-024-03347-y Gan Y et al (2021) Spatially defined single-cell transcriptional profiling characterizes diverse chondrocyte subtypes and nucleus pulposus progenitors in human intervertebral discs. Bone Res 9. 10.1038/s41413-021-00163-z He J et al (2021) Dissecting human embryonic skeletal stem cell ontogeny by single-cell transcriptomic and functional analyses. Cell Res 31:742–757. 10.1038/s41422-021-00467-z Zhou Y et al (2020) Single-cell RNA landscape of intratumoral heterogeneity and immunosuppressive microenvironment in advanced osteosarcoma. Nat Commun 11:6322. 10.1038/s41467-020-20059-6 (!!! INVALID CITATION !!! 61) Kang M et al (2024) Mapping single-cell developmental potential in health and disease with interpretable deep learning. bioRxiv: preprint Serv biology. 10.1101/2024.03.19.585637 Walter W, Sánchez-Cabo F, Ricote M (2015) GOplot: an R package for visually combining expression data with functional analysis. Bioinformatics 31:2912–2914. 10.1093/bioinformatics/btv300 Ba H et al (2025) RXFP2-positive mesenchymal stem cells in the antlerogenic periosteum contribute to postnatal development of deer antlers. Commun Biol 8:645. 10.1038/s42003-025-08085-w Li C, Suttie JM (2003) Tissue collection methods for antler research. Eur J Morphol 41:23–30. 10.1076/ejom.41.1.23.28106 Li C, Harris AJ, Suttie JM (2001) Tissue interactions and antlerogenesis: new findings revealed by a xenograft approach. J Exp Zool 290:18–30 Additional Declarations There is NO Competing Interest. Supplementary Files suppl.docx SUPPLEMENTARY INFORMATION Cite Share Download PDF Status: Under Review 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 Advisory Board 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-6919532","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Biological Sciences - Article","associatedPublications":[],"authors":[{"id":475057823,"identity":"888f0320-fbf3-40b9-9cc1-06e95847d64c","order_by":0,"name":"Hengxing Ba","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/ElEQVRIiWNgGAWjYBACewYeMA0kmQ9AxRLwazFsgGthS4CqJqDF4AAPjMljQJwWw/azBx8X/DosY86/5uPHnz8OM/Cz5xgw/NyBW4s9T16y8cy+NB7LGW83S/MkHGaQ7HljwNh7Bp9fcsykeXtseAxunN3GzADUYnAjx4CZsQ2PX86/Mf/N2yMB1HLmGeMPoBZ7glpu5Jgx8/wA2nK+h40B5DADCQJaDGe8MZbmbUgD2sJmLM2Tls4jceZZwcFePFrs+XMMP/P8OWxvcP7ww48/bKzl+NuTNz74iUcLGICdIZEAZoOj6QABDUDwB4j5iVA3CkbBKBgFIxMAAN6mULhc5GYXAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0003-0882-8841","institution":"Institute of Antler Science and Product Technology, Changchun Sci-Tech University","correspondingAuthor":true,"prefix":"","firstName":"Hengxing","middleName":"","lastName":"Ba","suffix":""},{"id":475057824,"identity":"a8bed784-ae34-4ca5-b665-bc4b5f28bbc0","order_by":1,"name":"Shidian He","email":"","orcid":"","institution":"BGI Research","correspondingAuthor":false,"prefix":"","firstName":"Shidian","middleName":"","lastName":"He","suffix":""},{"id":475057825,"identity":"1275632f-fc71-49ad-93f5-2e6cb99be657","order_by":2,"name":"Hai-Xi Sun","email":"","orcid":"https://orcid.org/0000-0002-8521-0554","institution":"University of Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Hai-Xi","middleName":"","lastName":"Sun","suffix":""},{"id":475057826,"identity":"7f4ea105-2ac6-4f19-84fc-0cbf85a7ec19","order_by":3,"name":"Xin Wang","email":"","orcid":"","institution":"BGI-Shenzhen","correspondingAuthor":false,"prefix":"","firstName":"Xin","middleName":"","lastName":"Wang","suffix":""},{"id":475057827,"identity":"55055513-223f-4be4-9433-5cd65753db71","order_by":4,"name":"Zhang Hang","email":"","orcid":"","institution":"Institute of Special Economic Animals and Plants, Chinese Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Zhang","middleName":"","lastName":"Hang","suffix":""},{"id":475057828,"identity":"89813a7c-1cf0-4834-8205-99aa92d93b61","order_by":5,"name":"Guo Qianchi","email":"","orcid":"","institution":"Institute of Antler Science and Product Technology, Changchun Sci-Tech University","correspondingAuthor":false,"prefix":"","firstName":"Guo","middleName":"","lastName":"Qianchi","suffix":""},{"id":475057829,"identity":"084626f5-7ee3-4b1f-af25-60f337c0dc46","order_by":6,"name":"Qiuting Deng","email":"","orcid":"","institution":"BGI-Shenzhen","correspondingAuthor":false,"prefix":"","firstName":"Qiuting","middleName":"","lastName":"Deng","suffix":""},{"id":475057830,"identity":"78fb7b79-02bf-4870-b3b3-696e8f7fd2d6","order_by":7,"name":"Yue Yuan","email":"","orcid":"","institution":"BGI-Shenzhen","correspondingAuthor":false,"prefix":"","firstName":"Yue","middleName":"","lastName":"Yuan","suffix":""},{"id":475057831,"identity":"de684864-a829-41ee-880f-1805374d1500","order_by":8,"name":"Chang Liu","email":"","orcid":"","institution":"BGI-Shenzhen","correspondingAuthor":false,"prefix":"","firstName":"Chang","middleName":"","lastName":"Liu","suffix":""},{"id":475057832,"identity":"0ca54e0b-8c78-4bf6-8164-8d113b1b3777","order_by":9,"name":"Wang Zhen","email":"","orcid":"","institution":"Institute of Antler Science and Product Technology, Changchun Sci-Tech University","correspondingAuthor":false,"prefix":"","firstName":"Wang","middleName":"","lastName":"Zhen","suffix":""},{"id":475057833,"identity":"714e0d7b-a493-4a63-9b95-0cb3a02298c3","order_by":10,"name":"Jiping Li","email":"","orcid":"","institution":"Institute of Antler Science and Product Technology, Changchun Sci-Tech University","correspondingAuthor":false,"prefix":"","firstName":"Jiping","middleName":"","lastName":"Li","suffix":""},{"id":475057834,"identity":"e2679704-f2bc-4ff4-96d3-95bc39da63d6","order_by":11,"name":"Xie Liuwei","email":"","orcid":"","institution":"Department of Police Dog Technology, Criminal Investigation Police University of China","correspondingAuthor":false,"prefix":"","firstName":"Xie","middleName":"","lastName":"Liuwei","suffix":""},{"id":475057835,"identity":"d460a5b1-5af7-476d-82e7-bdd5415adfc4","order_by":12,"name":"Hu Pengfei","email":"","orcid":"","institution":"Institute of Antler Science and Product Technology, Changchun Sci-Tech University","correspondingAuthor":false,"prefix":"","firstName":"Hu","middleName":"","lastName":"Pengfei","suffix":""},{"id":475057836,"identity":"7b04bb37-5760-4e39-a38c-1920c281954e","order_by":13,"name":"Qianqian Guo","email":"","orcid":"","institution":"Institute of Antler Science and Product Technology, Changchun Sci-Tech University","correspondingAuthor":false,"prefix":"","firstName":"Qianqian","middleName":"","lastName":"Guo","suffix":""},{"id":475057837,"identity":"a43a5c1a-bfab-43ad-a5a0-764196943462","order_by":14,"name":"Zhang Guokun","email":"","orcid":"","institution":"Institute of Antler Science and Product Technology, Changchun Sci-Tech University","correspondingAuthor":false,"prefix":"","firstName":"Zhang","middleName":"","lastName":"Guokun","suffix":""},{"id":475057838,"identity":"222ad47b-1c89-4acf-ae59-eef4961b8be1","order_by":15,"name":"Dawn Coates","email":"","orcid":"","institution":"Sir John Walsh Research Institute, Faculty of Dentistry, University of Otago","correspondingAuthor":false,"prefix":"","firstName":"Dawn","middleName":"","lastName":"Coates","suffix":""},{"id":475057839,"identity":"6efe1c54-cc6e-40cf-9de9-dfbe9b8ba99b","order_by":16,"name":"Ying Gu","email":"","orcid":"https://orcid.org/0000-0001-7822-0570","institution":"BGI Research","correspondingAuthor":false,"prefix":"","firstName":"Ying","middleName":"","lastName":"Gu","suffix":""},{"id":475057840,"identity":"a67fb8e0-90a9-4dbd-94e0-e2b39859bd01","order_by":17,"name":"Chuanyu Liu","email":"","orcid":"https://orcid.org/0000-0003-2258-0897","institution":"BGI Research","correspondingAuthor":false,"prefix":"","firstName":"Chuanyu","middleName":"","lastName":"Liu","suffix":""},{"id":475057841,"identity":"36c942f7-b736-48dc-aaa9-83beeb216fb3","order_by":18,"name":"Datao Wang","email":"","orcid":"https://orcid.org/0000-0002-8372-9133","institution":"Institute of Special Economic Animals and Plants, Chinese Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Datao","middleName":"","lastName":"Wang","suffix":""},{"id":475057842,"identity":"1b7a23fa-8cb8-4d9c-9dc6-77be72304dd0","order_by":19,"name":"Chunyi Li","email":"","orcid":"https://orcid.org/0000-0001-7275-4440","institution":"Changchun Sci-Tech University","correspondingAuthor":false,"prefix":"","firstName":"Chunyi","middleName":"","lastName":"Li","suffix":""}],"badges":[],"createdAt":"2025-06-18 06:01:01","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6919532/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6919532/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":86025516,"identity":"cdc6f99f-e1a6-436e-88aa-02a0644b3940","added_by":"auto","created_at":"2025-07-04 12:56:12","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":808843,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIntegrated snRNA-seq and snATAC-seq analyses of the antler growth center (AGC)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea.\u003c/strong\u003e Schematic illustration showing the anatomical location of the AGC and sampling sites for snRNA-seq, snATAC-seq, and spatial scRNA-seq. The AGC comprises five histologically disguisable layers arranged distoproximally: reserve mesenchyme (RM), pre-cartilage (PC), transition zone (TZ), cartilage (CA), and mineralized cartilage (MC). Note that Stereo-seq samples did not include the dermal region, but included proximal bone tissue.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb.\u003c/strong\u003e UMAP plot of snRNA-seq data showing all cells from the five tissue layers, color-coded for tissue layers.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec.\u003c/strong\u003e UMAP plot of snRNA-seq data showing 10 distinct cell types identified across the five layers.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed.\u003c/strong\u003e UMAP plots displaying the expression patterns of representative marker genes for the 10 identified cell types.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee.\u003c/strong\u003e UMAP plot of snATAC-seq data showing all cells from the five layers, color-coded for tissue layers.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef.\u003c/strong\u003e UMAP plot of snATAC-seq data showing 10 distinct cell types identified across the five layers.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eg.\u003c/strong\u003e UMAP plots showing chromatin accessibility at representative marker gene loci corresponding to the 10 identified cell types.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eh.\u003c/strong\u003e Integrated analysis of snRNA-seq and snATAC-seq data showing the concordance at cell type (left) and tissue level (right).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ei.\u003c/strong\u003e Bar plot to show the relative proportions of cell type from snRNA-seq (left) and snATAC-seq data (right).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ej.\u003c/strong\u003e Heatmap showing Pearson correlation coefficients between cell types identified from snRNA-seq and snATAC-seq data.\u003c/p\u003e","description":"","filename":"Fig.1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6919532/v1/95bcf3828ea552898c2371cc.jpg"},{"id":86025636,"identity":"22cca907-0b4a-4849-9c0a-ab8879bb3a8c","added_by":"auto","created_at":"2025-07-04 13:04:12","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1096264,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSpatial scRNA-seq analysis reveals the cellular architecture of the AGC\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea.\u003c/strong\u003e Spatial map showing the distribution of 11 distinct cell types. Notably, two cell types, PHEX⁺ cells and osteoblasts, were identified in this analysis, while mast cells were absent, likely due to their low abundance. The pie chart shows the percentage of each cell type in the whole cell population. Note that this map does not include the dermal region in the AGC; see Figure 1a for sampling reference.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb.\u003c/strong\u003e Spatial expression patterns of representative marker genes for the 11 cell types.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec.\u003c/strong\u003e UMAP plots showing the expression of RUNX2.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed.\u003c/strong\u003e Spatial map showing the distribution of RUNX2 expression in the AGC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee.\u003c/strong\u003e UMAP plots displaying chromatin accessibility at the RUNX2 locus.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef.\u003c/strong\u003e Immunohistochemical staining showing spatial expression of RUNX2 protein. Scale bar: 50 μm\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eg.\u003c/strong\u003e UMAP plots showing expression of MCAN.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eh.\u003c/strong\u003e Spatial map showing MCAN expression in the AGC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ei.\u003c/strong\u003e UMAP plots displaying chromatin accessibility at the MCAN locus.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ej.\u003c/strong\u003e Immunohistochemical staining showing spatial expression of MCAN protein. Scale bar: 50 μm\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ek.\u003c/strong\u003e Spatial expression patterns of osteoblast and chondroblast markers in proliferative and non-proliferative AnPCs.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003el.\u003c/strong\u003e Mapping of spatial cell-type relationships: Banks et al. (1982) and Li et al (2002) vs. the present study. RMC: Reserve mesenchyme, PCB: Prechondroblasts, CB: Chondroblasts, CC: Chondrocytes, HCC: Hypertrophic Chondrocytes, MCC: Mineralized Chondrocytes, OS: Osteoblasts, IR: inner RM layer, OR: outer RM layer, PC: pre-cartilage, TZ: transition zone, CA: cartilage, MC: mineralized cartilage.\u003c/p\u003e","description":"","filename":"Fig.2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6919532/v1/f7bd4d10d6572badff867d7c.jpg"},{"id":86025518,"identity":"515c4acc-ce5c-4f57-8d11-eb0579115c1d","added_by":"auto","created_at":"2025-07-04 12:56:12","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":512857,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSignaling from AnSCs promotes proliferation of AnPCs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea.\u003c/strong\u003e Spatial distribution of AnSCs, proliferative AnPCs, and non-proliferative AnPCs in the AGC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb.\u003c/strong\u003e Spatial cell cycle phase scores (G2/M and S) for AnSCs, proliferative AnPCs, and non-proliferative AnPCs.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec.\u003c/strong\u003e Spatial expression of IGF1 and its receptor IGF1R, and PTHLH and its receptor PTH1R, in AnSCs and AnPCs.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed.\u003c/strong\u003e Schematic diagram of predicted cell–cell communication between AnSCs and proliferative AnPCs, and among proliferative AnPCs.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee.\u003c/strong\u003e Circle plots showing interaction strength between AnSCs and proliferative AnPCs, and among proliferative AnPCs, based on snRNA-seq (left) and Stereo-seq (right) datasets.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef.\u003c/strong\u003e Bubble plots displaying significant ligand–receptor interactions between AnSCs and proliferative AnPCs, and within AnPCs, based on snRNA-seq (left) and Stereo-seq (right). Key ligand–receptor pairs are highlighted in red.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eg.\u003c/strong\u003e Spatial expression maps of representative ligand–receptor pairs.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eh.\u003c/strong\u003e Schematic of the co-culture experiment of AnSCs with AnPCs. NC: negative control.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ei.\u003c/strong\u003e AnPC proliferation at 12, 24, and 48 h, showing time-dependent increase (n=6). Significant differences were determined using a one-way analysis of variance, followed by Tukey’s post hoc test. ****P \u0026lt; 0.0001, **P \u0026lt; 0.01, *P \u0026lt; 0.05. Error bars: SEM.\u003c/p\u003e","description":"","filename":"Fig.3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6919532/v1/8d7f6105b433e80f31d48429.jpg"},{"id":86025520,"identity":"56e93f10-ddc3-4b91-abca-d3180fe342c0","added_by":"auto","created_at":"2025-07-04 12:56:12","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":624655,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCharacteristics of proliferative AnPCs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea.\u003c/strong\u003e Violin plots showing area under the curve (AUC) scores (top) and module scores (bottom) of key signaling pathways in proliferative versus non-proliferative AnPCs. Significant differences were determined using the Wilcoxon rank-sum test.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb.\u003c/strong\u003e Spatial map showing AnSCs and AnSC-derived cells in the AGC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec.\u003c/strong\u003e Spatial expression patterns of apoptosis-related genes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed.\u003c/strong\u003e Scatter plot of activated transcription factors (TFs) in proliferative and non-proliferative AnPCs. Activated TFs were defined as |log₂FC| ≥ 1 and AUC of motif deviation ≥ 0.6. Top five TFs are labeled.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee.\u003c/strong\u003e Network visualization of enriched TFs in proliferative AnPCs, their predicted targets, and their enriched GO BP terms based on DAVID analysis.\u003c/p\u003e","description":"","filename":"Fig.4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6919532/v1/7910278890624d9e34175d22.jpg"},{"id":86025635,"identity":"a276c6d7-b74b-4739-a20b-f18b2e6cd8cd","added_by":"auto","created_at":"2025-07-04 13:04:12","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1186522,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePHEX⁺ cells contribute to antler cartilage mineralization via Hedgehog (HH) signaling and may represent an intermediate stage of chondrocyte maturation during the transition toward an osteoblast\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eSpatial distribution of chondrocytes, hypertrophic chondrocytes, and PHEX⁺ cells.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eBar plot showing top five GO BP terms enriched in PHEX⁺ cells based on highly expressed genes, including osteoblast differentiation and bone mineralization.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eSpatial expression of representative bone mineralization genes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eCircle plot of predicted cell–cell interactions among chondrocytes, hypertrophic chondrocytes, and PHEX⁺ cells.