Intermittent parathyroid hormone employs autonomous and non-autonomous mechanisms to drive osteogenesis from Ebf3-expressing skeletal progenitor cells

Byron SK Chan; Bingzi Dong; Majd George; Juwell Wu; Birol Ay; Daniel J Brooks; Mary L Bouxsein; Benjamin Z Leder; Garyfallia Papaioannou; Karin Gustafsson; David T Scadden; Takashi Nagasawa; Henry M Kronenberg; Charles P Lin; Marc N Wein

doi:10.64898/2026.05.21.726951

This is a preprint.

It has not yet been peer reviewed by a journal.

The National Library of Medicine is running a pilot to include preprints that result from research funded by NIH in PMC and PubMed.

[Preprint]. 2026 May 26:2026.05.21.726951. [Version 1] doi: 10.64898/2026.05.21.726951

Intermittent parathyroid hormone employs autonomous and non-autonomous mechanisms to drive osteogenesis from Ebf3-expressing skeletal progenitor cells

Byron SK Chan ¹, Bingzi Dong ^1,², Majd George ¹, Juwell Wu ³, Birol Ay ¹, Daniel J Brooks ¹, Mary L Bouxsein ^1,⁴, Benjamin Z Leder ¹, Garyfallia Papaioannou ¹, Karin Gustafsson ^5,^6,⁷, David T Scadden ^5,^6,⁷, Takashi Nagasawa ⁸, Henry M Kronenberg ¹, Charles P Lin ³, Marc N Wein ^1,^6,^9,¹⁰

PMCID: PMC13232172 PMID: 42244729

Abstract

How systemic hormonal signals coordinate stem cell fate decisions in adult tissues remains incompletely understood. In bone marrow, Cxcl12-abundant reticular (CAR) cells, marked by Early B-cell Factor 3 (Ebf3) expression, are multipotent mesenchymal progenitors that maintain the hematopoietic stem cell niche and serves as a major osteoblast progenitor source during adult bone remodeling. Using inducible lineage tracing coupled with single-cell transcriptomics and conditional genetics in mice, we show that intermittent parathyroid hormone (iPTH; teriparatide) drives osteogenesis from CAR cells by simultaneously engaging cell-intrinsic and cell-extrinsic mechanisms. Directly, iPTH suppresses lineage-enforcing transcription factors Ebf3, Ebf1, and Foxc1, thereby destabilizing progenitor identity and priming CAR cells for osteogenic commitment. Simultaneously, iPTH stimulates osteoclastic bone resorption, releasing TGFß which recruits these primed progenitors to bone surfaces, a process abolished by osteoclast depletion. Preventing CAR cell maturation via Sp7 deletion abrogates iPTH-induced bone gain, establishing these progenitors as essential mediators of bone anabolism. This coupled mechanism, in which intrinsic transcriptional priming converges with extrinsic niche remodeling, is conserved in human CAR cells from teriparatide-treated postmenopausal women, which show concordant suppression of EBF3 and FOXC1 and elevated TGFß-responsive gene signatures. These findings reveal a general principle by which a systemic hormone orchestrates tissue remodeling through simultaneous reprogramming of progenitor identity and remodeling of the niche microenvironment.

Introduction

How systemic signals coordinate progenitor fate decisions in adult tissues is a fundamental question in stem cell biology. Progenitor differentiation depends on both cell-intrinsic transcriptional programs and extrinsic cues from the surrounding niche microenvironment^1,2. How these two levels of regulation are integrated, and whether a single extracellular signal can simultaneously engage, both to drive and coordinate fate decisions, remains poorly understood. The bone marrow provides an experimentally tractable model for addressing this question in that it harbors progenitor cells, a complex multicellular niche, and is subject to regulation by systemic cues. Osteoblasts originate from early mesenchymal progenitor cells through multiple differentiation pathways^3–7, and parathyroid hormone (PTH) acts on multiple cell types within the osteoblast lineage to regulate bone remodeling. Teriparatide, given once daily by subcutaneous injection as an intermittent parathyroid hormone (iPTH) therapy, is an effective and widely used medication to treat osteoporosis. The bone anabolic actions of iPTH are complex, but in part may be mediated by osteoblast progenitors⁸. However, osteoblast progenitors are heterogeneous, and multiple progenitor sources can contribute to the total osteoblast pool at different times and locations during skeletal growth and remodeling. Despite this, little is known about the effects of iPTH on specific subtypes of osteoblast progenitors.

CXC chemokine ligand 12 (Cxcl12)-abundant reticular (CAR) cells are a critical class of osteoblast progenitors during adult bone remodeling. These cells have a central role in supporting hematopoiesis, and also have the capacity to give rise to osteoblasts and adipocytes⁹. CAR cells are characterized by high expression of hematopoiesis-supportive paracrine factors including Cxcl12 and Stem cell factor (Kitl), as well as Leptin receptor (Lepr), and the transcription factors Forkhead box C1 (Foxc1)¹⁰ and Early B-cell Factor 3 (Ebf3). Ebf3, in particular, is a key transcription factor that maintains the ability of these cells to support hematopoiesis and inhibits their osteoblastic differentiation¹¹.

CAR cells are a major source of osteoblast progenitors responsible for homeostatic bone remodeling at adulthood, compared to other progenitor sources that play more prominent roles during earlier postnatal skeletal growth^6,12,13. Previously, we demonstrated that iPTH treatment in young/growing mice acts on a subset of growth-associated osteoblast progenitors marked by Sox9-CreERT2¹⁴. It is unclear whether these Sox9-expressing progenitors, versus other progenitor subtypes, respond to iPTH at skeletal maturity. Studies using the constitutive Lepr-Cre lineage tracing system^15–18 have suggested that iPTH promotes osteoblastic differentiation of Lepr-expressing cells. However, the lack of inducible labelling of Lepr-expressing cells at skeletal maturity in this system raises the possibility that progeny of Lepr-expressing cells, rather than CAR cells themselves, may also contribute to iPTH-induced increases in osteoblasts.

The actions of iPTH concomitantly stimulate both bone formation and bone resorption. PTH directly targets cells within the osteoblast lineage^8,14,19, but also exerts cell-extrinsic effects within the bone marrow microenvironment to enhance osteoblast differentiation. These non-cell-autonomous effects occur, in part, through iPTH-stimulated osteoclastic bone resorption, which liberates matrix-embedded growth factors such as insulin-like growth factor 1 (IGF1)²⁰ and transforming growth factor ß-1 (TGFß1)²¹ that enhance osteoblastic differentiation. In addition to their resorptive activity, osteoclasts secrete anabolic coupling factors, or clastokines, that directly increase osteoblast numbers²². Additionally, PTH directly acts on matrix-embedded osteocytes to inhibit expression of the Wnt antagonist sclerostin²³, thereby promoting osteoblast differentiation via Wnt signaling. While these cell-autonomous and non-cell-autonomous mechanisms are well established in mature osteoblast-lineage cells, it remains unclear whether and how iPTH employs similar mechanisms in early progenitors, such as CAR cells, to drive osteoblastic differentiation and bone formation.