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eCircle plot showing predicted HH pathway interactions among these cell types.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eSpatial expression patterns of representative HH signaling genes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eg.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eSpatial localization of PHEX⁺ cells and osteoblasts. Cell type localization (left upper) with the region of interest enlarged (left lower). Osteoblasts are categorized into OB-D (Red; located proximal to PHEX⁺ cells) and OB-P (Green: located distally) with no PHEX\u003csup\u003e+\u003c/sup\u003e cells, as illustrated (right).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eh.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eUMAP plot showing distinct clustering of OB-D and OB-P populations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ei.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eUMAP plot of osteoblast marker gene expression.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ej.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eMonocle3-inferred cellular trajectory embedded in UMAP space, showing the developmental progression from chondrocytes and hypertrophic chondrocytes through PHEX⁺ cells toward OB-D. Notably, OB-D cells form a distinct and separate cluster.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ek.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eDifferentiation flow along the StaVia-inferred pseudotime, embedded in the same UMAP space, illustrating the developmental progression from chondrocytes and hypertrophic chondrocytes through PHEX⁺ cells toward OB-D. Notably, two trajectory endpoints are observed: one at hypertrophic chondrocytes and the other at OB-D (indicated by black arrowheads).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003el.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eExpression dynamics of key marker genes along the pseudotime axis. The proportion of AnSC-derived cells along the pseudotime axis is shown by bins (top).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003em.\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eXenogeneic antler formation model with avascular cartilage (AC-antler). hema-toxylin and eosin (H\u0026amp;E)/Alcian blue staining at day 30 revealed a large nodule of avascular cartilage with early signs of ossification (highlighted by black boxes), which showed progressive expansion by day 60.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea.\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp;\u0026nbsp; \u003c/strong\u003eImmunofluorescence analysis of the xenogeneic antler formation model with vascularized cartilage (VC-antler, day 60), using RFP-expressing nude mice. BGLAP (cytoplasmic) and SP7 (nuclear) expression are shown. Red fluorescent cells (RFP⁺) originate from the host and are primarily involved in blood vessel formation. Osteoblasts derived from perivascular bone marrow-derived mesenchymal stem cells (BMSCs) are marked with yellow arrowheads (OB-P), and those transdifferentiated from hypertrophic chondrocytes/PHEX⁺ cells are marked with white arrowheads (OB-D); not all positive cells are indicated. Scale bar: 200 μm\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e","description":"","filename":"Fig.5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6919532/v1/c803542de0d6145e188df1bc.jpg"},{"id":86026566,"identity":"30057dab-23b5-4e1f-a5b7-878f3953f1be","added_by":"auto","created_at":"2025-07-04 13:12:12","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":306731,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eComparative transcriptomic analysis of antler versus other tissues\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea.\u003c/strong\u003e Schematic of analyzed tissues: intervertebral disc (ID; Gan et al., 2021), embryonic long bone (EL; He et al., 2021), and osteosarcoma (Os; Zhou et al., 2020), with UMAP plots showing cartilage/bone-related cell types.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb.\u003c/strong\u003e Schematic of SVM classifier based on AnSC-derived cells (top) and dot plots showing classification probabilities using scPred (bottom).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec.\u003c/strong\u003e Schematic of Classification of human bone-related cells (top) and violin plots of predicted classification probabilities of AnSC-derived cells across ID, EL, and Os cell populations (bottom).\u003c/p\u003e","description":"","filename":"Fig.6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6919532/v1/a76682b1f21287a879ffa095.jpg"},{"id":86025638,"identity":"fcaf53e1-38e9-40a5-af39-1713d6f3f6af","added_by":"auto","created_at":"2025-07-04 13:04:12","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":1014283,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAnSC-derived cells form a vascularized microenvironment and initiate vascular development programs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea.\u003c/strong\u003e Schematic showing AnSC-derived cells contribute to vascular microenvironment formation in the AGC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb.\u003c/strong\u003e Scatter plot of AUCell scores for vascular development-related signaling pathways along AnSC-derived cell trajectories. The proportion of AnSC-derived cells along the pseudotime axis is also displayed in bins (top panel); see Figure 5i for reference.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec.\u003c/strong\u003e Immunofluorescence showing vascular development in the AGC. Note that some vessels consist solely of endothelial cells (indicated by yellow arrowheads). Red: endothelial cells; green: mural cells.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed.\u003c/strong\u003e Violin plots of CytoTRACE2 scores for endothelial and mural cells across five tissue layers. Activated/Quiescent states defined by score threshold = 0.3.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee.\u003c/strong\u003e Dot plots of differential ligand–receptor interactions (probability delta) between activated and quiescent endothelial and mural cells.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef.\u003c/strong\u003e Scatter plots of TFs associated with activated/quiescent states. Pearson correlations between TF expression and CytoTRACE2 scores are shown; top five TFs are labeled.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eg.\u003c/strong\u003e Network diagram of activated TFs in activated endothelial cells, their targets, and enriched KEGG terms.\u003c/p\u003e","description":"","filename":"Fig.7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6919532/v1/8d614bd8d7f5b0d4fa1940ba.jpg"},{"id":86025523,"identity":"9ee204e5-997e-436b-9abd-71b2a7d23df4","added_by":"auto","created_at":"2025-07-04 12:56:12","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":601210,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eComparative transcriptomics of vascular cells from antler, intervertebral disc, and osteosarcoma\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea.\u003c/strong\u003e Schematic comparing vascular cells from AGC, ID (Gan et al., 2021), and Os (Zhou et al., 2020).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb.\u003c/strong\u003e Bar plots of cell cycle phase distributions (G1, G2/M, S) in endothelial and mural cells across tissues. AGC had more mitotically active vascular cells.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec.\u003c/strong\u003e Violin plots of CytoTRACE2 scores for endothelial and mural cells across tissues. Cells were classified into activated or quiescent based on a 0.3 threshold.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed.\u003c/strong\u003e Scatter plots of DEGs in AGC vs. ID and Os, showing substantial overlaps in co-upregulated and co-downregulated genes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee.\u003c/strong\u003e Circular plots of top KEGG pathways enriched in AGC vs. ID and Os. Outer circle: gene log₂FC (red = upregulated, blue = downregulated); inner bars: z-scores indicating enrichment (positive = activation, negative = suppression).\u003c/p\u003e","description":"","filename":"Fig.8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6919532/v1/91a28b50d0117c64bf43fac9.jpg"},{"id":86026697,"identity":"20f318bc-4d56-4741-844e-2f0765de8cd8","added_by":"auto","created_at":"2025-07-04 13:20:15","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":8115524,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6919532/v1/98eb72d2-502f-4da7-8a3d-1d07949cc5dd.pdf"},{"id":86025524,"identity":"1c3a6ac6-b46e-4ba0-8105-728fccf21d98","added_by":"auto","created_at":"2025-07-04 12:56:12","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":2423461,"visible":true,"origin":"","legend":"SUPPLEMENTARY INFORMATION","description":"","filename":"suppl.docx","url":"https://assets-eu.researchsquare.com/files/rs-6919532/v1/0a3957780360d7b4d0fc51c5.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Single-cell multi-omics reveals unique regulatory mechanisms that sustains unprecedented elongation rate of bone structure, the deer antler","fulltext":[{"header":"Introduction","content":"\u003cp\u003eDeer antlers, capable of elongating up to 2 cm per day, represent the fastest known form of skeletal regeneration in mammals \u003csup\u003e\u003cspan additionalcitationids=\"CR2 CR3 CR4 CR5\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e–\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. This remarkable annual growth challenges conventional models of bone development, which rely on avascular growth plates and rigidly zoned chondrocyte maturation with growth at 2 cm per year during puberty \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Bone fractures can rapidly heal however critical sized defects, comminuted fractures, bone trauma and tumors pose challenges that medicine is yet to fully address. In contrast, antlers overcome these constraints through a vascularized cartilage matrix and a continuous gradient of cellular proliferation and differentiation producing bone \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Antlers also contain a specialized antler growth center (AGC), which integrates dense vascular networks, abundant progenitor populations, and a hybrid ossification mechanism \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eCranial neural crest cell (CCNC)-derived antler stem cells (AnSCs) have been identified as central to this regenerative process \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan additionalcitationids=\"CR12\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e–\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e, however their interactions with downstream progenitors and vascular niches remains poorly understood. Elucidating the molecular mechanisms that enable such rapid yet orderly growth is crucial, not only for understanding evolutionary adaptations but also for informing regenerative strategies in high-demand tissues. In addition, elucidating these interactions may help explain how antlers sustain extreme growth rates without progressing to oncogenic transformation \u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, in contrast to pathological conditions such as osteosarcoma.\u003c/p\u003e \u003cp\u003eTo address this, we applied integrated single-nucleus RNA sequencing (snRNA-seq), chromatin accessibility profiling (snATAC-seq), and spatial transcriptomics to map the cellular and molecular architecture of the AGC (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). Our analysis revealed three key evolutionary innovations: 1) a stem–progenitor hierarchy maintained by AnSC-derived paracrine signals; 2) hypoxia-driven vascularization that fuels metabolic needs and supports appositional growth; 3) hybrid ossification involving both chondroclasia and the direct transdifferentiation of hypertrophic chondrocytes into osteoblasts via PHEX⁺ intermediates. Together, these features establish a \"continuous flow\" growth model that emphasizes speed without compromising genomic integrity.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eBy bridging comparative biology with regenerative medicine, this study redefines the principles of endochondral ossification. It uncovers transcriptional and epigenetic programs that balance rapid osseous growth with structural stability, demonstrating nature’s capacity to optimize developmental mechanisms for high-performance tissue regeneration. These findings offer actionable insights for enhancing bone repair and engineering vascularized skeletal tissues.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eIntegrated snRNA-seq and snATAC-seq analyses reveal the cellular landscape of the antler growth center\u003c/h2\u003e \u003cp\u003eTo resolve the cellular landscape of the AGC at both the transcriptional and chromatin accessibility levels, snRNA-seq data was integrated from the five distinguishable AGC tissue layers of RM, PC, TZ, CA, and MC \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). This stratified framework provided a comprehensive view of the spatial and functional organization of the AGC. Following stringent quality control filtering (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ea, S1b and S1c), a total of 53,949 high-quality cells were retrieved for downstream analysis (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ed). Unsupervised clustering identified 10 distinct cell types (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb, \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec, \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed), each corresponding to a specific cluster (Figure S2a), as defined by established gene markers (Figure S2b) \u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. These included NT5E⁺, TWIST2⁺, SFRP2⁺ AnSCs; TNC⁺, TNN⁺ antler progenitor cells (AnPCs) and with MK167 proliferative AnPCs; COL2A1⁺ chondrocytes; COL10A1⁺ hypertrophic chondrocytes; ACP5⁺ chondroclasts; ACTA2⁺ mural cells; PECAM1⁺CD34⁺CDH5⁺ endothelial cells, CSF1R⁺ monocytes/macrophages, and TPSB2⁺ mast cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). Enrichment analysis of differentially expressed genes within these clusters provided additional validation for the identified cell types (Figure S2c). Notably, AnPCs clearly comprised both proliferative and non-proliferative populations, suggesting that cellular proliferation is one of the key contributors to the rapid growth of antlers.\u003c/p\u003e \u003cp\u003eAfter application of quality control thresholds to filter out low-quality nuclei (Figure S3a), unsupervised clustering of the snATAC-seq data identified the same 10 cell types based on the same marker genes, resulting in a total of 57,722 high-quality cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ee, \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ef, \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eg and Figure S3b). The transcriptional activity and chromatin accessibility exhibited a high degree of concordance at cell type level, with an 87.6% consistency rate (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eh), along with a complete (98.9%) agreement in their tissue-specific distributions. The proportional distributions of the ten cell types across the two omics datasets were also highly similar (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ei). Data showed, AnSCs were exclusively localized in the RM, while both proliferative and non-proliferative AnPCs were predominantly found in the RM and PC. Chondrocytes were primarily distributed across the other four tissue layers (excluding RM), with hypertrophic chondrocytes specifically enriched in the CA and MC. Chondroclasts and monocytes/macrophages were largely confined to the MC. In contrast, vascular-associated cells, including mural cells and endothelial cells, were present across all five tissue layers with a predominance in RM. We also utilized previously generated bulk RNA-seq data \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e from the five tissue layers and performed deconvolution analysis, which revealed a consistent distribution of cell type proportions (Figure S2d). Furthermore, the global transcriptome and chromatin accessibility profiles of the identified cell types exhibited strong correlations (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ej). Correlation analysis revealed three major groups: 1) AnSCs, AnPCs, and mural cells (Pearson correlation\u0026rsquo;s r\u0026thinsp;=\u0026thinsp;0.87\u0026thinsp;~\u0026thinsp;0.97), 2) chondrocytes and hypertrophic chondrocytes (r\u0026thinsp;=\u0026thinsp;0.93\u0026thinsp;~\u0026thinsp;0.99), and 3) all cells derived from peripheral blood (r\u0026thinsp;=\u0026thinsp;0.74\u0026thinsp;~\u0026thinsp;0.99). Overall, our results demonstrate a highly consistent cellular composition, reflected at both the transcriptional and chromatin accessibility levels.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eSpatially resolved cellular architecture in the antler growth center\u003c/h3\u003e\n\u003cp\u003eWe utilized Stereo-RNAseq to explore the spatial distribution of cell types within the AGC. After stringent quality control filtering (Figure S4a, S4b and S4c), a total of 57,311 cells were retrieved for the unsupervised clustering, which resided in 17 clusters (Figure S5a). These clusters were subsequently annotated into 11 cell types based on known marker genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea, \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb; Figure S5b, S5c and S5d). All the previously identified cell types from snRNA-seq and snATAC-seq analyses were detected, except mast cells, likely due to their low abundance. Notably, these identified cells exhibited a spatially sequential organization: AnSCs were located in the distal-most region of the AGC, followed by proliferative and non-proliferative AnPCs, which correspond to RM cells in the inner RM layer and prechondroblasts in the PC layer, and then chondrocytes, hypertrophic chondrocytes and PHEX\u003csup\u003e+\u003c/sup\u003e cells in the CA and MC layers. This arrangement reflects the differentiation trajectory originating from the AnSCs. Chondroclasts, which accounted for 18% of the total cell population, were interspersed among hypertrophic chondrocytes in longitudinal columns. Monocytes/macrophages were distributed alongside with chondroclasts, supporting the notion that chondroclasts arise from differentiation of monocytes.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAdditional cell types, including PHEX⁺ cells and osteoblasts, were identified by Stereo-seq but were not detected in the snRNA-seq and snATAC-seq data. PHEX⁺ cells, accounting for approximately 5% of the total cell population, and were predominantly located in the vicinity of chondrocytes, particularly in the vicinity of specialized hypertrophic chondrocytes. The PHEX gene is known to encode a membrane-bound metalloproteinase that plays a critical role in phosphate metabolism and bone mineralization \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, some hypertrophic chondrocytes may undergo further differentiation into PHEX⁺ cells with enhanced control of mineralization. Osteoblasts accounted for 13% of total cells and were predominantly localized at the proximal region of AGC, corresponding to the ossification zone below the MC, and are responsible for sustaining rapid bone formation. Consistent with the snRNA-seq data of the five tissue layers, mural cells and endothelial cells showed a continuous distribution from the distal to the proximal regions of the AGC.\u003c/p\u003e \u003cp\u003eRUNX2, a key transcription factor that regulates both early chondrocyte and osteoblast differentiation and skeletal mineralization \u003csup\u003e\u003cspan additionalcitationids=\"CR19\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. It was broadly expressed across the AnSC-derived lineages, including AnPCs, chondrocytes, and hypertrophic chondrocytes, with particularly high expression levels observed in AnPCs and hypertrophic chondrocytes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed). Chromatin accessibility analysis revealed strong RUNX2 activity specifically in the RM region, encompassing both AnSCs and AnPCs (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee), underscoring its essential role in AnSC differentiation. MCAM, encodes a critical adhesion factor for angiogenesis and is primarily expressed in vascular cells \u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. It exhibited high expression levels in both endothelial and mural cells in the AGC (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eg and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eh), accompanied by marked chromatin accessibility at its genomic locus (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ei). Immunohistochemical staining further confirmed the consistent expression patterns of both RUNX2 and MCAM at the protein-level (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ej).\u003c/p\u003e \u003cp\u003eIn the AGC, AnPCs further differentiate into chondrocytes and, as expected, express early chondrogenic marker genes, including SOX9, SOX5, SOX6, ACAN and COL9A1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ek), as well as COL2A1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb), which also is typically expressed in chondroprogenitor cells \u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. Interestingly, in addition to RUNX2, AnPCs also express osteogenic marker genes such as SP7, IBSP, SPP1, COL1A1, and CDH11, indicating that AnPCs possess both chondrogenic and osteogenic features and differentiation potential. Bank and Newbrey (1982) reported the presence of RM cells, prechondroblasts (proliferative capacity), and chondroblasts in the distal AGC (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003el). In the present study, we redefined the corresponding cell populations as AnSCs, proliferative AnPCs, and non-proliferative AnPCs. Furthermore, within the previously defined chondrocyte region, we delineated three distinct cell types: chondrocytes, hypertrophic chondrocytes and PHEX⁺ cells (mineralized chondrocytes). The observed spatial distribution of these cell types also aligned well with the established tissue layer architecture described in our previous studies \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Overall, our comprehensive spatial mapping of the cellular architecture in the AGC not only confirms and further refines the finding of previously published histological structure, but also provides novel insights into the mechanism underlying rapid formation of bone in antlers at the cellular level.\u003c/p\u003e\n\u003ch3\u003eSignaling from AnSCs promotes proliferation of AnPCs\u003c/h3\u003e\n\u003cp\u003eWe found that AnPCs consisted of distal proliferative a subset (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb), suggesting the rapid addition of AnPCs-derived tissues, at least in part, to the unprecedented growth rate of antlers. Previous studies have demonstrated that IGF1 strongly stimulates AnSCs \u003cem\u003ein vitro\u003c/em\u003e \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. In the present study, IGF1 was specifically expressed in AnSCs, while its receptor, IGF1R, was highly expressed in proliferative AnPCs (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). A similar expression pattern was observed for PTHLH and its receptor PTH1R, which are also known to regulate chondrocyte proliferation and differentiation \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. These findings suggest that AnSCs promote AnPC proliferation through paracrine signaling (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed). To further explore potential signaling interactions, we employed CellChat to investigate cell\u0026ndash;cell communication between AnSCs and proliferative AnPCs, as well as among proliferative AnPCs themselves. Both snRNA-seq and Stereo-seq data revealed significantly stronger signals outgoing from the AnSCs to proliferative AnPCs than in the reverse direction (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eSeveral active ligand\u0026ndash;receptor (LR) pairs mediating intercellular communication between AnSCs and proliferative AnPCs were identified, including WNT5A/11-FZD, THBS4\u0026ndash;SDC1/CD47, THBS1\u0026ndash;SDC1, PDGFA\u0026ndash;PDGFRA, PTN\u0026ndash;SDC1, PTN\u0026ndash;SDC2, IGF1\u0026ndash;IGF1R, and FGF16\u0026ndash;FGFR2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef). Notably, MDK, a ligand uniquely expressed by AnSCs, interacted with multiple receptors located on the proliferative AnPCs, including SDC1, SDC2, NCL and LRP1. MDK has high homology with PTN and is well known to promote cell proliferation and inhibit apoptosis, particularly in rapidly growing and regenerating tissues \u003csup\u003e\u003cspan additionalcitationids=\"CR27 CR28\" citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. In addition to the AnSC\u0026ndash;AnPC signaling, we also observed significant signaling activities that occurred among the proliferative AnPCs themselves, mediated by LR pairs such as TNC\u0026ndash;SDC1, TNN\u0026ndash;SDC1, PTN\u0026ndash;SDC1, PTN\u0026ndash;SDC2, PTN\u0026ndash;NCL, FN1\u0026ndash;ITGA4/ITGB1, and FN1\u0026ndash;SDC1. These LR interactions were further spatially mapped in the AGC (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg).\u003c/p\u003e \u003cp\u003eTo validate these signaling interactions experimentally, we established an \u003cem\u003ein vitro\u003c/em\u003e co-culture system for AnSCs and AnPCs (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eh). The results demonstrated that putative substances secreted by AnSCs significantly promoted AnPC proliferation in a time-dependent manner (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ei). Together, these findings highlight the critical role of AnSC-derived molecular signaling in driving the rapid proliferation of AnPCs, in turn promoting elongation of antlers.\u003c/p\u003e\n\u003ch3\u003eEnhanced genomic stability and multipotent potential of proliferative AnPCs\u003c/h3\u003e\n\u003cp\u003eThe unprecedented growth rate of antlers imposes considerable stress on cellular and genomic integrity. To investigate how the AGC copes with this stress, we assessed the activity (AUCell scores) of several key pathways in proliferative AnPCs. The results revealed that proliferative AnPCs exhibited significantly higher activity in DNA damage response (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;2.22 \u0026times; 10\u003csup\u003e\u0026ndash;16\u003c/sup\u003e) and DNA repair pathways (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;2.22 \u0026times; 10\u003csup\u003e\u0026ndash;16\u003c/sup\u003e) compared to their non-proliferative counterparts (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). Interestingly, although statistically significant, only minor differences were observed in apoptotic signaling (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;2.22 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;7\u003c/sup\u003e) and in DNA damage response signal transduction via p53 signaling (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;2.75 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;8\u003c/sup\u003e) between the proliferative and non-proliferative AnPCs. Chromatin accessibility analysis showed consistent results (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea) and demonstrated that proliferative AnPCs possess an enhanced capacity to maintain genomic stability without excessive activation of apoptotic pathways, supporting the \u0026ldquo;fast yet stable\u0026rdquo; growth characteristic of antlers. Moreover, the spatial expression patterns of apoptotic factors (e.g., CASP9, ANXA5, ANXA6, BAK1, BID and TP53) in proliferative and non-proliferative AnPCs, relative to other AnSC-derived cell types (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec), may represent a safeguard mechanism. This apoptotic preparedness likely enables the rapid elimination of damaged cells during intense proliferation and differentiation \u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, thereby preserving tissue integrity and ensuring genomic stability during the process of rapid growth.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eNext, we integrated transcriptome and chromatin accessibility to identify active transcription factors (TFs) in the proliferative AnPCs. This analysis revealed that TWIST2, MEOX2, EGR2, PRRX2, and ZNF37A were the top five most active TFs in the proliferative AnPCs (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed). Previous studies have shown that TWIST2 and PRRX2 are signature genes of CNCC-derived ectomesenchyme \u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e, and that AnSCs originate from CNCCs \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. Given that TWIST2 and PRRX2 are associated with high cellular plasticity \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e, we hypothesized that proliferative AnPCs may retain transcriptional signatures linked to multipotent differentiation potential. To further investigate this, we performed DAVID Gene Ontology (GO) enrichment analysis using these five TFs and their target genes. The results showed that enrichment in biological processes (BPs) was associated with embryo development (MEOX2), blood vessel development (MEOX2 and EGR2), neurogenesis (EGR2), and osteoblast differentiation (TWIST2) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee), further supporting the notion that proliferative AnPCs possess multipotent differentiation potential. In contrast, GO enrichment analysis of non-proliferative AnPCs did not reveal these terms (Figure S6a). We also applied CytoTRACE2 to calculate scores reflecting the pluripotency of proliferative and non-proliferative AnPCs. The results showed that proliferative AnPCs exhibited higher scores compared to non-proliferative AnPCs (Figure S6b). Thus, the proliferative AnPCs have a unique phenotype associated with genomic and apoptotic protection and have TFs suggestive they retain multipotent differentiation potential.\u003c/p\u003e\n\u003ch3\u003ePHEX⁺ cells contribute to antler cartilage mineralization via hedgehog signaling\u003c/h3\u003e\n\u003cp\u003eAs PHEX⁺ cells can contribute to approximately 5% of the total cell population they were investigated further. Spatial analysis of this transmembrane peptidase involved in mineralization revealed that PHEX⁺ cells were predominantly localized around hypertrophic chondrocytes (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea). Genes highly expressed in PHEX⁺ cells were significantly enriched in BPs related to bone mineralization (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;8.16 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;7\u003c/sup\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb and S7a). Moreover, the spatial expression patterns of bone mineralization-related genes (e.g., LGR4, FITM5, ASPN, IBSP, LOX, BGLAP, MMP9 and MMP13) showed strong concordance with the distribution of PHEX⁺ cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo further investigate the molecular signals driving bone mineralization in the AGC, we applied CellChat to infer intercellular communication among chondrocytes, hypertrophic chondrocytes and PHEX⁺ cells. The results revealed that chondrocytes and hypertrophic chondrocytes acted as the main signaling sources, transmitting significantly strong signals to PHEX⁺ cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed). Among the identified pathways (Figure S7b), Hedgehog (HH) signaling emerged as a key signaling axis (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ee). HH signaling is a crucial regulator in cell fate determination and tissue patterning, particularly in cartilage mineralization \u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. In the HH pathway, the ligand IHH was highly expressed in chondrocytes and hypertrophic chondrocytes, whereas the receptors PTCH1 and the transducer SMO were predominantly expressed in PHEX⁺ cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ef and Figure S7c). Notably, HHIP, a soluble inhibitory factor that binds extracellular IHH and limits its interaction with PTCH1, was also highly expressed in PHEX⁺ cells. This indicates a feedback mechanism within PHEX⁺ cells to fine-tune HH signaling and maintain mineralization homeostasis. Additionally, interaction between IHH and its receptors PTCH1/SMO stimulates vascular invasion into the hypertrophic zone, activates osteoblast differentiation, and promotes bone matrix deposition \u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. High expression of MMP9 and MMP13 in PHEX⁺ cells further facilitates vascular invasion by degrading the extracellular matrix, creating conditions favorable for osteoblast migration and bone matrix deposition \u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eTo further validate the identity and signaling involvement of PHEX⁺ cells, we performed unsupervised clustering analysis based on snRNA-seq data, which revealed four distinct chondrocyte clusters (C0\u0026ndash;C3, Figure S7d). Among these clusters, C3 displayed markedly higher expression of PHEX, identifying it as the PHEX⁺ cell population (Figure S7e and S7f). This cluster was predominantly localized in the MC layer (Figure S7g). Consistent with their role in mineralization, C3 cells also showed elevated expression of IBSP, IFITM5, MMP9 and MMP13 (Figure S7h). The expression profiles of IHH, PTCH1, HHIP, and SMO across the four clusters closely mirrored the spatial expression pattern (Figure S7i), for example, IHH was highly expressed in C0, while PTCH1 was enriched in C3. Collectively, these findings highlight the pivotal role of PHEX⁺ cells in driving antler cartilage mineralization, orchestrated by HH signaling derived from surrounding chondrocytes and hypertrophic chondrocytes.\u003c/p\u003e \u003cp\u003e \u003cb\u003ePHEX⁺ cells act as pivotal transitional intermediates in the direct transdifferentiation of hypertrophic chondrocytes into osteoblasts.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eAs a bony appendage, the antler growth is contingent on osteoblast origin, differentiation, and maturation to sustain its rapid and extensive expansion for the tissue. Interestingly, enrichment analysis revealed that genes (e.g., IBSP, BGLAP, SPP1, SP7, LOX, and COL1A1) being highly expressed in PHEX⁺ cells are involved in osteoblast differentiation (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;6.83 \u0026times; 10\u003csup\u003e\u0026ndash;13\u003c/sup\u003e; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb). This result indicates that PHEX\u003csup\u003e+\u003c/sup\u003e cells exhibit osteoblast-like transcriptional feature. Previous studies have reported that, during endochondral ossification, a subset of hypertrophic chondrocytes can avoid apoptosis and instead directly transdifferentiate into osteoblasts \u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e,\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. Moreover, RUNX2 has been shown to inhibit terminal hypertrophic chondrocyte apoptosis and promote their transdifferentiation into osteoblasts \u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e,\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. In the AGC, we observed strong RUNX2 expression not only in AnPCs but also in hypertrophic chondrocytes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec, \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed, \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef). Importantly, PHEX⁺ cells were predominantly localized in the vicinity of hypertrophic chondrocytes (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea). These findings suggest that antler hypertrophic chondrocytes may give rise to osteoblasts via an intermediate PHEX⁺ cell type. Since deer are not yet established as transgenic model organisms, genetic tools such as \u003cem\u003ePHEX-Cre\u003c/em\u003e for lineage tracing are currently unavailable. As a result, directly testing this hypothesis by tracking the fate of PHEX\u003cb\u003e⁺\u003c/b\u003e cells remains technically challenging.