Here, we used inducible Ebf3-CreERT2 and Sox9-CreERT2 lineage tracing systems, coupled with single-cell transcriptomics, to define the contributions of CAR cells and growth-associated osteoblast progenitors to iPTH-induced bone anabolism at skeletal maturity. We examined both cell-autonomous and non-cell-autonomous mechanisms through which iPTH regulates CAR cell fate, and investigated inter-cellular communication between osteoclasts and CAR cells during iPTH-stimulated bone remodeling. Finally, we tested whether the transcriptional programs governing iPTH action in murine CAR cells are conserved in human CAR cells from postmenopausal women receiving teriparatide.

Results

Intermittent PTH treatment recruits Ebf3-lineage cells to bone surfaces and induces their differentiation into mature osteoblasts

To assess the role of CAR cells and their descendants in response to intermittent PTH (iPTH), we performed lineage tracing of CAR cells using Ebf3-CreERT2; Rosa26-LSL-tdTomato; Bglap-GFP mice¹¹. In this system, a single dose of tamoxifen treatment transiently activates Cre recombinase to induce the permanent expression of tdTomato in CAR cells and their descendants. Additionally, the Bglap-GFP (osteocalcin-GFP) transgene, which is highly expressed only by mature osteoblasts, affords visualization of labeled cell differentiation into mature osteoblasts. Ebf3-expressing cells were labeled with one dose of tamoxifen (2 mg) in 10-week-old mice. Thereafter, mice were treated with PTH amino acids 1–34 (100 μg/kg) or vehicle once per day and sacrificed 2 days, 10 days, and 2 months later (Figure 1A). No tdTomato+ cells were observed in the absence of tamoxifen treatment (Supplemental Figure 1A, B).

Figure 1. — **(A)** Overview of lineage tracing strategy to study Ebf3-lineage cells using Ebf3-CreERT2; Rosa26-LSL-tdTomato; Bglap-GFP mice. Tamoxifen was administered at 10 weeks of age and sacrificed after 2 days, 10 days, and 2 months of iPTH or vehicle treatment. (B) Representative fluorescent images of the proximal tibia showing Ebf3-CreERT2-labeled tdTomato+ cells accumulate on bone surfaces (outlined by Bglap-GFP+ osteoblasts) and differentiate into mature osteoblasts (tdTomato+GFP+) and matrix-embedded osteocytes under steady state condition, and this process is enhanced with iPTH treatment. Red = Ebf3-CreERT2-tdTomato; green = Bglap-GFP; blue = DAPI. Scale bar = 500 μm (lower magnification) and 50 μm (higher magnification). (C) Quantification of tdTomato+ cells and tdTomato+GFP+ osteoblasts on metaphyseal trabecular and (D) endocortical bone surfaces, (E) tdTomato+ osteocytes in the cortical bone, and (F) tdTomato+ cells in the central bone marrow. Distribution (area fraction) and cell morphology (aspect ratio) analysis of tdTomato+ cells relative to defined distances from the endocortical bone surface in the proximal tibia following (G) 2 days, (H) 10 days, and (I) 2 months of iPTH or vehicle treatment. Statistical test: Two-way ANOVA with (**C-E**) treatment and time, or (**G-I**) treatment and specified distance bins as factors followed by Šidák’s multiple comparisons test. Data is expressed as mean ± SD, n = 3–4 mice/group.

In vehicle-treated mice at 2 days post-tamoxifen, labeled cells were distributed throughout the bone marrow and absent from bone surfaces (Figure 1B). There were also rare tdTomato+ cells present on periosteal surfaces. At 10 days post-tamoxifen, tdTomato+ cells were present on trabecular and endocortical bone surfaces and some differentiated into tdTomato+GFP+ osteoblasts (Figure 1B, arrowheads). Furthermore, labeled cells near and on bone surfaces appeared less reticular compared to those localized in the marrow. At 2 months post-tamoxifen in vehicle-treated mice, tdTomato+ and tdTomato+GFP+ cells continued to be present on bone surfaces. At this time point, but not earlier, tdTomato+ cells were embedded in cortical bone as terminally-differentiated osteocytes (Figure 1B, yellow arrowheads). In iPTH-treated mice, labeled cells were already localizing near trabecular and endocortical surfaces by 2 days post-tamoxifen and had adopted an elongated morphology. Following 10 days of iPTH treatment, there was an exuberant expansion of tdTomato+ cells on and near bone surfaces which was accompanied by enhanced differentiation into tdTomato+GFP+ osteoblasts. After 2 months of iPTH, the majority of trabecular and endocortical bone surfaces were occupied by tdTomato+ cells and tdTomato+GFP+ osteoblasts. Additionally, there was a greater density of tdTomato+ osteocytes in the trabecular and cortical bone upon iPTH treatment (Figure 1B).

Quantification across skeletal compartments confirmed that iPTH progressively recruits Ebf3-lineage cells to bone surfaces and drives their osteoblastic differentiation over time (Figures 1C–F, Supplementary Figures 2A–D). At 2 days, iPTH increased labeled cells on metaphyseal and diaphyseal surfaces without evidence of osteoblast differentiation. By 10 days, both progenitor and osteoblast numbers were significantly increased at trabecular and endocortical sites. After 2 months, labeled osteocytes accumulated in cortical bone, an effect seen at both metaphyseal and diaphyseal regions, while central marrow tdTomato+ cell numbers remained unchanged at all timepoints. These effects were comparable in male and female mice (Supplementary Figure 2E).

To further assess the effects of time and iPTH on tdTomato+ cell distribution and morphology relative to bone surfaces, MATLAB-based image analysis (see Methods) was performed. After 2 days, iPTH increased the area fraction (proportion of tissue area occupied by tdTomato+ pixels) and aspect ratio (major axis length/minor axis length of tdTomato+ cell) within 10 μm of endocortical bone surfaces (Figure 1G). After 10 days (Figure 1H) and 2 months (Figure 1I), iPTH treatment increased the tdTomato+ area fraction within 30 μm distance from endocortical bone. Cell shape changes, however, were confined to within 10 μm of bone, indicating that PTH-induced morphological remodeling is spatially restricted to the endocortical surface. These results demonstrate that some Ebf3-lineage cells differentiate into osteoblasts and, eventually, osteocytes, over time. Moreover, Ebf3-expressing progenitors may respond to PTH via enhanced recruitment to trabecular and endocortical bone surfaces and subsequent differentiation into mature osteoblasts.