\u003c/p\u003e \u003cp\u003eTo circumvent this limitation, we leveraged the spatially resolved cellular architecture of the AGC. Specifically, we identified two subtypes from osteoblasts based on their proximity to PHEX⁺ cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eg): 1) OB-D, located exclusively adjacent to PHEX⁺ cells; and 2) OB-P, situated in regions lacking direct contact with PHEX⁺ cells. We hypothesized that OB-D cells are derived from hypertrophic chondrocytes through direct transdifferentiation, while OB-P cells are peripheral blood or bone marrow mesenchymal stem cells (BMSCs) through osteogenic differentiation. Dimension reduction analysis confirmed that OB-D and OB-P represent two distinct osteoblast populations (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eh). Both subtypes expressed canonical osteoblast markers (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ei), including IBSP, MGP, BGLAP, OMD, SP7, and SPP1. Among them SP7, a critical transcription factor regulating BMSC commitment to the osteoblast lineage \u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e, BGLAP, a marker of mature osteoblasts \u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, and OMD, involved in osteoblast function and matrix mineralization \u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e, were relatively highly expressed in OB-P cells. In contrast, MGP, which modulates the balance between cartilage and bone mineralization \u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e, and IBSP, which facilitates osteoblast adhesion and mineralization \u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e, were more highly expressed in OB-D cells. These results support our proposed origins of the two subtypes. Furthermore, differentiation flow along the pseudotime trajectory illustrated a developmental progression from chondrocytes and hypertrophic chondrocytes, through PHEX⁺ cells, toward OB-D cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ej, \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ek and Figure S8a, S8b and S8c). Notably, two trajectory endpoints were observed: one at hypertrophic chondrocytes and the other at OB-D cells. This suggests that a small subset of hypertrophic chondrocytes transdifferentiated into OB-D cells via the PHEX⁺ intermediate type. Along this pseudotime trajectory, the expression of osteoblast-related genes (e.g., IBSP, SPP1, BGLAP and SP7) progressively increased, whereas the hypertrophic chondrocyte marker COL10A1 was downregulated during the transdifferentiation process (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003el).\u003c/p\u003e \u003cp\u003eTo further validate whether OB-D cells are formed via transdifferentiation, we employed an \u003cem\u003ein vivo\u003c/em\u003e model using nude mice for creation of xenogeneic antlers that consist of avascularized cartilage via transplantation of antler stem cell tissue (AP) \u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e. On day 30 of AP transplantation, osteoblasts began to emerge in the center of the large avascularized cartilage nodules without visible signs of osteogenesis via chondroclasia, and by day 60, osteoblasts in the trabeculae and woven bone were found to be broadly distributed (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003em). In parallel, we also generated xenogeneic antlers that consists of vascularized cartilage using the RFP-expressing nude mice through the same procedure but the skull periosteum on the transplantation site thoroughly scraped. Our previous studies have shown that the vascular endothelial cells in xenogeneic antlers are primarily derived from the RFP-expressing host mice \u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. Immunofluorescent analysis of osteoblast marker genes, BGLAP and SP7, revealed two distinct regions of expression (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003en): one surrounding blood vessels (OB-P), indicating osteoblasts differentiated from perivascular BMSCs; and another in the avascularized cartilage/bone nodules (OB-D), implicating that the osteoblasts are transdifferentiated from hypertrophic chondrocytes or PHEX⁺ cells. These findings strongly support that a subset of osteoblasts are derived via transdifferentiation from chondrocyte-lineage cells.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eAntler exhibits a distinct transcriptional profiling from osteosarcoma despite their rapid growth\u003c/h2\u003e \u003cp\u003eTo further characterize the transcriptional profiling of AnSC-derived cells, we compared their transcriptomic signatures with those of other cartilage- and bone-associated tissues, including normal intervertebral discs (ID), embryonic long bones (EL), and pathological osteosarcoma (Os) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). We first leveraged publicly available single-cell RNA-seq datasets from ID, EL, and Os to extract cartilage/bone-related cell populations (Figure S9a, S9b and S9c). Subsequently, we employed the scPred tool to compare the transcriptional profiles of AnSC-derived cells with those of these non-antler tissues (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb). Our scPred analysis revealed that AnSCs exhibited striking similarity in expression profiling to stromal mesenchymal stem cells (MSCs) in Os, with an average classification probability exceeding 88% (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ec). Notably, MSCs represent normal cells and not of tumor origin. However, they have been shown to promote tumor cell proliferation and metastasis in various types of cancer, including osteosarcoma \u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. AnPCs showed partial similarity to osteoprogenitors from EL, while mature AGC chondrocytes demonstrated greater similarity to chondrocytes from both EL and ID, and only limited resemblance to Os chondroblasts. These findings highlight the unique, non-pathological nature of rapid antler growth, emphasizing that the developmental trajectory and transcriptomic profiling of AnSC-derived cells are fundamentally distinct from those of tumorigenic processes, despite superficial similarities in growth dynamics.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eAnSC-derived cells provide a vascularized microenvironment rapid for cartilage growth and ossification\u003c/h3\u003e\n\u003cp\u003eUnlike cartilage in other parts of the body, antler cartilage is highly vascularized. We hypothesized that AnSC-derived cells contribute to the formation of a microenvironment that promotes vascular development (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ea). To investigate how the AGC microenvironment influences vascular development, we assessed the AUC scores of several key GO BPs along the developmental trajectory of AnSC-derived cells (Figure S8d). Our analysis revealed that Hypoxia inducible factor 1alpha signaling (Pearson correlation, r\u0026thinsp;=\u0026thinsp;0.84), regulation of cellular response to hypoxia (r\u0026thinsp;=\u0026thinsp;0.89), and regulation of blood vessel endothelial cell migration (r\u0026thinsp;=\u0026thinsp;0.84) predominantly occurred in the proximal region of the AGC, which is primarily composed of hypertrophic chondrocytes and likely creates a hypoxic microenvironment (Figure S10a). Under these conditions, HIF1A induces the expression of VEGFA, which in turn promotes angiogenesis, as indicated by enrichment in GO BPs such as angiogenesis involved in wound healing (r\u0026thinsp;=\u0026thinsp;0.84). Consistently, the expression levels of both HIF1A and VEGFA were significantly upregulated in this region (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eb). In contrast, BPs such as branching involved in blood vessel morphogenesis (r\u0026thinsp;=\u0026thinsp;0.77) and blood vessel maturation (r\u0026thinsp;=\u0026thinsp;0.82) were mainly observed in the distal region of the AGC (Figure S10a). Genes associated with vascular maturation and stabilization (e.g., PDGFB, ANGPT1, and ANGPT2) were highly expressed in the distal AGC (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eb). Immunofluorescence staining further confirmed these findings, showing that mural cells were predominantly localized in the distal AGC (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ec). These results indicate that the AGC provides distinct vascular microenvironments along its distal-to-proximal axis.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eVascular cells, including mural and endothelial cells, were primarily mitotically active in the RM and PC layers (Figure S10b), suggesting that these cells are activated by signals from the AGC microenvironment. To further explore this, we applied CytoTRACE2 to calculate scores reflecting the differentiation potential of mural and endothelial cells. Based on CytoTRACE2 scores, we defined two cellular states (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ed): an activated state (CytoTRACE2 score\u0026thinsp;\u0026gt;\u0026thinsp;0.3) and a quiescent state (CytoTRACE2 score\u0026thinsp;\u0026lt;\u0026thinsp;0.3). We then used CellChat to compare the cell\u0026ndash;cell signaling networks between these two states. In endothelial cells, LR pairs associated with proliferation and differentiation, such as PTN\u0026ndash;NCL, MDK\u0026ndash;NCL, MDK\u0026ndash;ITGA6_ITGB1, and WNT5A\u0026ndash;MCAM, were significantly upregulated in the activated state compared to the quiescent state (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ee). Conversely, BMP2/4/6\u0026ndash;BMPR1A_BMPR2 signaling was notably downregulated in the activated state. Previous studies have shown that BMP signaling is associated with vascular calcification \u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e, consistent with its primary activity during chondrocyte hypertrophy. Interestingly, VEGFA signaling exhibited receptor-specific regulation: VEGFA\u0026ndash;KDR signaling was upregulated, whereas VEGFA\u0026ndash;FLT1 signaling was downregulated in the activated state. This aligns with the distinct functional roles of the two receptors, KDR promotes vasculogenesis \u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e, while FLT1 acts as a decoy receptor that inhibits angiogenic signaling \u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. In mural cells, we observed a similar upregulation of PTN, MDK, and WNT5A signaling, along with additional signals such as TNC, TNN, and FN1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ee), all known to facilitate antler regeneration \u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e,\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. In contrast, platelet-derived growth factor LR signaling (PDGFA/D\u0026ndash;PDGFRB), critical for recruiting mural cells to nascent vessels and promoting vascular maturation and stability \u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e, was downregulated in the activated state. This downregulation likely represents a functional shift, even as mural cells undergo reactivation. Overall, these findings demonstrate that AnSC-derived cells establish a microenvironment that favors vascular formation and development through spatially distinct mechanisms, and they reveal unique LR interactions that drive this process.\u003c/p\u003e\n\u003ch3\u003eEnriched TF network drives vascular niche cell activation\u003c/h3\u003e\n\u003cp\u003eNext, we performed a joint analysis of transcriptional and chromatin accessibility profiles to identify differentially enriched TFs between activated and quiescent states in both endothelial and mural cells. The results revealed that TF enrichment levels in activated endothelial and mural cells were positively correlated with CytoTRACE2 scores, indicating that increased TF enrichment was associated with enhanced cellular differentiation (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ef). Top enriched TFs in activated endothelial cells included E2F7, E2F8, CENPN, ETV4, and THAP11, whereas those enriched in activated mural cells included PRRX1, CEBPA, GTF3A, ZBTB32, and YBX1. Most of these TFs are known to be involved in cell cycle regulation. PRRX1, a marker of mesenchymal cells, has been shown to promote cell activation, including increased migratory and invasive capacities \u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e. KEGG enrichment analysis of these TFs and their target genes further revealed that, in the activated state, they are primarily involved in cell cycle progression, genomic stability and developmental signaling pathways (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eg and S10c). In contrast, TFs highly enriched in quiescent endothelial and mural cells, along with their targets, showed no enrichment in these categories but were instead associated with inflammatory and immune-related pathways, such as the TNF signaling pathway (Figure S10d). These findings suggest that the activation of vascular niche cells triggers distinct transcriptional programs that promote both proliferative and differentiation capacity, potentially contributing to their rapid growth.\u003c/p\u003e \u003cp\u003e \u003cb\u003eActivated vascular cells in AGC exhibit distinct metabolic signatures compared to other counterparts in normal and pathological tissues\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo further characterize antler vascular cells in the activated state, we compared their transcriptomic profiles with those of other cartilage/bone-associated tissues, including normal ID and pathological Os (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003ea). Notably, the proportion of mitotically active vascular cells was significantly higher in the AGC compared to both ID and Os (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eb). Similarly, cells in the activated state, defined as those with CytoTRACE2 scores greater than 0.3 in mural and endothelial populations, were further analyzed (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003ec). We identified the differentially expressed genes (DEGs) in activated AGC vascular cells relative to those in ID and Os (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003ed). Interestingly, the number of overlapping DEGs was markedly higher in the co-upregulated group (corresponding to ID_down \u0026amp; Os_down; 1,371 for endothelial cells and 1,425 for mural cells) and the co-downregulated group (ID_up \u0026amp; Os_up; 2,648 and 2,429, respectively) than in the discordant groups (ID_up \u0026amp; Os_down or ID_down \u0026amp; Os_up; 44 and 48). These findings suggest that activated vascular cells in the AGC exhibit a transcriptional program distinct from both ID and Os. KEGG enrichment analysis of the co-upregulated and co-downregulated genes further revealed an upregulation of the citrate cycle (TCA cycle) pathway (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;6.17 \u0026times; 10⁻⁴ for endothelial cells and \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.94 \u0026times; 10⁻⁶ for mural cells) and a downregulation of oxidative phosphorylation (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.57 \u0026times; 10⁻\u0026sup1;⁶ and \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;4.34 \u0026times; 10⁻\u0026sup2;\u0026sup2;, respectively) (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003ed and Figure S11a and S11b). This metabolic shift suggests that activated antler vascular cells may prioritize anabolic biosynthetic processes over mitochondrial oxidative metabolism, a pattern commonly observed in rapid growth tissues. The thermogenic uncoupling seen is these rapidly growing cells is also fascinating. This results is consistent with our previous finding that antler cells depend on the TCA cycle to obtain their energy \u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Additionally, the pronounced downregulation of the ribosome pathway (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.82 \u0026times; 10⁻\u0026sup1;⁵ and \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;4.80 \u0026times; 10⁻\u0026sup1;⁵) may reflect a shift from global protein synthesis to selective translational control, ensuring the efficient production of key regulatory proteins needed to support rapid and energy-efficient growth.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe unprecedented elongation rate of deer antlers, which can reach up to 2 cm per day, represents the fastest known skeletal growth in vertebrates. By integrating single-cell multi-omics, this study reveals how antlers overcome the metabolic and structural constraints that typically limit somatic bone growth. We identify three key evolutionary innovations: expansive progenitor proliferation, vascularized cartilage, and hybrid ossification. Collectively, these mechanisms enable antlers to achieve exceptional growth velocity while maintaining tissue stability.\u003c/p\u003e \u003cp\u003eAntler growth challenges the conventional model of endochondral ossification by redefining spatial tissue organization, regulatory dynamics, and cell fate decisions. In contrast to somatic growth plates, which maintain strict zonal segregation (resting, proliferative, hypertrophic) to accommodate mechanical loading, the AGC functions as a continuous system optimized for maximal elongation. Within the AGC, overlapping zones of proliferation and differentiation allow AnPCs to undergo sustained divisions (2 cm/day) \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, far exceeding those observed in growth plates (2 cm/year) \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e, before committing to hypertrophy. This continuous flow architecture is regulated by two levels of control: systemic endocrine cues, such as IGF1 surges that coincide with peak antler growth \u003csup\u003e\u003cspan additionalcitationids=\"CR57 CR58\" citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e, and localized TF cues of autocrine or paracrine signals and receptor activation, including AnSC-derived IGF1, PTHLH, and MDK identified in this study.\u003c/p\u003e \u003cp\u003eThis evolutionary adaptation reflects a clear trade-off. Antlers sacrifice zonal precision, which is critical for structural integrity under mechanical stress, in order to support rapid and extensive tissue expansion. A striking example is the vascularization of the AGC cartilage, which contrasts sharply with the avascular structure of somatic growth plates. Vascular channels permeate the cartilage matrix, ensuring continuous delivery of oxygen and nutrients. This enables chondrocytes to reach larger sizes and persist in greater numbers than their counterparts in somatic plates. The presence of vasculature also accelerates ossification, as osteoblasts can reach mineralization fronts through preformed vascular conduits, bypassing the delay caused by metaphyseal vascular invasion \u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e. These adaptations effectively resolve the metabolic bottlenecks that typically constrain hypertrophic expansion in avascular cartilage.\u003c/p\u003e \u003cp\u003eA central feature of antler growth is the amplification and maintenance of a proliferative reservoir. This stem\u0026ndash;progenitor hierarchy is sustained by AnSC-derived ligands, including IGF1, PTHLH, and MDK, which collectively ensure the continuous generation of proliferating AnPCs. These AnPCs exhibit elevated activity of cell cycle regulators and enhanced DNA repair capabilities, enabling rapid yet genomically stable cell division. In contrast, somatic growth plates restrict chondrocyte proliferation to preserve mechanical integrity, prioritizing quality control over growth rate \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e. Notably, the IHH\u0026ndash;PTHLH feedback loop, a conserved mechanism in skeletal development \u003csup\u003e\u003cspan additionalcitationids=\"CR63\" citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e, is markedly amplified in antlers. Chondrocyte-derived IHH maintains PTHLH expression in AnSCs, thereby sustaining the proliferative pool of AnPCs. In parallel, IGF1\u0026ndash;IGF1R signaling works synergistically with mitogenic pathways such as WNT, FGF, and PDGF to drive cellular hyperplasia. This multilayered regulatory architecture ensures that antler development remains precisely controlled through mechanisms such as regulated apoptosis, effectively preventing the uncontrolled proliferation characteristic of pathological conditions like osteosarcoma. These findings align with our transcriptomic comparison, which demonstrated that AnPCs share greater molecular similarity with osteoprogenitors located in embryonic long bone, while exhibiting marked differences from osteosarcoma cells. In addition, key regulatory networks such as TWIST2, EGR2, and MEOX2 were identified, suggesting that proliferating AnPCs possess multipotent potential. In our previous xenogeneic antler implantation experiments using fluorescent nude mice, we observed that stem\u0026ndash;progenitor cells were capable of differentiating into mural-like cells \u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eAntlers resolve a fundamental biological paradox: sustaining centimeter-scale daily growth in a tissue where nutrient supply is typically constrained by diffusion across millimeter-scale distances \u003csup\u003e\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e. This is achieved through the vascularization of the cartilage matrix, which represents a striking deviation from conventional growth plate biology. In the present study, we identified extensive ligand and receptor interactions between AnSC-derived microenvironmental cells and activated vascular cells. These interactions are closely associated with cell proliferation and differentiation and are accompanied by enriched TFs and targets involved in cell cycle progression and developmental signaling pathways. Angiogenesis within the AGC gives rise to a dense vascular network that permeates the antler cartilage \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. This network not only delivers oxygen and nutrients but also provides a structural scaffold for appositional growth, which is the principal mechanism of antler elongation. It divides the cartilage into longitudinal trabeculae lined with perivascular osteoblasts.\u003c/p\u003e \u003cp\u003eCompared to the somatic growth plate, this vascular strategy in AGC cartilage confers three key advantages. First, it offers enhanced metabolic support, as hypertrophic chondrocytes maintain direct contact with vasculature, thereby preventing hypoxia-induced apoptosis. Second, it provides greater spatial flexibility, enabling new cartilage to be added more efficiently in an appositional manner to existing trabeculae. Third, it increases ossification efficiency, as osteoblasts can immediately colonize mineralization fronts via vascular routes, bypassing the delayed metaphyseal invasion that occurs in growth plates \u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e. In contrast to growth plates, where angiogenic factors such as VEGF are restricted to hypertrophic zones, antlers exhibit continuous expression of these factors throughout the cartilage matrix, creating a sustained pro-angiogenic environment. For example, proximally, angiogenesis is driven by hypoxia via HIF1α and VEGFA, while distally, vascular maintenance and maturation are promoted by PDGFB and ANGPT1. This spatially coordinated regulatory system ensures that vascular expansion keeps pace with the rapid tissue growth of the antler.\u003c/p\u003e \u003cp\u003eOsteogenesis in the somatic growth plate is primarily achieved through chondroclasia \u003csup\u003e\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e\u003c/sup\u003e. In contrast, our study shows that antler ossification incorporates an additional mechanism: the direct transdifferentiation of hypertrophic chondrocytes into osteoblasts, a process that is absent in somatic growth plates. These cells downregulate chondrogenic markers such as COL10A1 and upregulate osteogenic markers including IBSP and BGLAP. PHEX\u003csup\u003e+\u003c/sup\u003e cells eventually differentiate into mature osteoblasts. At the molecular level, this transdifferentiation may be driven by sustained RUNX2 expression, which help reduce the lag between cartilage resorption and bone formation \u003csup\u003e\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e\u003c/sup\u003e. Importantly, IHH produced by chondrocytes-derived cells activates PTCH1/SMO signaling in PHEX\u003csup\u003e+\u003c/sup\u003e cells, thereby promoting vascular invasion and osteoblast activation. Simultaneously, matrix remodeling is facilitated by MMP9 and MMP13 secretion, which prepares the environment for new bone deposition \u003csup\u003e\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e\u003c/sup\u003e. This form of hybrid ossification, integrating both chondroclasia and direct transdifferentiation, illustrates how antlers achieve rapid bone formation without compromising structural fidelity.\u003c/p\u003e \u003cp\u003eAntlers and somatic growth plates exemplify divergent evolutionary strategies. Antlers exhibit spatially and temporally expanded proliferative zones to maximize growth rate, unburdened by mechanical loading except for gravity, as they grow from the antler tip. In contrast, somatic growth plates are optimized for structural precision to withstand mechanical stress, such as body weight. Antlers incorporate vasculature early to support their intense metabolic demands, whereas growth plates maintain avascularity to preserve zonal organization. During ossification, antlers repurpose apoptotic signals into cues for chondrocyte transdifferentiation, while growth plates enforce terminal differentiation and lineage cessation. These distinctions parallel industrial paradigms: growth plates function as assembly lines with compartmentalized processes, whereas antlers operate like continuous flow systems. By minimizing transitional delays, antlers achieve exceptional growth speed and scale, while growth plates prioritize mechanical fidelity and structural precision.\u003c/p\u003e \u003cp\u003eThe case of antler elongation illustrates how extreme growth rates can evolve through modifications of conserved developmental programs. Classical constraints such as limited proliferation, avascularity, and obligatory chondrocyte apoptosis are relaxed to permit extensive growth, vascular integration, and cell fate plasticity. These adaptations likely evolved in response to selective pressures for rapid regeneration within a brief seasonal window, during which antlers must reach the size that is proportional to the deer body-size (around 1.5 meters in length and up to 30 kilograms in weight, within approximately 90 days in large deer species, such as wapti) \u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e,\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e\u003c/sup\u003e. The translational potential of these findings is substantial, with implications for bone grafting and fracture repair. The concept of vascularized cartilage may advance tissue engineering for metabolically demanding structures such as articular cartilage. Moreover, leveraging transdifferentiation pathways could accelerate bone healing or enable therapeutic reprogramming of chondrocytes into osteoblasts for the treatment of osteoarthritis. Future research should investigate the feasibility of replicating these mechanisms in human tissues, particularly in contexts requiring rapid tissue regeneration.\u003c/p\u003e "},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eEthical approval\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe experimental protocols involving deer and mice were approved by the Ethics Committee of Changchun Sci-Tech University (Permit Number: CKARI202109). We have complied with all relevant ethical regulations for animal use.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSamples collection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe experimental protocols involving deer and mice were approved by the Ethics Committee of Changchun Sci-Tech University (Permit Number: CKARI201911), and all relevant ethical manipulations for animal use were strictly followed.\u003c/p\u003e\n\u003cp\u003eFor sampling AGC tissues used in snRNA-seq and snATAC-seq, tissues were collected from three healthy 2-year-old sika deer approximately 30 days of growth after casting of their previous hard antlers. The distal 8 cm portion of the AGC was excised and sectioned sagittally along the longitudinal axis. Five distinguishable tissue layers of the AGC were immediately identified, dissected and processed as previously described\u003csup\u003e6\u003c/sup\u003e. For Stereo-seq analysis, a relevant area of 1 cm \u0026times; 2 cm was selected, which is the maximum size that can be accommodated by the technique. To fit this size, a first growing antler (approximately 10 days old) from a healthy 1-year-old sika deer was used, allowing the 1 cm \u0026times; 2 cm section to encompass all five tissue layers along with the underlying bone tissue. Following layer dissection, AGC tissues were embedded in Tissue-Tek OCT Compound (Sakura Finetek), respectively, snap-frozen in liquid-nitrogen\u0026ndash;precooled isopentane, and stored at \u0026minus;80 \u0026deg;C prior to Stereo-seq. Cryosections were prepared at a thickness of 10 \u0026mu;m using a cryostat (Leica CM3050 S, Leica Biosystems).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCell dissociation from AGC tissues\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEach tissue sample was cut into pieces smaller than 1 mm\u0026sup3;, transferred to a 50-ml tube, and digested in DMEM (Sigma-Aldrich) containing 100 \u0026mu;g/ml collagenase type I and II (Invitrogen) at 37 \u0026deg;C for 60\u0026ndash;100 minutes with gentle shaking. Digestion was terminated by adding 10% fetal bovine serum (Gibco) once more than 50,000 cells were released. The suspension was filtered through a 70-\u0026mu;m strainer and centrifuged at 500 g for 5 minutes at 4 \u0026deg;C, and treated with 1\u0026times; RBC lysis buffer (Beyotime) to remove red blood cells. Cells were washed with PBS and live cells were counted using an AO/PI staining kit (Beyotime).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003esnRNA-seq and snATAC-seq library preparation and sequencing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSingle-cell RNA sequencing libraries were prepared following established protocols using the DNBelab C Series Single-Cell Library Prep Set (MGI, catalog #1000021082). The procedure involved several key steps: First, single-cell suspensions were processed through droplet-based encapsulation to partition individual cells. After emulsion breakage, mRNA-captured beads were collected for downstream processing. Subsequent steps included reverse transcription of captured mRNA, followed by cDNA amplification and purification.\u003c/p\u003e\n\u003cp\u003eFor single-cell ATAC-seq library construction, we employed the DNBelab C Series Single-Cell ATAC Library Prep Set (MGI, catalog #1000021878). Following the manufacturer\u0026apos;s recommendations, we prepared one library per 10,000 viable cells. The library preparation workflow involved several critical steps: First, transposed single-nucleus suspensions were partitioned using droplet-based microfluidics. The encapsulated nuclei then underwent preamplification before emulsion disruption. Subsequent processing included DNA amplification and purification to generate barcoded sequencing libraries.\u003c/p\u003e\n\u003cp\u003eFor both RNA and ATAC sequencing libraries, we implemented a standardized quality control pipeline. Library quantification was performed using the Qubit ssDNA Assay Kit (Thermo Fisher Scientific, Q10212) to ensure accurate DNA concentration measurements. This rigorous quality assessment was applied uniformly across all library types prior to sequencing.\u003c/p\u003e\n\u003cp\u003eFor both RNA-seq and ATAC-seq analyses, sequencing was performed on DIPSEQ T1 platform at China National GeneBank (CNGB). RNA-seq libraries were sequenced on the DIPSEQ T1 platform (China National GeneBank) using a paired-end 130 bp format, consisting of Read 1 (30 bp; containing 10 bp cell barcode 1, 10 bp cell barcode 2, and 10 bp unique molecular identifier) and Read 2 (100 bp transcript sequence), with an additional 10 bp sample index. ATAC-seq libraries were processed on the BGISEQ-500 platform with 50 bp paired-end reads. Both sequencing approaches implemented stringent quality control measures throughout the experimental workflow to ensure data reliability.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003esnRNA-seq data pre-processing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFirst, PISA (v.1.1)\u003csup\u003e71\u003c/sup\u003e was used to extract bead barcodes and unique molecular identifier (UMI) sequences from raw sequencing data. Next, processed reads were aligned to the reference genome\u003csup\u003e72\u003c/sup\u003e using STAR (v.2.5.3)\u003csup\u003e73\u003c/sup\u003e. To ensure data quality, beads with UMI counts below the set threshold were filtered out, and those representing the same cell were merged. Finally, gene expression profiles for each barcode were quantified using PISA.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003esnATAC-seq data pre-processing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRaw sequencing reads were processed using PISA, and then aligned to the genome\u003csup\u003e72\u003c/sup\u003e using BWA (v.0.7.15)\u003csup\u003e74\u003c/sup\u003e to generate BAM files. The BAM files were subsequently processed using bap2 (v.0.6.2)\u003csup\u003e75\u003c/sup\u003e to generate fragment files for each snATAC-seq library.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eComputational analysis of snRNA-seq Data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGene-barcode expression count matrices were loaded into R (v.4.4.1) and converted to Seurat objects for downstream analysis using the Seurat package (v.4.4.0)\u003csup\u003e76\u003c/sup\u003e. Poor-quality cells, defined as those detecting fewer than 200 genes, and genes detected in fewer than 3 cells, were filtered out. Potential doublets were identified and excluded using DoubletFinder (v.2.0.3)\u003csup\u003e77\u003c/sup\u003e to minimize the impact of technical artifacts on the analysis. Cells with a high mitochondrial gene percentage (\u0026gt;5%) were also excluded, as these may indicate apoptosis or cell lysis. After quality control, a total of 56,254 high-quality barcodes were retained for downstream analysis. Log-normalization was conducted using the \u0026lsquo;NormalizeData\u0026rsquo; function, which scaled each cell\u0026rsquo;s total read count to 10,000. The top 2,000 highly variable genes (HVGs) were identified using the \u0026lsquo;FindVariableFeatures\u0026rsquo; function with the \u0026lsquo;vst\u0026rsquo; method. Gene expression for each cell was then standardized using the \u0026lsquo;ScaleData\u0026rsquo; function. Dimensionality reduction was performed using principal component analysis (PCA). Batch effects across samples were corrected using the Harmony (v.1.2.0)\u003csup\u003e78\u003c/sup\u003e algorithm. The Harmony-generated embeddings were used for clustering by constructing a shared nearest-neighbor (SNN) graph. The clusters identified by the Louvain algorithm were evaluated for the expression patterns of well-known cell-type marker genes. The \u0026lsquo;CellCycleScoring\u0026rsquo; function from Seurat was used to assign a cell-cycle score to each cell and predict its phase (G2M, S, or G1).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eComputational analysis of snATAC-seq Data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFor the analysis of single-nucleus chromatin accessibility, downstream processing was conducted using the ArchR package (v.1.0.2)\u003csup\u003e79\u003c/sup\u003e, which generated an Arrow file for each sample. All Arrow files were then integrated into an ArchRProject object for further analysis. Genome and gene annotations required for the ArchR workflow were constructed using the \u0026lsquo;createGenomeAnnotation\u0026rsquo; and \u0026lsquo;createGeneAnnotation\u0026rsquo; functions, respectively. Mitochondrial-derived fragments were excluded, and quality metrics were calculated for each nucleus. Nuclei were excluded if their transcription start site (TSS) enrichment score was less than 1, if their unique fragment count was below 1,000, or if their nucleosome signal score was exceeded 2. Potential doublets for each sample were identified using the \u0026lsquo;addDoubletScores\u0026rsquo; function with the following parameters: nTrials = 20, k = 10, knnMethod = \u0026lsquo;UMAP\u0026rsquo;, dimsToUse = 1:30, and were subsequently removed using the \u0026lsquo;filterDoublets\u0026rsquo; function with \u0026lsquo;filterRatio\u0026rsquo; set to 1. After quality control, 103,278 nuclei passed the filtration.\u003c/p\u003e\n\u003cp\u003eTo assess chromatin accessibility at the gene level, the \u0026lsquo;addGeneScoreMatrix\u0026rsquo; function was used to generate gene scores. Dimensionality reduction was performed using iterative latent semantic indexing (LSI) on the TileMatrix via the \u0026lsquo;addIterativeLSI\u0026rsquo; function. The top 30 LSI components were selected for batch effect correction across samples using the Harmony algorithm. Cell clusters were then identified by applying the \u0026lsquo;addClusters\u0026rsquo; function to the Harmony embeddings, with the following parameters: resolution = 1.2, dimsToUse = 1:30, maxClusters = 50. Activated genes for each cluster were identified using the \u0026lsquo;getMarkerFeatures\u0026rsquo; function based on gene scores with the following parameters: useMatrix = \u0026lsquo;GeneScoreMatrix\u0026rsquo;, bias = c(\u0026lsquo;TSSEnrichment\u0026rsquo;, \u0026lsquo;log10(nFrags)\u0026rsquo;), and testMethod = \u0026lsquo;wilcoxon\u0026rsquo;. Cell type annotation for the snATAC-seq clusters was performed based on the same markers used in the snRNA-seq analysis. For visualization, Uniform Manifold Approximation and Projection (UMAP) was performed using the \u0026lsquo;addUMAP\u0026rsquo; function.