Adipogenic differentiation of Ebf3-lineage cells is inhibited by PTH

As CAR cells can become bone marrow adipocytes^11,24 and PTH inhibits progenitor adipogenesis^8,14, we assessed whether iPTH alters the adipogenic fate of Ebf3-lineage cells (Supplementary Figure 3A). 2 months after tamoxifen-induced labeling, the total number of perilipin+ and tdTomato+perilipin+ adipocytes in the proximal tibia metaphysis increases in vehicle-treated mice and was significantly reduced by iPTH (Supplementary Figure 3B, C). Thus, iPTH may skew the fate of Ebf3-lineage cells from adipogenic to osteogenic commitment.

Ebf3-lineage osteogenic/stromal subpopulations are increased by PTH

To characterize the phenotypic heterogeneity of Ebf3-lineage cells, we performed flow cytometric analysis on enzymatically-digested marrow plus bone fragments using a panel of cell surface markers that identifies early subpopulations of skeletal stem and progenitors (mSSC, pre-BCSP, and BCSP) that give rise to lineage-restricted osteochondrogenic (Thy1+) and HSC-supporting stroma (6C3+) subpopulations²⁵ (Supplementary Figure 4A). First, we defined where initially labeled Ebf3-lineage CAR cells track in these FACS-defined subpopulations. 10-week-old Ebf3-CreERT2; Rosa26-LSL-tdTomato; Cxcl12-GFP transgenic mice were studied 2 days post-tamoxifen (Supplementary Figure 5A). Flow cytometry revealed Ebf3-CreERT2 marks the majority (~85%) of Cxcl12-GFP expressing cells (Supplementary Figure 5B). Moreover, the parental Ebf3-lineage CAR cells identified as Cxcl12-GFP+tdTomato+ cells predominantly occupy the 6C3+ FACS subpopulation (Thy1–6C3+; Supplementary Figure 5C).

Next, we assessed the effects of iPTH on the distribution of Ebf3-lineage cells with respect to these skeletal stem and progenitor subpopulations. At 2 days post-tamoxifen (Figure 2A), iPTH treatment led to modest but significant increases in CD45–Ter119–CD31– (hereafter Lineage–) tdTomato+ cells (Figure 2B) while the subsets of tdTomato+ skeletal progenitors did not change (Figure 2C). Following 10 days of iPTH (Figure 2D), there were no differences in overall numbers of Lineage–tdTomato+ cells (Figure 2E). However, within tdTomato+ cells, iPTH decreased numbers of pre-BCSP cells while increasing 6C3+ cells (Figure 2F). Therefore, within the Ebf3-lineage, iPTH promotes the differentiation of early subsets of skeletal progenitors (Pre-BCSP) towards 6C3+ stromal cells.

Figure 2. — Representative flow cytometry plots and analysis of non-hematopoietic Ebf3-CreERT2-labeled tdTomato+ cells and their skeletal stem and progenitor cell surface markers following 2 (**A-C)** and 10 days (**D-F)** of iPTH or vehicle treatment. Red box: CD45–Ter119–CD31–tdTomato+ fraction. Yellow Box: CD45–Ter119–CD31–tdTomato+Thy1–6C3– fraction. mSSC: Mouse skeletal stem cell; Pre-BCSP: Pre-bone/cartilage/stromal progenitor; BCSP: Bone/cartilage/stromal progenitor. (G) Experimental overview for *in vitro* culturing of FACS-isolated non-hematopoietic Ebf3-CreERT2-labeled tdTomato+ cells. (H) Cyclic adenosine monophosphate (cAMP) analysis, (I) immunocytochemical staining and (J) flow cytometry analysis for phospho-CREB, of *in vitro* cultured tdTomato+ cells thirty minutes after PTH, forskolin, or vehicle treatment. Statistical test: Student’s T-test (**B, E, H, and J**) and Two-way ANOVA with treatment and subpopulations (**C and F**) as factors followed by Šidák’s multiple comparisons test. Data is expressed as mean ± SD, n = 3–5 mice/group.

Across multiple experiments (see below), we observed that increased tdTomato+ cells detected by histology were not consistently reflected by flow cytometry. We speculated that our serial enzymatic digestion inefficiently liberates strongly bone-adherent tdTomato+ cells (presumably pre-osteoblasts, osteoblasts, and osteocytes). RT-qPCR on bone fragments remaining post-serial digestion showed higher tdTomato mRNA with iPTH, suggesting a greater number of tdTomato+ cells remains associated with bone fragments following iPTH (Supplementary Figure 5D).

Ebf3-lineage cells can respond directly to PTH

To test potential direct actions of PTH on Ebf3-lineage cells, Lineage–tdTomato+ cells were isolated by FACS 2 days after tamoxifen treatment in 10-week-old Ebf3-CreERT2; Rosa26-LSL-tdTomato mice. After short-term (7 day) in vitro culture (Figure 2G), cells were treated for 30 minutes with PTH which led to increases in cAMP (Figure 2H) and phospho-CREB (Figure 2I, J). As detailed below, Pth1r mRNA expression is also noted in CAR cells and their descendants. These results indicate that Ebf3-lineage cells can directly respond to PTH for their commitment towards the osteogenic differentiation. iPTH can increase osteoblast progenitor and osteoblast numbers via anti-apoptotic mechanisms^14,26. Therefore, we assessed whether iPTH affected the proliferation and apoptosis of Ebf3-lineage cells by EdU and Annexin V staining, respectively. After 10, but not 2, days, iPTH decreased the percentage of Annexin V+tdTomato+ cells (Supplementary Figure 6A). iPTH did not impact the percentage of EdU+tdTomato+ cells (Supplementary Figure 6B).