\u003c/p\u003e\n\u003cp\u003eTo integrate the snATAC-seq and snRNA-seq datasets with high precision, we performed a constrained integration using the \u0026lsquo;addGeneIntegrationMatrix\u0026rsquo; function. Cell types were categorized into four groups: group 1 (AnSCs, AnPCs, proliferative AnPCs, and mural cells), group 2 (chondrocytes), group 3 (endothelial cells, monocytes/macrophages, mast cells, and chondroclasts), and group 4 (hypertrophic chondrocytes). Within each group, cells were further stratified by tissue layers (RM, PC, TZ, CA, and MC) to account for potential biological heterogeneity. The snATAC-seq datasets were then integrated with their corresponding snRNA-seq datasets. During the integration, each cell in the snATAC-seq dataset was assigned a predicted RNA cell barcode, generating an integrated gene expression profile, stored in the \u0026lsquo;GeneIntegrationMatrix\u0026rsquo;.\u003c/p\u003e\n\u003cp\u003eThe \u0026lsquo;addGroupCoverages\u0026rsquo; function was used to create pseudo-bulk replicates at the cell type scale. Next, the \u0026lsquo;addReproduciblePeakSet\u0026rsquo; function was used to call reproducible and fixed-width peaks (501 bp) within each cell type using MACS2\u003csup\u003e80\u003c/sup\u003e. Peaks were annotated based on their relative position to the nearest gene, and classified into four types: promoter, distal, exonic, and intronic.\u003c/p\u003e\n\u003cp\u003eTF binding motifs for humans were downloaded from the CIS-BP (v.2.0.0)\u003csup\u003e81\u003c/sup\u003e database. Peaks containing TF binding sites were annotated using the \u0026lsquo;addMotifAnnotations\u0026rsquo; function. The activity of each TF motif enriched in individual cells was calculated using the \u0026lsquo;addDeviationsMatrix\u0026rsquo; function, and the results were stored in the \u0026lsquo;MotifMatrix\u0026rsquo;.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStereo-seq chip preparation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTissue sections were placed on the Stereo-seq chip (BGI) and incubated at 37\u0026deg;C for 3 minutes. The sections were then fixed in methanol at -20\u0026deg;C for 30 minutes before Stereo-seq preparation. The tissue sections were washed with 0.1x SSC buffer (Thermo) containing RNase inhibitor (NEB), then permeabilized with 0.1% pepsin (Sigma) in 0.01 M HCl at 37\u0026deg;C for 5 minutes. RNA was captured by DNBs and reverse transcribed overnight at 42\u0026deg;C using SuperScript II (Invitrogen) with appropriate reagents. After reverse transcription, tissue sections were washed and digested with Tissue Removal buffer at 55\u0026deg;C for 10 minutes. cDNA was released and purified using VAHTS\u0026trade; DNA Clean Beads (0.8\u0026times;). The cDNA was amplified using KAPA HiFi HotStart ReadyMix (Roche) with cDNA-PCR primer. PCR conditions were as follows: 95\u0026deg;C for 5 minutes, 15 cycles of 98\u0026deg;C for 20 seconds, 58\u0026deg;C for 20 seconds, 72\u0026deg;C for 3 minutes, and a final extension at 72\u0026deg;C for 5 minutes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStereo-seq library preparation and sequencing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe concentration of amplified cDNA was measured using the Qubit\u0026trade; dsDNA Assay Kit (Thermo Fisher). A total of 20 ng of DNA was fragmented using in-house Tn5 transposase at 55 \u0026deg;C for 10 minutes. The reaction was stopped by adding 0.02% SDS and gently mixed at 37 \u0026deg;C for 5 minutes. For library amplification, 25 \u0026mu;l of the fragmentation product was combined with 1\u0026times; KAPA HiFi HotStart ReadyMix, 0.3 mM Stereo-seq-Library-F primer, and 0.3 mM Stereo-seq-Library-R primer, and nuclease-free water to a total volume of 100 \u0026mu;l. The PCR conditions were: 95 \u0026deg;C for 5 minutes; 13 cycles of 98 \u0026deg;C for 20 seconds, 58 \u0026deg;C for 20 seconds, and 72 \u0026deg;C for 30 seconds; and a final extension at 72 \u0026deg;C for 5 minutes. The final PCR products were purified using AMPure XP beads at bead-to-sample ratios of 0.63\u0026times; and 0.153\u0026times;, used for DNA nanoball (DNB) generation, and sequenced on the MGI DNBSEQ-Tx platform (Novogene).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStereo-seq data processing and annotation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe FASTQ files generated by the sequencer were processed using the publicly available software suite SAW (v.7.1), accessible at https://github.com/STOmics/SAW. We followed the standard workflow provided by SAW to generate spatial gene expression matrices in GEF format. The raw spatial expression matrix was converted into bins of 100 \u0026times; 100 DNBs, each representing approximately one cell. This resulted in the detection of a median of 5,430 mRNA molecules and 1,166 genes per spot.\u003c/p\u003e\n\u003cp\u003eThe spatial transcriptomics profile was analyzed using the Seurat framework. Initially, normalization and variance stabilization of molecular counts were performed with the SCTransform v2 method, introducing the \u0026quot;SCT\u0026quot; array. PCA was then conducted, and the first 30 principal components were selected for cluster identification using the Louvain algorithm, with a resolution parameter of 0.935.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGene set activity scoring\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo assess the activity of specific gene sets in the snRNA-seq dataset, we employed the AUCell (v.1.24.0)\u003csup\u003e82\u003c/sup\u003e package to calculate the area under the curve (AUC) score for each cell. For the snATAC-seq dataset, gene set activity scores were calculated using the \u0026lsquo;addModuleScore\u0026rsquo; function in the ArchR package.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCell-Cell communication inference\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo investigate key cell-cell communication signals associated with the proliferation of Proliferative AnPCs, we used the CellChat (v.2.1.2)\u003csup\u003e83\u003c/sup\u003e toolkit to construct cell\u0026ndash;cell communication networks based on the snRNA-seq and spatial transcriptomics datasets independently. Communication probabilities between cell types were computed using the \u0026lsquo;computeCommunProb\u0026rsquo; function. To visualize key signaling interactions, the\u0026nbsp;netVisual_bubble() function was applied to display significant LR pairs mediating communication from 1) AnSCs to Proliferative AnPCs and 2) Proliferative AnPCs to themselves (autocrine signaling).\u003c/p\u003e\n\u003cp\u003eTo explore the dynamics of cell\u0026ndash;cell communication in the vascular microenvironment provided by antler stem cell-derived cells, we applied CellChat to construct intercellular signaling networks based on the snRNA-seq dataset. We compared communication probabilities between activated and quiescent subpopulations within endothelial and mural cells across all significant LR pairs. The delta probability was computed and visualized as a dot plot.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentification of enriched TFs in proliferative and non-proliferative AnPCs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo identify TFs with potential regulatory roles in proliferative versus non-proliferative AnPCs, we first performed differential analysis of TF motif activity and TF expression using the \u0026lsquo;getMarkerFeatures\u0026rsquo; function. Specifically, TF motif activity was assessed by setting\u0026nbsp;useMatrix = \u0026lsquo;MotifMatrix\u0026rsquo;, while TF expression was evaluated by setting\u0026nbsp;useMatrix = \u0026lsquo;GeneIntegrationMatrix\u0026rsquo;. TFs were considered enriched in proliferative AnPCs if they exhibited a log\u003csub\u003e2\u003c/sub\u003e fold change (log\u003csub\u003e2\u003c/sub\u003eFC) in expression \u0026ge; 1 and an AUC score \u0026ge; 0.6. Conversely, TFs enriched in non-proliferative AnPCs were identified if their log\u003csub\u003e2\u003c/sub\u003eFC \u0026le; \u0026minus;1 and AUC \u0026le; 0.4.\u003c/p\u003e\n\u003cp\u003eNext, we sought to identify downstream target genes of the enriched TFs by integrating chromatin accessibility and gene expression data. Gene expression imputation was performed using the \u0026lsquo;imputeMatrix\u0026rsquo; function with the Markov Affinity-based Graph Imputation of Cells (MAGIC) algorithm. Differentially accessible peaks between proliferative and non-proliferative AnPCs were identified using the \u0026lsquo;getMarkerFeatures\u0026rsquo; function (with\u0026nbsp;useMatrix = \u0026quot;GeneIntegrationMatrix\u0026quot;), applying thresholds of log\u003csub\u003e2\u003c/sub\u003eFC \u0026ge; 1 and \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05 for proliferative AnPCs, and log\u003csub\u003e2\u003c/sub\u003eFC \u0026le; \u0026minus;1 and \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05 for non-proliferative AnPCs. A TF was considered functionally linked to a downstream gene if the corresponding motif was able to bind to peaks (located in the promoter or distal regulatory regions) that were differentially accessible in the relevant AnPC population, and if the TF-target gene pair exhibited a significant positive correlation in gene expression (Pearson\u0026rsquo;s r \u0026gt; 0.3, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05).\u003c/p\u003e\n\u003cp\u003eTo gain insights into the biological functions associated with the enriched TFs in proliferative AnPCs and their potential targets, we performed functional annotation using the DAVID (v.2023q4) database\u003csup\u003e84\u003c/sup\u003e. Finally, we used Cytoscape (v.3.9.0)\u003csup\u003e85\u003c/sup\u003e to construct a network visualizing TF-target regulatory relationships and pathway-gene annotations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTrajectory analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCellular trajectories were independently inferred using Monocle3 (v1.3.4)\u003csup\u003e86\u003c/sup\u003e and StaVia (Via 2.0)\u003csup\u003e87\u003c/sup\u003e. Monocle3 applied principal graph learning on the UMAP embedding to order cells along lineage progression, while StaVia used higher-order lazy-teleporting random walks with memory to capture both local transitions and global topology. Fine-grained vector fields generated by StaVia were visualized using the \u0026lsquo;via_streamplot\u0026rsquo; function to depict the flow of cellular state transitions. Finally, Pearson correlation was calculated between the pseudotime values inferred by Monocle3 and StaVia across all shared cells.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eProcessing of published datasets\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe scRNA-seq expression profiles of human adult intervertebral discs (ID) were obtained from the Gene Expression Omnibus (GEO) database (accession number: GSE160756) \u003csup\u003e88\u003c/sup\u003e. Cells with fewer than 300 or more than 4,000 detected genes, fewer than 2,000 or more than 20,000 UMIs, and apoptotic or lysed cells (defined by high mitochondrial gene percentage \u0026gt; 5% or high ribosomal gene percentage \u0026gt; 40%) were removed. Normalization and identification of highly variable features were performed. Dimensionality reduction was performed with PCA, followed by batch effect correction using Harmony based on the first 50 principal components. UMAP visualization was performed, and clustering was conducted with a resolution of 0.3. Cell type annotation was performed using markers from the original publication.\u003c/p\u003e\n\u003cp\u003eThe scRNA-seq datasets of human embryonic long bones (EL) were downloaded from the GEO database (accession number: GSE143753)\u003csup\u003e89\u003c/sup\u003e. Low-quality cells (fewer than 300 or more than 6,000 detected genes, fewer than 2,000 or more than 40,000 UMIs) and apoptotic or lysed cells (mitochondrial gene percentage \u0026gt; 5% or ribosomal gene percentage \u0026gt; 60%) were excluded. Normalization and identification of highly variable features were conducted, followed by dimensionality reduction using PCA and batch effect correction with Harmony based on the first 50 principal components. UMAP was used for visualization, and clustering analysis was performed at a resolution of 0.13. Cell type annotation was performed using markers from the original publication.\u003c/p\u003e\n\u003cp\u003eThe scRNA-seq datasets of osteosarcoma (Os) were acquired from the GEO database (accession number: GSE152048)\u003csup\u003e90\u003c/sup\u003e. Poor-quality nuclei (fewer than 300 or more than 8,000 detected genes, fewer than 2,000 or more than 50,000 UMIs) and apoptotic or lysing cells (high mitochondrial gene percentage \u0026gt; 10% or ribosomal gene percentage \u0026gt; 40%) were removed. Normalization and identification of highly variable features were performed. Dimensionality reduction was done using PCA, followed by batch effect correction using Harmony based on the first 50 principal components. UMAP visualization and clustering analysis were performed with a resolution of 0.15. Cell type annotation was based on markers from the original publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eComparison of cell types in published datasets with AnSC-derived cell types\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe utilized ScPred V1.9.2\u003csup\u003e91\u003c/sup\u003e which employs all principal components as gene feature space, to train the classifiers using AnSC-derived cells as references. By default, prediction models use a support vector machine with a radial kernel. Subsequently, we predicted the similar probability of cell types in ID, EL and Os by comparing them with the reference cell types.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDefinition of activated and quiescent states of vascular cells\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe used the CytoTRACE2 (v.1.0.0)\u003csup\u003e92\u003c/sup\u003e algorithm to characterize the cellular differentiation potential of endothelial and mural cells based on gene expression profiles, with higher values indicating a more progenitor-like, transcriptionally active state. These cells were then stratified into activated and quiescent states using a fixed threshold of 0.3, determined from the score distribution across regions. Cells with scores above the threshold were classified as activated state, while those below were assigned to the quiescent state.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentification of enriched TFs in activated and quiescent states of both endothelial and mural cells\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo identify enriched TFs in activated and quiescent states of both endothelial and mural cells, we first integrated the snATAC-seq and snRNA-seq datasets using the \u0026lsquo;addGeneIntegrationMatrix\u0026rsquo; function in the ArchR package. Each cell in the snATAC-seq dataset was assigned the label of its most transcriptionally similar neighbor in the snRNA-seq dataset, thereby classifying endothelial and mural cells in the snATAC-seq dataset into activated and quiescent states.\u003c/p\u003e\n\u003cp\u003eThe enriched TFs were identifined through integrated transcriptional and chromatin accessibility analyses. For each TF, differential gene expression between activated and quiescent states was assessed using the \u0026lsquo;FindMarkers\u0026rsquo; function in the Seurat package, while differential chromatin accessibility was calculated using the \u0026lsquo;getMarkerFeatures\u0026rsquo; function in the ArchR package (with\u0026nbsp;\u003euseMatrix = \u0026quot;GeneScoreMatrix\u0026quot;). TFs showing positive log₂FC in both RNA expression and ATAC accessibility were designated as activated-state-enriched TFs, whereas those with negative log₂FC in both modalities were considered enriched in the quiescent state. Downstream target gene prediction for enriched TFs was performed as described in the \u0026ldquo;Identification of enriched TFs in proliferative and non-proliferative AnPCs\u0026rdquo; section.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDual-comparison differential expression analysis for AGC vascular cells\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo investigate the transcriptomic uniqueness of vascular cells within AGC, we performed differential gene expression analyses comparing antler endothelial and mural cells with their corresponding cell types in human adult intervertebral discs (ID) and osteosarcomas (Os). Independent comparisons for each cell type were performed using the \u0026lsquo;FindMarkers\u0026rsquo; function in Seurat, with default parameters. Genes with an average log₂FC \u0026gt; 0.25 and \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05 were considered differentially expressed. The genes were categorized into four groups: 1) Upregulated in AGC relative to both ID and Os (double-positive); 2) Downregulated in AGC relative to both ID and Os (double-negative); 3) Upregulated compared to Os but downregulated compared to ID; 4) Upregulated compared to ID but downregulated compared to Os.\u003c/p\u003e\n\u003cp\u003eTo interpret the biological significance of genes specifically upregulated or downregulated in antler-derived vascular cells, we performed KEGG enrichment analysis for double-positive and double-negative genes sets separately using the DAVID database. Pathways with \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05 were considered significantly enriched. To infer the directionality of biological changes, Z-scores were calculated for each enriched term as follows:\u0026nbsp;\u0026nbsp;, where\u0026nbsp;\u0026nbsp;\u0026nbsp;and\u0026nbsp;\u0026nbsp;refer to the number of upregulated and downregulated genes, respectively, and N is the total number of genes annotated to that term\u003csup\u003e93\u003c/sup\u003e.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCo-culture of AnSCs and AnPCs and assessment of AnPC viability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe THY1⁺RXFP2⁺\u0026nbsp;AnSCs\u003csup\u003e94\u003c/sup\u003e and TNN⁺TNC⁺\u0026nbsp;AnPCs\u003csup\u003e53\u003c/sup\u003e were isolated from AGC as previously described. All cells were cultured in Dulbecco\u0026apos;s Modified Eagle Medium (DMEM; Gibco) supplemented with 100 U/mL penicillin, 100 \u0026mu;g/mL streptomycin, and 10% fetal bovine serum (FBS; HyClone), in a humidified incubator at 37 \u0026deg;C with 5% CO₂. For the co-culture assay, AnPCs were seeded in the lower chamber, while AnSCs were seeded in the upper chamber of transwell inserts with 0.4 \u0026mu;m pores. The co-culture was maintained for one week under standard culture conditions (37 \u0026deg;C, 5% CO₂). Following co-culture, AnPCs were subjected to cell viability analysis using the Cell Counting Kit-8 (CCK-8, Cat. No. K1018, Apexbio). Viability was determined by measuring absorbance and comparing it to that of untreated control cells.