Transcriptomic effects of PTH on Ebf3-lineage cells

To understand the mechanisms by which iPTH promotes osteoblastic differentiation of Ebf3-lineage cells, we performed single cell RNA-sequencing on FACS-isolated CD45–Ter119–tdTomato+ cells following 2 and 10 days of vehicle or iPTH treatment. Mice were treated with PTH or vehicle 30 minutes prior to sacrifice to capture acute transcriptomic effects of PTH at both time points. After merging all libraries with Harmony²⁷, an algorithm for integrating multiple single cell RNA-sequencing datasets, tdTomato mRNA was expressed in all cell clusters (Supplementary Figure 7A). Unsupervised clustering revealed 15 distinct groups of Ebf3-lineage cells: 6 CAR, 1 pre-osteoblast, 1 mature osteoblast, 3 periosteal, 2 pericyte, 1 endothelial, and 1 myogenic (Figure 3A). Relative to other clusters, CAR subsets were identified based on the highest expression of Cxcl12, Lepr, Ebf3, Adipoq, and Foxc1. A pre-osteoblast cluster showed high expression of Limch1, Kcnk2, Wif1, Dlx5, Lrp4, Spp1, and Alpl. Notably, the expression of these pre-osteoblastic genes overlaps with genes that are expressed in the CAR6 cluster, suggesting that this may be an osteogenic-primed CAR subset (so-called OsteoCAR), as other groups have reported^28,29. The mature osteoblast cluster was identified based on highest Col1a1, Bglap, Dmp1, and Pth1r expression and lower expression of the pre-osteoblast defining genes. Periosteal clusters were identified based on expression of Prg4, Mfap5, and Clec3b²⁸. The remaining clusters represented pericytes, endothelial cells, and a rare myogenic population, identified by canonical markers (Figure 3B).

Figure 3. — **(A)** Harmony-integrated UMAP of single cell RNA-sequencing data from FACS-isolated CD45–Ter119–tdTomato+ cells labeled by Ebf3-CreERT2 following 2 and 10 days of iPTH or vehicle treatment (2 days: vehicle = 13,058 cells and PTH = 12,214 cells; 10 days: vehicle = 34,735 cells and PTH = 19,809 cells; n = 4 mice/group). (B) Dot plot summarizing the expression levels of cell-type marker genes across subclusters within Ebf3-lineage cells. Dot size represents the fraction of positive cells within the cluster expressing the gene of interest, and color scale denotes average gene expression level. Purple = higher expression and yellow = lower expression. (C) RNA velocity-based fate probability analysis using the CellRank⁴² to infer Ebf3-lineage CAR cell fates at steady state or in response to PTH. Color scale represents fate probability of each cell. Violin plots show the fate probability of Ebf3-lineage CAR subclusters transitioning toward the pre-osteoblast cluster under steady state or PTH-treated conditions. (D) Volcano plot showing the total number of significant differentially expressed genes (DEGs) of all Ebf3-lineage CAR subclusters regulated by PTH (adjusted P-value < 0.05 and |log2FC| > 0.5). (E) Gene set enrichment analysis of significantly up- and down-regulated DEGs by PTH using the WikiPathways 2024 Mouse database. Top 12 significantly enriched gene sets are shown (adjusted P-value < 0.05). (F) PTH-responsive module scores visualized by Feature plots of each Ebf3-lineage cell (top) and violin plots (bottom) across Ebf3-lineage CAR and osteoblast clusters in response to iPTH or vehicle treatment. Purple = higher expression and yellow = lower expression. Cliff’s delta (δ) = non-parametric effect size used to measure differences between module score distributions. (G) Violin plots showing the expression levels of canonical PTH target genes, (H) adipogenic genes, (I) blood supportive genes, and (J) CAR cell fate regulating transcription factor genes, across Ebf3-lineage CAR and osteoblast clusters in response to iPTH or vehicle treatment. Differentially expressed genes between PTH and control cells within each cluster was assessed using the MAST hurdle model⁷⁴. Asterisk (*) denotes significance (adjusted P-value < 0.05 and |log2FC| > 0.5).

To further resolve the heterogeneity within the 6 different CAR clusters, we performed DEG analysis of each CAR cluster versus all the other CAR subsets (Supplementary Figure 7B). While all CAR subsets express high levels of Cxcl12, its expression in CAR1 is the highest. The CAR1 subset is also enriched with genes such as Adipoq, Sar1a, Srpr, P4hb, Os9, and Nfe2l1 that are involved with secretion^30,31 (Supplementary Figure 7C) and may be most similar to previous reports of an adipogenic CAR subset^29,32. CAR2 has the lowest Cxcl12 and highest Ebf3 expression and is enriched in genes (like Dock1, Trio, Arhgap35) which play roles in cytoskeletal remodeling and cell migration^33–35. CAR3 is enriched in expression of Foxc1, Snai2, and Abcg2, all genes associated with stem and progenitor cells^10,36,37. CAR4 is enriched in Pecam1, Cx3cr1, and Sema4a which may be involved with endothelial cell interactions³⁸. CAR5 is enriched in expression of Ier5 and Nfkbia, which are involved in early stress responses^39,40. As noted above, CAR6 is enriched in expression of Kcnk2, Alpl, Wif1, Limch1, Lrp4, Spp1, and Nav2, a profile of enriched markers that is consistent with an osteogenic CAR subset^28,29,41.

To study the relationship between CAR subsets and how they are regulated by iPTH, we performed RNA velocity-based fate probability analysis using CellRank⁴² (Figure 3C). Three putative macrostates, which represents cells that share similar transition probabilities, were predicted: CAR1, CAR3, and pre-osteoblast. CAR1 and pre-osteoblast were identified as terminal states with high self-transition probabilities suggesting a differentiated transcriptional state. Notably, when comparing between all CAR clusters in the vehicle treated group, CAR6 shows the highest fate probability towards pre-osteoblasts. Remarkably, iPTH treatment increases the fate probability of all CAR subsets towards pre-osteoblasts and decreases their fate probability towards adipocyte-primed CAR1 cells (Figure 3C).

To understand the transcriptional mechanisms associated with iPTH-induced osteoblastic CAR cell differentiation, we next performed cluster-specific DEG analysis comparing vehicle versus iPTH in all CAR and osteoblast clusters. A total of 4,134 upregulated and 4,068 downregulated DEGs were identified to be altered by PTH (Figure 3D). Several canonical PTH target genes like Tnfsf11 (RANKL), Mmp13, and Vdr were identified as upregulated DEGs, and adipogenesis-related genes Pparg, Zfp467, and Cebpa were identified as downregulated DEGs. Next, we performed gene set enrichment analysis (GSEA) of all the upregulated and downregulated DEGs (Figure 3E). The top pathways with significant (p_adj < 0.05) enrichment amongst the upregulated DEGs are those associated with protein translation, endochondral ossification, and TGFß signaling. The top pathways amongst genes downregulated by PTH are those related to adipogenesis, Notch signaling, and mechanisms associated with pluripotency.