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImmunohistochemistry\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParaffin-embedded AGC tissue sections (2-3 mm in thickness) were de-parafnized and rehydrated. Endogenous peroxidase was quenched with 3% H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e for 5 min. Antigen retrieval was performed by boiling in 10 mM sodium citrate buffer (pH 6.0) for 10 min. The non-specifc binding sites were blocked in PBS plus 10% normal goat serum for 30 min and then incubated with primary antibodies:\u0026nbsp;RUNX2, (1:500,\u0026nbsp;cat. No. 20700-1-AP, Proteintech) and MCAM (1:200, A13927, ABclonal) at 37 \u0026deg;C for 2h. After rinsing with PBS, sections were incubated with secondary antibody for 30 min. After rinsing in PBS, all sections were stained with DAB chromogen reaction solution.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCreation of ectopic antlers either with vascularized or with avascularized cartilage on the nude mice foreheads\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA 9-month-old male sika deer calf was selected prior to pedicle initiation (March) for antler stem cell tissue (antlerogenic periosteum, AP) collection and transplantation into nude mice, aiming to generate two types of ectopic antlers: one comprising vascularized cartilage (VC-antler) in red fluorescent protein (RFP)-expressing nude mice, and the other comprising avascular cartilage (AC-antler) in conventional nude mice. The AP from one antler was divided into six pieces for VC-antler induction, and the AP from the contralateral side was similarly divided for AC-antler formation. Detailed surgical procedures were described in our previous studies\u003csup\u003e47,95,96\u003c/sup\u003e. Briefly, to induce VC-antlers, the periosteum at the implantation site on each mouse skull was completely removed prior to subcutaneous implantation of the AP. In contrast, to induce AC-antlers, the periosteum was left intact. In the absence of host periosteum, the implanted AP fused more effectively with the underlying periosteum-free skull, leading to the formation of larger ectopic antlers with vascularized cartilage. When the host periosteum was preserved, fusion was inhibited, resulting in smaller ectopic antlers with avascular cartilage.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImmunofluorescence staining\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFrozen tissue samples were cryosectioned at a thickness of 10 \u0026mu;m, thawed, and washed with PBS. To block non-specific binding, sections were incubated for 1 hour in PBS containing 10% goat serum and 0.3% Triton X-100. Primary antibodies (CD31, 1:1000, ab281583, Abcam; \u0026alpha;-Smooth Muscle Actin, 1:500, ab7817, Abcam; BGLAP, 1:200, A20800, ABclonal; SP7, 1:500, ab209484, Abcam) were diluted in PBS and applied to the sections overnight at 4 \u0026deg;C. After washing three times with PBS, sections were incubated with secondary antibodies for 1 hour at room temperature. Sections were then washed with PBS for 3 minutes and counterstained with DAPI (1:500) for 5 minutes. Fluorescence images were acquired using an EVOS M5000 microscope (Thermo Fisher).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical and Reproducibility\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe statistical methods used are indicated in the figure legends. All calculations and visualizations were performed in R. No statistical methods were used to predetermine the sample size, and the experiment was not randomized.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis project was supported by National Natural Science Foundation of China (Grant No.: U23A20523, 32470892 and 32370899), Natural Science Foundation of Jilin Province (Grant No.: 20240602030RC), Shenzhen Science and Technology Program (Grant No.\u0026nbsp;RCJC20221008092804002) and Jilin Merit Aid Study Abroad Programs (2024).\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.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eGoss R (1963) In: Sognnaes RF (ed) Mechanism of Hard Tissue Destruction. American Association for the Advancement of Science, pp 339\u0026ndash;369\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoss RJ (1983) Deer Antlers. Regeneration, Function and Evolution. Academic\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi C, Yang F, Sheppard A (2009) Adult stem cells and mammalian epimorphic regeneration-insights from studying annual renewal of deer antlers. Curr Stem Cell Res Therapy 4:237\u0026ndash;251. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2174/157488809789057446\u003c/span\u003e\u003cspan address=\"10.2174/157488809789057446\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKierdorf U, Kierdorf H, Szuwart T (2007) Deer antler regeneration: cells, concepts, and controversies. J Morphol 268:726\u0026ndash;738. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jmor.10546\u003c/span\u003e\u003cspan address=\"10.1002/jmor.10546\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrice J, Allen S (2004) Exploring the mechanisms regulating regeneration of deer antlers. Philos Trans R Soc Lond B Biol Sci 359:809\u0026ndash;822. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1098/rstb.2004.1471\u003c/span\u003e\u003cspan address=\"10.1098/rstb.2004.1471\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi C, Clark DE, Lord EA, Stanton JA, Suttie JM (2002) Sampling technique to discriminate the different tissue layers of growing antler tips for gene discovery. Anat Rec 268:125\u0026ndash;130\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSissons HA, Kember NF (1977) Longitudinal bone growth of the human femur. Postgrad Med J 53:433\u0026ndash;437. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/pgmj.53.622.433\u003c/span\u003e\u003cspan address=\"10.1136/pgmj.53.622.433\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClark DE, Lord EA, Suttie JM (2006) Expression of VEGF and pleiotrophin in deer antler. \u003cem\u003eThe anatomical record. Part A, Discoveries in molecular, cellular, and evolutionary biology\u003c/em\u003e 288, 1281\u0026ndash;1293. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/ar.a.20393\u003c/span\u003e\u003cspan address=\"10.1002/ar.a.20393\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClark DE, Li C, Wang W, Martin SK, Suttie JM (2006) Vascular localization and proliferation in the growing tip of the deer antler. \u003cem\u003eThe anatomical record. Part A, Discoveries in molecular, cellular, and evolutionary biology\u003c/em\u003e 288, 973\u0026ndash;981. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/ar.a.20364\u003c/span\u003e\u003cspan address=\"10.1002/ar.a.20364\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBanks JWN (1983) In: Banks RD (ed) W. J. in Antler development in Cervidae. Caesar Kleburg Wildlife Research Institute, pp 279\u0026ndash;306\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLandete-Castillejos T et al (2019) Antlers - Evolution, development, structure, composition, and biomechanics of an outstanding type of bone. Bone 128:115046. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.bone.2019.115046\u003c/span\u003e\u003cspan address=\"10.1016/j.bone.2019.115046\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBa H et al (2024) Insights into the molecular characteristics of embryonic cranial neural crest cells and their derived mesenchymal cell pools. Commun Biology 7:1347. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s42003-024-07056-x\u003c/span\u003e\u003cspan address=\"10.1038/s42003-024-07056-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBa H, Wang D, Yau TO, Shang Y, Li C (2019) Transcriptomic analysis of different tissue layers in antler growth Center in Sika Deer (Cervus nippon). BMC Genomics 20:173. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12864-019-5560-1\u003c/span\u003e\u003cspan address=\"10.1186/s12864-019-5560-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi C et al (2023) Deer antlers: the fastest growing tissue with least cancer occurrence. Cell Death Differ 30:2452\u0026ndash;2461. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41418-023-01231-z\u003c/span\u003e\u003cspan address=\"10.1038/s41418-023-01231-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBa H et al (2022) Single-cell transcriptome reveals core cell populations and androgen-RXFP2 axis involved in deer antler full regeneration. Cell regeneration (London England) 11:43. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13619-022-00153-4\u003c/span\u003e\u003cspan address=\"10.1186/s13619-022-00153-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang D et al (2019) Deer antler stem cells are a novel type of cells that sustain full regeneration of a mammalian organ-deer antler. Cell Death Dis 10:443. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41419-019-1686-y\u003c/span\u003e\u003cspan address=\"10.1038/s41419-019-1686-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAddison WN, Nakano Y, Loisel T, Crine P, McKee MD (2008) MEPE-ASARM peptides control extracellular matrix mineralization by binding to hydroxyapatite: an inhibition regulated by PHEX cleavage of ASARM. J bone mineral research: official J Am Soc Bone Mineral Res 23:1638\u0026ndash;1649. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1359/jbmr.080601\u003c/span\u003e\u003cspan address=\"10.1359/jbmr.080601\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang X et al (2017) Analysis of novel RUNX2 mutations in Chinese patients with cleidocranial dysplasia. PLoS ONE 12:e0181653. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0181653\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0181653\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJung YJ, Bae HS, Ryoo HM, Baek SH (2018) A novel RUNX2 mutation in exon 8, G462X, in a patient with Cleidocranial Dysplasia. J Cell Biochem 119:1152\u0026ndash;1162. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jcb.26283\u003c/span\u003e\u003cspan address=\"10.1002/jcb.26283\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeng L et al (2017) Functional analysis of novel RUNX2 mutations in cleidocranial dysplasia. Mutagenesis 32:437\u0026ndash;443. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/mutage/gex012\u003c/span\u003e\u003cspan address=\"10.1093/mutage/gex012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShih IM, Elder DE, Speicher D, Johnson JP, Herlyn M (1994) Isolation and functional characterization of the A32 melanoma-associated antigen. Cancer Res 54:2514\u0026ndash;2520\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBardin N et al (1996) S-Endo 1, a pan-endothelial monoclonal antibody recognizing a novel human endothelial antigen. Tissue Antigens 48:531\u0026ndash;539. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.1399-0039.1996.tb02666.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1399-0039.1996.tb02666.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcAlinden A (2014) Alternative splicing of type II procollagen: IIB or not IIB? Connect Tissue Res 55:165\u0026ndash;176. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3109/03008207.2014.908860\u003c/span\u003e\u003cspan address=\"10.3109/03008207.2014.908860\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi C, Pearson A, McMahon C (2013) Morphogenetic mechanisms in the cyclic regeneration of hair follicles and deer antlers from stem cells. \u003cem\u003eBioMed research international\u003c/em\u003e 643601. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1155/2013/643601\u003c/span\u003e\u003cspan address=\"10.1155/2013/643601\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVortkamp A et al (1998) Recapitulation of signals regulating embryonic bone formation during postnatal growth and in fracture repair. Mech Dev 71:65\u0026ndash;76. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/s0925-4773(97)00203-7\u003c/span\u003e\u003cspan address=\"10.1016/s0925-4773(97)00203-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXing H, Wang Q, Ma Y, Han R, Li H (2024) The significance of MDK growth factor in the antler development of sika deer (Cervus nippon): An in-depth analysis. Gene Expr Patterns 119388. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.gep.2024.119388\u003c/span\u003e\u003cspan address=\"10.1016/j.gep.2024.119388\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong Z, Li C, Coates D (2021) PTN-PTPRZ signalling is involved in deer antler stem cell regulation during tissue regeneration. J Cell Physiol 236:3752\u0026ndash;3769. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jcp.30115\u003c/span\u003e\u003cspan address=\"10.1002/jcp.30115\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIshimoto H et al (2006) Midkine, a heparin-binding growth factor, selectively stimulates proliferation of definitive zone cells of the human fetal adrenal gland. J Clin Endocrinol Metab 91:4050\u0026ndash;4056. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1210/jc.2006-1139\u003c/span\u003e\u003cspan address=\"10.1210/jc.2006-1139\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSinha S et al (2022) Fibroblast inflammatory priming determines regenerative versus fibrotic skin repair in reindeer. Cell 185:4717\u0026ndash;4736e4725. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cell.2022.11.004\u003c/span\u003e\u003cspan address=\"10.1016/j.cell.2022.11.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eColitti M, Allen SP, Price JS (2005) Programmed cell death in the regenerating deer antler. J Anat 207:339\u0026ndash;351. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.1469-7580.2005.00464.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1469-7580.2005.00464.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong Z, Coates D (2021) Bioactive Molecular Discovery Using Deer Antlers as a Model of Mammalian Regeneration. J Proteome Res 20:2167\u0026ndash;2181. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acs.jproteome.1c00003\u003c/span\u003e\u003cspan address=\"10.1021/acs.jproteome.1c00003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCalamari ZT, Flynn JJ (2024) Gene expression supports a single origin of horns and antlers in hoofed mammals. Commun Biol 7:509. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s42003-024-06134-4\u003c/span\u003e\u003cspan address=\"10.1038/s42003-024-06134-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoldatov R et al (2019) Spatiotemporal structure of cell fate decisions in murine neural crest. Sci (New York N Y) 364. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/science.aas9536\u003c/span\u003e\u003cspan address=\"10.1126/science.aas9536\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBechtold TE et al (2016) Osteophyte formation and matrix mineralization in a TMJ osteoarthritis mouse model are associated with ectopic hedgehog signaling. Matrix biology: J Int Soc Matrix Biology 52\u0026ndash;54:339\u0026ndash;354. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.matbio.2016.03.001\u003c/span\u003e\u003cspan address=\"10.1016/j.matbio.2016.03.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eColnot Cl et al (2005) Indian hedgehog synchronizes skeletal angiogenesis and perichondrial maturation with cartilage development. \u003cem\u003eDevelopment\u003c/em\u003e 132, 1057\u0026ndash;1067. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1242/dev.01649%J\u003c/span\u003e\u003cspan address=\"10.1242/dev.01649%J\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Development\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrtega N, Wang K, Ferrara N, Werb Z, Vu TH (2010) Complementary interplay between matrix metalloproteinase-9, vascular endothelial growth factor and osteoclast function drives endochondral bone formation. \u003cem\u003eDisease Models \u0026amp; Mechanisms\u003c/em\u003e 3, 224\u0026ndash;235. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1242/dmm.004226%J\u003c/span\u003e\u003cspan address=\"10.1242/dmm.004226%J\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Disease Models \u0026amp; Mechanisms\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWolff LI, Hartmann C (2019) A Second Career for Chondrocytes-Transformation into Osteoblasts. Curr Osteoporos Rep 17:129\u0026ndash;137. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s11914-019-00511-3\u003c/span\u003e\u003cspan address=\"10.1007/s11914-019-00511-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou X et al (2014) Chondrocytes transdifferentiate into osteoblasts in endochondral bone during development, postnatal growth and fracture healing in mice. PLoS Genet 10:e1004820. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pgen.1004820\u003c/span\u003e\u003cspan address=\"10.1371/journal.pgen.1004820\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKomori T (2022) Whole Aspect of Runx2 Functions in Skeletal Development. Int J Mol Sci 23. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/ijms23105776\u003c/span\u003e\u003cspan address=\"10.3390/ijms23105776\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQin X et al (2020) Runx2 is essential for the transdifferentiation of chondrocytes into osteoblasts. PLoS Genet 16:e1009169. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pgen.1009169\u003c/span\u003e\u003cspan address=\"10.1371/journal.pgen.1009169\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSinha KM, Zhou X (2013) Genetic and molecular control of osterix in skeletal formation. J Cell Biochem 114:975\u0026ndash;984. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jcb.24439\u003c/span\u003e\u003cspan address=\"10.1002/jcb.24439\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHojo H, Ohba S (2020) Gene regulatory landscape in osteoblast differentiation. Bone 137:115458. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.bone.2020.115458\u003c/span\u003e\u003cspan address=\"10.1016/j.bone.2020.115458\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRehn AP, Cerny R, Sugars RV, Kaukua N, Wendel M (2008) Osteoadherin is upregulated by mature osteoblasts and enhances their in vitro differentiation and mineralization. Calcif Tissue Int 82:454\u0026ndash;464. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00223-008-9138-1\u003c/span\u003e\u003cspan address=\"10.1007/s00223-008-9138-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuo G et al (1997) Spontaneous calcification of arteries and cartilage in mice lacking matrix GLA protein. Nature 386:78\u0026ndash;81. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/386078a0\u003c/span\u003e\u003cspan address=\"10.1038/386078a0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBianco P, Fisher LW, Young MF, Termine JD, Robey PG (1991) Expression of bone sialoprotein (BSP) in developing human tissues. Calcif Tissue Int 49:421\u0026ndash;426. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/bf02555854\u003c/span\u003e\u003cspan address=\"10.1007/bf02555854\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi C et al (2009) Development of a nude mouse model for the study of antlerogenesis\u0026ndash;mechanism of tissue interactions and ossification pathway. J Exp Zool B Mol Dev Evol 312:118\u0026ndash;135. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jez.b.21252\u003c/span\u003e\u003cspan address=\"10.1002/jez.b.21252\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang D et al (2022) Chimeric blood vessels sustained development of the xenogeneic antler: a unique model for xenogeneic organ generation. Life Med 2. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/lifemedi/lnac021\u003c/span\u003e\u003cspan address=\"10.1093/lifemedi/lnac021\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng Y, Wang G, Chen R, Hua Y, Cai Z (2018) Mesenchymal stem cells in the osteosarcoma microenvironment: their biological properties, influence on tumor growth, and therapeutic implications. Stem Cell Res Ther 9:22. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13287-018-0780-x\u003c/span\u003e\u003cspan address=\"10.1186/s13287-018-0780-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNiu Z et al (2022) Vascular Calcification: New Insights Into BMP Type I Receptor A. Front Pharmacol 13:887253. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fphar.2022.887253\u003c/span\u003e\u003cspan address=\"10.3389/fphar.2022.887253\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWaltenberger J, Mayr U, Pentz S, Hombach V (1996) Functional upregulation of the vascular endothelial growth factor receptor KDR by hypoxia. Circulation 94:1647\u0026ndash;1654. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1161/01.cir.94.7.1647\u003c/span\u003e\u003cspan address=\"10.1161/01.cir.94.7.1647\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFong GH, Rossant J, Gertsenstein M, Breitman ML (1995) Role of the Flt-1 receptor tyrosine kinase in regulating the assembly of vascular endothelium. Nature 376:66\u0026ndash;70. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/376066a0\u003c/span\u003e\u003cspan address=\"10.1038/376066a0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang D, Landete-Castillejos T (2023) Stem cells drive antler regeneration. Sci (New York N Y) 379:757\u0026ndash;758. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/science.adg9968\u003c/span\u003e\u003cspan address=\"10.1126/science.adg9968\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQin T et al (2023) A population of stem cells with strong regenerative potential discovered in deer antlers. Science 379:840\u0026ndash;847. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/science.add0488\u003c/span\u003e\u003cspan address=\"10.1126/science.add0488\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHellberg C, Ostman A, Heldin CH (2010) PDGF and vessel maturation. Recent results cancer Res Fortschr der Krebsforschung Progres dans les recherches sur le cancer 180:103\u0026ndash;114. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-3-540-78281-0_7\u003c/span\u003e\u003cspan address=\"10.1007/978-3-540-78281-0_7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOca\u0026ntilde;a OH et al (2012) Metastatic Colonization Requires the Repression of the Epithelial-Mesenchymal Transition Inducer Prrx1. Cancer Cell 22:709\u0026ndash;724. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ccr.2012.10.012\u003c/span\u003e\u003cspan address=\"10.1016/j.ccr.2012.10.012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSuttie JM et al (1985) Insulin-like growth factor 1 (IGF-1) antler-stimulating hormone? \u003cem\u003eEndocrinology\u003c/em\u003e 116, 846\u0026ndash;848. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1210/endo-116-2-846\u003c/span\u003e\u003cspan address=\"10.1210/endo-116-2-846\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi C, Littlejohn RP, Suttie JM (1999) Effects of insulin-like growth factor 1 and testosterone on the proliferation of antlerogenic cells in vitro. J Exp Zool 284:82\u0026ndash;90\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSadighi M, Haines SR, Skottner A, Harris AJ, Suttie JM (1994) Effects of insulin-like growth factor-I (IGF-I) and IGF-II on the growth of antler cells in vitro. J Endocrinol 143:461\u0026ndash;469. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1677/joe.0.1430461\u003c/span\u003e\u003cspan address=\"10.1677/joe.0.1430461\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSuttie JM, Fennessy PF, Gluckman PD, Corson ID (1988) Elevated plasma IGF 1 levels in stags prevented from growing antlers. Endocrinology 122:3005\u0026ndash;3007. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1210/endo-122-6-3005\u003c/span\u003e\u003cspan address=\"10.1210/endo-122-6-3005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilsman NJ, Van Sickle DC (1970) The relationship of cartilage canals to the initial osteogenesis of secondary centers of ossification. 168:381\u0026ndash;391. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ar.1091680305\u003c/span\u003e\u003cspan address=\"10.1002/ar.1091680305\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrighton CT (1984) The growth plate. Qld Gov Min J 15:571\u0026ndash;595\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVortkamp A et al (1996) Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Sci (New York N Y) 273:613\u0026ndash;622. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/science.273.5275.613\u003c/span\u003e\u003cspan address=\"10.1126/science.273.5275.613\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSt-Jacques B, Hammerschmidt M, McMahon AP (1999) Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes Dev 13:2072\u0026ndash;2086. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1101/gad.13.16.2072\u003c/span\u003e\u003cspan address=\"10.1101/gad.13.16.2072\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFaucheux C et al (2004) Recapitulation of the parathyroid hormone-related peptide-Indian hedgehog pathway in the regenerating deer antler. Dev Dyn 231:88\u0026ndash;97\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStockwell R (1979) Biology of Cartilage Cells. (Cambridge University\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMagne D et al (2005) Cartilage formation in growth plate and arteries: from physiology to pathology. BioEssays: news reviews Mol Cell Dev biology 27:708\u0026ndash;716. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/bies.20254\u003c/span\u003e\u003cspan address=\"10.1002/bies.20254\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWuelling M, Vortkamp A (2011) Chondrocyte proliferation and differentiation. Endocr Dev 21:1\u0026ndash;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1159/000328081\u003c/span\u003e\u003cspan address=\"10.1159/000328081\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang MC, Nakamura M, Kageyama Y, Igari Y, Sasano Y (2023) Age-Related Gene and Protein Expression in Mouse Mandibular Condyle Analyzed by Cap Analysis of Gene Expression and Immunohistochemistry. Gerontology 69:1295\u0026ndash;1306. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1159/000533921\u003c/span\u003e\u003cspan address=\"10.1159/000533921\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoss RJ (1970) Problems of antlerogesis. Clin Orthop Relat Res 69:227\u0026ndash;238\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi C et al (2024) Bone metabolism associated with annual antler regeneration: a deer insight into osteoporosis reversal. Biol Direct 19:123. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13062-024-00561-3\u003c/span\u003e\u003cspan address=\"10.1186/s13062-024-00561-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShi Q, Liu S, Kristiansen K, Liu L (2022) The FASTQ\u0026thinsp;+\u0026thinsp;format and PISA. Bioinformatics 38:4639\u0026ndash;4642. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/bioinformatics/btac562\u003c/span\u003e\u003cspan address=\"10.1093/bioinformatics/btac562\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBa H et al (2020) Chromosome-level genome assembly of Tarim red deer, Cervus elaphus yarkandensis. Sci data 7:187. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41597-020-0537-0\u003c/span\u003e\u003cspan address=\"10.1038/s41597-020-0537-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDobin A et al (2013) STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29:15\u0026ndash;21. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/bioinformatics/bts635\u003c/span\u003e\u003cspan address=\"10.1093/bioinformatics/bts635\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi H (2013) Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. \u003cem\u003earXiv: Genomics\u003c/em\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLareau CA et al (2019) Droplet-based combinatorial indexing for massive-scale single-cell chromatin accessibility. Nat Biotechnol 37:916\u0026ndash;924. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41587-019-0147-6\u003c/span\u003e\u003cspan address=\"10.1038/s41587-019-0147-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHao Y et al (2021) Integrated analysis of multimodal single-cell data. Cell 184:3573\u0026ndash;3587e3529. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cell.2021.04.048\u003c/span\u003e\u003cspan address=\"10.1016/j.cell.2021.04.048\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcGinnis CS, Murrow LM, Gartner ZJ, DoubletFinder (2019) Doublet Detection in Single-Cell RNA Sequencing Data Using Artificial Nearest Neighbors. Cell Syst 8:329\u0026ndash;337e324. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cels.2019.03.003\u003c/span\u003e\u003cspan address=\"10.1016/j.cels.2019.03.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKorsunsky I et al (2019) Fast, sensitive and accurate integration of single-cell data with Harmony. Nat Methods 16:1289\u0026ndash;1296. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41592-019-0619-0\u003c/span\u003e\u003cspan address=\"10.1038/s41592-019-0619-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGranja JM et al (2021) ArchR is a scalable software package for integrative single-cell chromatin accessibility analysis. Nat Genet 53:403\u0026ndash;411. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41588-021-00790-6\u003c/span\u003e\u003cspan address=\"10.1038/s41588-021-00790-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang Y et al (2008) Model-based analysis of ChIP-Seq (MACS). \u003cem\u003eGenome biology\u003c/em\u003e 9, R137. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/gb-2008-9-9-r137\u003c/span\u003e\u003cspan address=\"10.1186/gb-2008-9-9-r137\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeirauch MT et al (2014) Determination and inference of eukaryotic transcription factor sequence specificity. Cell 158:1431\u0026ndash;1443. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cell.2014.08.009\u003c/span\u003e\u003cspan address=\"10.1016/j.cell.2014.08.009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAibar S et al (2017) SCENIC: single-cell regulatory network inference and clustering. Nat Methods 14:1083\u0026ndash;1086. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/nmeth.4463\u003c/span\u003e\u003cspan address=\"10.1038/nmeth.4463\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJin S, Plikus MV, Nie Q (2025) CellChat for systematic analysis of cell-cell communication from single-cell transcriptomics. Nat Protoc 20:180\u0026ndash;219. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41596-024-01045-4\u003c/span\u003e\u003cspan address=\"10.1038/s41596-024-01045-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSherman BT, Panzade G, Imamichi T, Chang W (2024) DAVID Ortholog: an integrative tool to enhance functional analysis through orthologs. Bioinformatics 40. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/bioinformatics/btae615\u003c/span\u003e\u003cspan address=\"10.1093/bioinformatics/btae615\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShannon P et al (2003) Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 13:2498\u0026ndash;2504. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1101/gr.1239303\u003c/span\u003e\u003cspan address=\"10.1101/gr.1239303\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCao J et al (2019) The single-cell transcriptional landscape of mammalian organogenesis. Nature 566:496\u0026ndash;502. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41586-019-0969-x\u003c/span\u003e\u003cspan address=\"10.1038/s41586-019-0969-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStassen SV et al (2024) StaVia: spatially and temporally aware cartography with higher-order random walks for cell atlases. Genome Biol 25:224. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13059-024-03347-y\u003c/span\u003e\u003cspan address=\"10.1186/s13059-024-03347-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGan Y et al (2021) Spatially defined single-cell transcriptional profiling characterizes diverse chondrocyte subtypes and nucleus pulposus progenitors in human intervertebral discs. Bone Res 9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41413-021-00163-z\u003c/span\u003e\u003cspan address=\"10.1038/s41413-021-00163-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHe J et al (2021) Dissecting human embryonic skeletal stem cell ontogeny by single-cell transcriptomic and functional analyses. Cell Res 31:742\u0026ndash;757. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41422-021-00467-z\u003c/span\u003e\u003cspan address=\"10.1038/s41422-021-00467-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou Y et al (2020) Single-cell RNA landscape of intratumoral heterogeneity and immunosuppressive microenvironment in advanced osteosarcoma. Nat Commun 11:6322. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41467-020-20059-6\u003c/span\u003e\u003cspan address=\"10.1038/s41467-020-20059-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003e(!!! INVALID CITATION !!! 61)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKang M et al (2024) Mapping single-cell developmental potential in health and disease with interpretable deep learning. bioRxiv: preprint Serv biology. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1101/2024.03.19.585637\u003c/span\u003e\u003cspan address=\"10.1101/2024.03.19.585637\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWalter W, S\u0026aacute;nchez-Cabo F, Ricote M (2015) GOplot: an R package for visually combining expression data with functional analysis. Bioinformatics 31:2912\u0026ndash;2914. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/bioinformatics/btv300\u003c/span\u003e\u003cspan address=\"10.1093/bioinformatics/btv300\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBa H et al (2025) RXFP2-positive mesenchymal stem cells in the antlerogenic periosteum contribute to postnatal development of deer antlers. Commun Biol 8:645. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s42003-025-08085-w\u003c/span\u003e\u003cspan address=\"10.1038/s42003-025-08085-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi C, Suttie JM (2003) Tissue collection methods for antler research. Eur J Morphol 41:23\u0026ndash;30. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1076/ejom.41.1.23.28106\u003c/span\u003e\u003cspan address=\"10.1076/ejom.41.1.23.28106\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi C, Harris AJ, Suttie JM (2001) Tissue interactions and antlerogenesis: new findings revealed by a xenograft approach. J Exp Zool 290:18\u0026ndash;30\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"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":"antler, unprecedented bone elongation, genomic stability, vascularized niche, PHEX⁺cells, hybrid ossification, transdifferentiation, single-cell multi-omics","lastPublishedDoi":"10.21203/rs.3.rs-6919532/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6919532/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eBone growth and regeneration remain clinically important problems in medicine, and understanding the mechanism of rapid bone growth is a key to new therapeutic approaches. Deer antlers represent the fastest-growing bone structure in mammals, undergoing regeneration through endochondral ossification and exhibiting extraordinary elongation rates of up to 2 cm per day, far exceeding human epiphyseal growth plate extension of approximately 2 cm annually. This research aimed to systematically map the cellular and molecular architecture of the antler growth center by integrating single-nucleus RNA sequencing (snRNA-seq), chromatin accessibility profiling (snATAC-seq), and spatial transcriptomics. Our analysis revealed that antler mesenchymal stem cells (AnSCs) drive the proliferation of antler progenitor cells (AnPCs) through paracrine signaling. These rapidly proliferating cells maintain genomic stability and evade oncogenic transformation, while displaying distinct molecular signatures that differentiate them from osteosarcoma. AnSC-derived cells also establish a vascularized niche that supports robust angiogenesis to meet the high metabolic demands essential for rapid antler elongation. Furthermore, antlers utilize a hybrid ossification strategy that combines classical endochondral ossification with the direct transdifferentiation of hypertrophic chondrocytes into osteoblasts via PHEX⁺ intermediates. These findings redefine the key principles of endochondral ossification and offer novel insights for the development of regenerative therapies.\u003c/p\u003e","manuscriptTitle":"Single-cell multi-omics reveals unique regulatory mechanisms that sustains unprecedented elongation rate of bone structure, the deer antler","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-04 12:56:08","doi":"10.21203/rs.3.rs-6919532/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":"2ae258ae-6081-493b-b11b-df56c0ef8ea9","owner":[],"postedDate":"July 4th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":50441288,"name":"Biological sciences/Developmental biology/Angiogenesis"},{"id":50441289,"name":"Biological sciences/Developmental biology/Bone development"},{"id":50441290,"name":"Biological sciences/Developmental biology/Cell growth"}],"tags":[],"updatedAt":"2025-07-04T12:56:08+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-04 12:56:08","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6919532","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6919532","identity":"rs-6919532","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}