To assess overall PTH responsiveness, we computed a module score from 13 canonical PTH target genes (Tnfsf11, Fos, Fosl1, Junb, Egr1, Crem, Nr4a2, Nr4a3, Nfkbia, Il6, Pde4d, Igf1, and Pde10a). As expected, iPTH increased this module score in CAR subsets, pre-osteoblast, and mature osteoblast clusters. Of all CAR subsets, CAR6 showed the largest increase module score with PTH treatment (Figure 3F).

Next, we queried the expression levels of canonical PTH target genes across the Ebf3-lineage CAR and osteoblast clusters. Expression levels of Tnfsf11, Mmp13, Vdr, and Cebpb, were induced by PTH in all CAR and osteoblast clusters. Igf1 was increased in all CAR and pre-osteoblast clusters. Interestingly, other PTH target genes are uniquely increased in specific CAR subsets. For example, Il6 expression is increased in CAR4/5/6 and pre-osteoblast; Atf4 only in CAR6; Nr4a2 and Nr4a3 in CAR3–6 and osteoblast clusters; and Pde10a was increased in CAR2–6 and osteoblast clusters (Figure 3G).

As demonstrated earlier, iPTH exerts anti-apoptotic actions in Ebf3-lineage cells (Supplementary Figure 6A). We queried the list of upregulated and downregulated DEGs that may be involved in apoptosis and observed that PTH downregulates a variety of pro-apoptotic genes (Supplementary Figure 8A). An anti-apoptosis/survival module score was computed using a curated gene module (Bcl2, Bcl2l1, Mcl1, Bcl2a1a, Bcl2a1b, Cflar, Xiap, Birc2, Birc3, Traf1, Traf2, Akt1. Akt2, Rel, Rela, Nfkb1, Nfkbia) to further assess the anti-apoptotic effects of iPTH on Ebf3-lineage CAR cells and osteoblasts. In response to PTH, anti-apoptosis module scores increased, albeit to a more modest degree than the PTH-responsive score, across CAR2–6, pre-osteoblast, and mature osteoblast clusters, with CAR6 and mature osteoblast clusters exhibiting the strongest scores (Supplementary Figure 8B).

iPTH suppressed the adipogenic transcription factors Pparg, Cebpa, and Zfp423^43,44 across CAR3–6 and pre-osteoblast clusters (Figure 3H). CAR1 cells were a notable exception: they expressed the highest levels of Adipoq and Apoe, and neither gene changed with PTH treatment, consistent with CAR1 representing an adipogenic-committed subset. The osteoblast differentiation factors Sp7 and Runx2, along with their target Bglap, were acutely downregulated in CAR6 and osteoblast clusters (Supplementary Figure 7D), a pattern previously described as part of the acute transcriptional response to PTH in osteoblast-lineage cells⁴⁵. Despite unchanged Col1a1, iPTH upregulated Col3a1, Col4a1, and Col6a1 in CAR3–6 and pre-osteoblasts, suggesting active remodeling of the extracellular matrix during CAR cell differentiation.

CAR cells are crucial components of the hematopoietic supportive niche. While short term iPTH treatment did not affect Cxcl12 levels in any CAR subsets, Kitl was significantly decreased in all CAR and osteoblast clusters (Figure 3I). Other marrow niche supportive genes such as Angpt1⁴⁶ and Vcam1⁴⁷ were downregulated by PTH in CAR2–6 and pre-osteoblasts. In contrast, Vegfa and Thbd was increased in CAR and pre-osteoblast clusters. Thus, PTH actions on CAR subsets modulates the expression of paracrine-acting genes that participate in inter-cellular communication in the bone marrow microenvironment.

Given the effects of PTH on osteogenic differentiation of CAR cells, we queried expression levels of critical transcription factors (Ebf1, Ebf3, and Foxc1) that enforce the CAR cell phenotype^10,11. Ebf1 and Ebf3 prevent osteoblastic differentiation of CAR cells¹¹, and expression of both of these factors (which show lower expression in pre-osteoblasts and osteoblasts versus CAR subsets) is significantly reduced by PTH (Figure 3J). Expression of the essential CAR cell-enforcing TF, Foxc1, was also significantly downregulated in response to PTH. Taken together, these findings highlight the heterogeneity of CAR cells and their subset-specific responses to PTH. Most notably, these data suggest that PTH destabilizes the CAR cell state by suppressing Ebf3, Ebf1, and Foxc1 expression to facilitate their differentiation into osteoblasts.

A subset of Ebf3-lineage endothelial cells does not change in numbers in response to PTH

Osteogenesis and angiogenesis are intricately linked during bone growth and response to parathyroid hormone^48,49. Subsets of vasculature-associated CAR cells have been previously identified to support the hematopoietic niche⁵. Single-cell RNA-sequencing identified two distinct CD31-expressing populations within Ebf3-lineage cells: a true endothelial cluster expressing canonical markers (Pecam1, Emcn, Flt1, Tek, Cdh5, Ptprb) and CAR4, which expresses lower CD31 without other endothelial-specific transcripts (Figures 3A, B; Supplementary Figure 7C, 9A). Immunostaining for CD31 and endomucin confirmed that a subset of labeled cells displayed endothelial morphology; notably, these CD31+tdTomato+ cells were absent from bone surfaces (Supplementary Figure 9B). Flow cytometry showed iPTH did not affect the proportion of CD31+tdTomato+ cells within the non-hematopoietic labeled population (Supplementary Figure 9C).

Sox9-lineage cells rarely differentiate into mature osteoblasts following PTH treatment in skeletally mature mice

Sox9-CreERT2 marks osteoblast progenitors in the growth plate and perichondrium that respond to iPTH in early postnatal mice^6,14. To determine whether this progenitor source remains active at skeletal maturity, we performed Sox9-lineage tracing at two ages. When labeled at 3 weeks of age, Sox9-lineage cells responded to iPTH with increased progenitor and osteoblast numbers in the primary spongiosa and cortical bone (Supplementary Figure 10A, B), and is consistent with previous findings¹⁴. Single-cell RNA-sequencing of these young-labeled cells identified mature osteoblast, pre-osteoblast, and CAR cell clusters (Supplementary Figure 10C, D), the latter consistent with CAR cells being Sox9-lineage descendants during skeletal growth^6,50. In contrast, Sox9-lineage cells labeled at 10 weeks of age, using an identical tamoxifen pulse-chase protocol, showed no iPTH-induced changes in cell numbers or osteoblast differentiation (Supplementary Figure 10E–G). Single-cell RNA-sequencing revealed that these adult-labeled cells consist predominantly of chondrocytes (Prg4, Acan, Col2a1), with PTH target genes induced in Pth1r-expressing chondrocyte clusters but no osteoblast or CAR cell clusters present (Supplementary Figure 10H–J). Thus, while Sox9-lineage progenitors contribute to osteoblasts and give rise to CAR cells during postnatal growth, they are supplanted by Ebf3-lineage CAR cells as the primary iPTH-responsive osteoprogenitor pool at skeletal maturity.

PTH-induced osteoblastic differentiation of Ebf3-lineage cells is inhibited by Sp7 deletion

Osterix (Sp7) is an essential transcription factor for osteoblast differentiation⁵¹. In our scRNA-seq data of Ebf3-lineage cells we noted that Sp7 is expressed only in CAR6, pre-osteoblasts, and osteoblasts, and not in CAR1–5 subsets (Figure 4A). We deleted Sp7 in Ebf3-lineage cells in Sp7 f/f mice (Ebf3-CreERT2; Sp7 f/f; Rosa26-LSL-tdTomato; Bglap-GFP). Wild type controls (Ebf3-CreERT2; Sp7 +/+; Rosa26-LSL-tdTomato; Bglap-GFP) and Sp7 f/f mice were pulsed with a single dose of tamoxifen to induce tdTomato reporter expression and deletion of Sp7 in Ebf3-lineage cells followed by vehicle or PTH treatment for 10 days (Figure 4B).

Figure 4. — **(A)** Violin plots showing Sp7 expression level across Ebf3-lineage CAR and osteoblast clusters. (B) Overview of lineage tracing strategy to study the role of Sp7 in Ebf3-lineage cells following 10 days of iPTH or vehicle treatment. (C) Representative fluorescent images of the proximal tibia from wild type and Sp7 f/f mice following 10 days of iPTH or vehicle treatment. Red = Ebf3-CreERT2-tdTomato; green = Bglap-GFP; blue = DAPI. Scale bar = 500 μm (lower magnification) and 50 μm (higher magnification). (D) Quantification of tdTomato+ cells and tdTomato+GFP+ osteoblasts on metaphyseal trabecular and (E) endocortical bone surfaces of wild type and Sp7 f/f mice following 10 days of iPTH or vehicle treatment. (F) Distribution (area fraction) and (G) cell morphology (aspect ratio) analysis of tdTomato+ cells relative to defined distances from bone surfaces in the proximal tibia of wild type and Sp7 f/f mice following 10 days of iPTH or vehicle treatment. (H) Representative flow cytometry plots and analysis of non-hematopoietic Ebf3-CreERT2-labeled tdTomato+ cells (I) and their skeletal stem/progenitor cell surface markers (J) from wild type and Sp7 f/f mice following 10 days of iPTH or vehicle treatment. (**K and L)** RT-qPCR analysis of FACS-isolated Lineage–tdTomato+ cells from wild type and Sp7 f/f mice following 10 days of iPTH and vehicle treatment. Statistical test: Two-way ANOVA with treatment and genotype (**D, E, I, and L**), or treatment and specified distance bins (**F and G**), or treatment and subpopulations (J) as factors followed by Šidák’s multiple comparisons test, and Student’s T-test (K). Data is expressed as mean ± SD, n = 4–5 mice/group.

At baseline, Sp7 f/f mice showed fewer labelled cells on metaphyseal trabecular and endocortical surfaces than wild type controls, with a similar trend at diaphyseal endocortical sites (Figures 4C–E, Supplementary Figure 11A). In wild type mice, iPTH increased labeled progenitors and osteoblasts across all examined surfaces, consistent with prior results. In Sp7 f/f mice, iPTH still recruited labeled progenitors to trabecular surfaces but not endocortical surfaces, and, as expected, failed to increase tdTomato+GFP+ osteoblasts at any site. Central marrow numbers and periosteal progenitor responses were comparable between genotypes, and osteoblast differentiation was not observed on periosteal surfaces in either group (Supplementary Figure 11B–D). Spatially, iPTH increased labeled cell density and elongation within 0–20 μm of endocortical bone in both genotypes; however, within the 0–10 μm zone where pre-osteoblasts and osteoblasts reside, Sp7 deletion specifically impaired the iPTH-stimulated increase in cell number (Figures 4F, G). Flow cytometry showed no difference in overall Lineage–tdTomato+ cell numbers at baseline, and iPTH modestly increased these cells in both genotypes (Figures 4H, I). The iPTH-induced shift from upstream BCSPs to downstream 6C3+ stromal cells was preserved across genotypes, indicating that early progenitor differentiation proceeds independently of Sp7 (Figure 4J). RT-qPCR on sorted labeled cells confirmed reduced Sp7 and Bglap mRNA expression in Sp7 f/f mice, while iPTH-induced changes in Tnfsf11, Ebf3, and Foxc1 were comparable between genotypes (Figures 4K, L).

Overall, these results indicate that the early stages of CAR cell differentiation towards osteoblasts, such as increases in stromal 6C3+ cells and their initial recruitment to bone surfaces, occur independently of Sp7. However, Sp7 plays a key role in controlling the final stages of osteoblast maturation (recruitment close to endocortical bone surfaces and expression of Bglap) at steady state and in response to iPTH.

Non-cell-autonomous recruitment of Ebf3-lineage cells to bone surfaces

Since Ebf3-lineage cells can directly respond to PTH in vitro (Figure 2), and scRNA-seq indicates that its receptor, Pth1r, expression is present in CAR3–6 and osteoblasts (Figure 5A), we next tested whether iPTH induces osteoblastic differentiation of CAR cells in a cell-autonomous manner in vivo by studying Pth1r f/f mice (Ebf3-CreERT2; Pth1r f/f; Rosa26-LSL-tdTomato; Bglap-GFP) and wild type control animals (Ebf3-CreERT2; Pth1r +/+; Rosa26-LSL-tdTomato; Bglap-GFP) (Figure 5B).

Figure 5. — **(A)** Violin plots showing Pth1r expression level across Ebf3-lineage CAR and osteoblast clusters. (B) Overview of lineage tracing strategy to study the role of Pth1r in Ebf3-lineage cells following 10 days of iPTH or vehicle treatment. (C) Representative fluorescent images of the proximal tibia from wild type and Pth1r f/f mice following 10 days of iPTH or vehicle treatment. Red = Ebf3-CreERT2-tdTomato; green = Bglap-GFP; blue = DAPI. Scale bar = 500 μm (lower magnification) and 50 μm (higher magnification). (D) Quantification of tdTomato+ cells and tdTomato+GFP+ osteoblasts on metaphyseal trabecular and (E) endocortical bone surfaces of wild type and Pth1r f/f mice following 10 days of iPTH or vehicle treatment. (F) Distribution (area fraction) and (G) cell morphology (aspect ratio) analysis of tdTomato+ cells relative to defined distances from bone surfaces in the proximal tibia of wild type and Pth1r f/f mice following 10 days of iPTH or vehicle treatment. (H) Representative flow cytometry plots and analysis of non-hematopoietic Ebf3-CreERT2-tdTomato+ cells (I) and their skeletal stem/progenitor cell surface markers (J) from wild type and Sp7 f/f mice following 10 days of iPTH or vehicle treatment. (**K and L)** RT-qPCR analysis of FACS-isolated Lineage–tdTomato+ cells from wild type and Sp7 f/f mice following 10 days of iPTH and vehicle treatment. Statistical test: Two-way ANOVA with treatment and genotype (**D, E, I, and L**), or treatment and specified distance bins (**F and G**), or treatment and subpopulations (J) as factors followed by Šidák’s multiple comparisons test, and Student’s T-test (K). Data is expressed as mean ± SD, n = 4–7 mice/group.

At baseline, labeled cell numbers and osteoblast differentiation were comparable between Pth1r f/f and wild type mice across all examined surfaces (Figures 5C–E, Supplementary Figure 12A). In response to iPTH, progenitor recruitment to trabecular and endocortical surfaces was similarly increased in both genotypes; however, Pth1r deletion blocked iPTH-induced osteoblast differentiation at all sites. Central marrow numbers were unaffected by genotype or treatment, and periosteal osteoblast differentiation was not observed in either group, though iPTH expanded labeled progenitors on periosteal surfaces in wild type but not Pth1r f/f mice (Supplementary Figures 12B–D). Spatially, both genotypes showed comparable iPTH-induced increases in labeled cell density and elongation within 10 μm of bone; at 11–40 μm from bone surfaces, however, Pth1r deletion attenuated these effects (Figures 5F, G). Flow cytometry revealed no iPTH-induced change in overall Lineage–tdTomato+ cell numbers in either genotype (Figures 5H, I). Within this population, Pth1r deletion reversed the direction of the iPTH-induced subset shift: whereas iPTH promoted BCSP contraction and 6C3+ expansion in wild type mice, Pth1r f/f mice showed increased BCSPs and decreased 6C3+ cells (Figure 5J). RT-qPCR confirmed reduced Pth1r mRNA expression and attenuated iPTH-induced upregulation of Tnfsf11 and Vdr in Pth1r f/f mice (Figures 5K, L). Critically, iPTH-induced downregulation of the CAR cell-enforcing transcription factors Ebf3 and Foxc1 was abolished by Pth1r deletion.

Pth1r expression in Ebf3-lineage cells is required for iPTH-induced suppression of CAR cell-enforcing transcription factors, expansion of the 6C3+ subset, and terminal differentiation into Bglap+ osteoblasts, but is dispensable for progenitor recruitment to bone surfaces. This dissociation prompted us to consider non-cell-autonomous mechanisms of intercellular communication that govern this early step in the progenitor response to iPTH.

PTH-stimulated recruitment of Ebf3-lineage cells to bone surfaces requires osteoclasts and is associated with TGFß-responsiveness

Given the key roles of osteoclasts in PTH-regulated bone remodeling⁵² and our focus on inter-cellular communication between CAR cells and other cell types in the bone marrow microenvironment, we performed TRAP staining to define the relationship between Ebf3-lineage cells and osteoclasts. In vehicle-treated mice, labeled cells occupied trabecular surfaces that harbored TRAP+ osteoclasts but lacked Bglap-GFP+ osteoblasts, and iPTH further enhanced this co-localization (Figure 6A). Quantification confirmed that labeled cells were on average closer to TRAP+ osteoclasts than to Bglap-GFP+ osteoblasts, with more cells found within a 50 μm radius of osteoclasts than osteoblasts (Figures 6B, C). Therefore, Ebf3-lineage progenitors preferentially associate with osteoclasts at steady state, an association amplified by iPTH.

To test whether osteoclasts mediate non-cell-autonomous recruitment of Ebf3-lineage cells, we depleted osteoclasts using OPG-Fc (Figure 6D). OPG-Fc reduced serum P1NP and CTX and prevented iPTH-induced increases in both markers, confirming effective osteoclast suppression (Supplemental Figure 13A). Consistent with this, OPG-Fc ablated TRAP+ osteoclasts on bone surfaces and, remarkably, reduced labeled cell numbers on trabecular and endocortical surfaces (Figures 6E–G). While iPTH alone again increased labeled cell recruitment, combined iPTH and OPG-Fc abolished this effect (Figure 6H). Spatially, OPG-Fc reduced labeled cell density within 10 μm of endocortical bone, and co-treatment with iPTH failed to restore the iPTH-stimulated increases in cell density and elongation at 11–40 μm from bone surfaces (Figures 6I, J).

Flow cytometry showed no difference in overall Lineage–tdTomato+ cell numbers across treatment groups, and OPG-Fc did not alter progenitor subset composition at baseline (Figures 6K, L). Notably, despite blocking iPTH-stimulated recruitment to bone surfaces, OPG-Fc did not impair iPTH-induced 6C3+ subset expansion. Thus, osteoclast activity is selectively required for Ebf3-lineage cell recruitment to bone surfaces, while the cell-autonomous effects of iPTH (transcriptional priming and 6C3+ expansion) proceed independently.

MicroCT of L5 vertebrae confirmed that OPG-Fc, iPTH, and combined treatment each increased trabecular bone parameters compared to vehicle (Supplemental Figures 13B, C), and spine histology showed iPTH-stimulated labeled cell recruitment and osteoblast differentiation at vertebral surfaces (Supplemental Figure 13D). Thus, Ebf3-lineage progenitors contribute to osteoblasts in response to iPTH in both the axial and appendicular skeleton.

To understand the cell-extrinsic mechanisms used by osteoclasts to promote iPTH-stimulated recruitment of Ebf3-lineage cells to bone surfaces, we performed single cell RNA-sequencing on FACS-isolated CD45–Ter119–tdTomato+ cells following 10 days of vehicle, iPTH, OPG-Fc, or OPG-Fc plus iPTH treatment (Figure 7A). Unsupervised clustering again identified 6 conserved CAR subsets, 1 pre-osteoblast, and 1 mature osteoblast with similar transcriptomic profiles as above (Figure 7B). For each CAR and pre-osteoblast cluster, we performed cluster-specific DEG analysis comparing vehicle versus PTH, OPG-Fc, or OPG-Fc + PTH treatment (Figure 7C). Notably, OPG-Fc treatment downregulated osteoblastic genes Alpl, Spp1, and Col1a1 and upregulated the skeletal progenitor associated gene Grem1⁷. As expected, PTH-responsive module scores were higher in both PTH and OPG-Fc + PTH conditions relative to vehicle in CAR and pre-osteoblast clusters (Figure 7D).

Figure 7. — **(A)** Harmony-integrated UMAP of single cell RNA-sequencing data from FACS-isolated CD45–Ter119–tdTomato+ cells labeled by Ebf3-CreERT2 following 10 days of iPTH, OPG-Fc, combination or vehicle treatment (Vehicle = 1,155 cells, PTH = 1,638 cells, OPG-Fc = 1,720, and OPG-Fc + PTH = 1,773; n = 5–6 mice/group). Inset shows feature plot of tdTomato mRNA expression across clusters. (B) Violin plots showing mRNA expression levels of CAR cell, pre-osteoblast, and mature osteoblast marker genes in Ebf3-lineage cell clusters. (C) Volcano plots showing the total number of significant differentially expressed genes (DEGs) of all Ebf3-lineage CAR subclusters regulated by PTH, OPG-Fc, or OPG-Fc + PTH (adjusted P-value < 0.05 and |log2FC| > 0.5). (D) PTH-responsive module scores visualized by Feature plots of each Ebf3-lineage CAR cell and pre-osteoblast (top) and violin plots (bottom) across Ebf3-lineage CAR and osteoblast clusters in response to iPTH, OPG-Fc, combination, or vehicle treatment. (E) Violin plots showing the expression levels of canonical PTH target genes and CAR cell fate regulating transcription factor genes across Ebf3-lineage CAR and pre-osteoblast clusters in response to iPTH, OPG-Fc, combination, or vehicle treatment. (F) Venn-diagrams comparing differentially expressed genes altered by PTH versus OPG-Fc + PTH treatment. (G) Gene set enrichment analysis of all significantly up- and down-regulated DEGs by PTH but not OPG-Fc + PTH in Ebf3-lineage CAR clusters using the MSigDB Hallmark 2020 database. The top 5 significantly enriched gene sets are shown (adjusted P-value < 0.05). (H) Violin plots showing the expression levels of TGFß-responsive genes across Ebf3-lineage CAR and pre-osteoblast clusters in response to iPTH, OPG-Fc, combination, or vehicle treatment. (I) TGFß-responsive module scores visualized by Feature plots of each Ebf3-lineage CAR cell and pre-osteoblast (top) and violin plots (bottom) across Ebf3-lineage CAR and osteoblast clusters in response to iPTH, OPG-Fc, combination, or vehicle treatment. Cliff’s delta (δ) = non-parametric effect size used to measure differences between module score distributions. Differential gene expression between treatment and control cells, and treatment interactions within each cluster was assessed using the MAST hurdle model. Black asterisks (*) denotes genes that are significantly differentially expressed in current dataset (adjusted P-value < 0.05 and |log2FC| > 0.5). Red asterisks indicate genes that were significantly regulated by iPTH in the single cell RNA-sequencing dataset from Figure 3, but did not reach statistical significance in the current experiment, likely due to limited power. Blue hashtags (#) indicates a significant antagonistic treatment interaction between OPG-Fc and PTH. (J) Representative confocal images of immunohistochemical staining for phospho-SMAD2/3 in proximal tibia tissue sections from Ebf3-CreERT2; Rosa26-LSL-tdTomato; Bglap-GFP mice following 10 days of iPTH, OPG-Fc, combination or vehicle treatment. Red = Ebf3-CreERT2-tdTomato; green = Bglap-GFP; white = phosphor-SMAD2/3; blue = DAPI. Scale bar = 50 μm.

Concurrent PTH and OPG-Fc treatment did not affect the PTH-induced expression of canonical PTH target genes (Tnfsf11, Mmp13, and Pde10a) and CAR cell transcription factors Ebf1 and Foxc1 (Figure 7E). To identify genes that are involved with PTH-stimulated recruitment of Ebf3-lineage cells to bone surfaces, we performed GSEA on DEGs that were altered only by PTH, but not OPG-Fc + PTH (Figure 7F), and identified ‘Epithelial Mesenchymal Transition’ (Figure 7G). Within this gene set, the expression of several downstream TGFß-responsive genes (Snai1⁵³, Pmepa1⁵⁴, Plod2^55,56 and Tgm2⁵⁷) was significantly increased by PTH and blunted by PTH/OPG-Fc cotreatment (Figure 7H). To assess whether OPG-Fc was attenuating the TGFß-responsiveness of Ebf3-lineage cells to PTH, a TGFß-responsive module score was computed using a literature-curated gene set (Serpine1, Ccn2, Postn, Fn1, Col1a1, Col1a2, Tagln, Timp1, Vim, Itga5, Mmp2, Mmp14, Plod2, Tgm2). In response to PTH, CAR6 and pre-osteoblast clusters showed the largest increase in TGFß-responsive module scores (Figure 7I). Notably, in pre-osteoblasts, OPG-Fc cotreatment inhibited PTH-stimulated induction of the TGFß-responsive module despite having an increase in canonical PTH-target genes (Figure 7D). To confirm whether TGFß-responsiveness of Ebf3-lineage cells was blunted by OPG-Fc co-treatment, we stained for phospho-SMAD2/3 and observed fewer phospho-SMAD2/3+ cells (Figure 7J) with OPG-Fc alone or combined with iPTH treatment.

Together, these findings indicate that osteoclasts are required for the iPTH-stimulated recruitment of Ebf3-lineage cells to bone surfaces and their subsequent differentiation into mature osteoblasts. Osteoclastic bone resorption releases latent TGFß from bone matrix into the local microenvironment⁵⁸. In contrast, early steps in PTH-induced Ebf3-lineage cell differentiation (6C3+ cell expansion, Ebf3 downregulation, reduction of adipogenic factors) occurs independently of osteoclastic bone resorption.

Ebf3-lineage cells are important contributors to overall bone anabolism of iPTH

To assess the relative contribution of Ebf3-lineage cells to iPTH-induced bone anabolism, we returned to Ebf3-CreERT2; Sp7 f/f mice, a model that selectively blocks CAR cell maturation into osteoblasts without affecting other progenitor sources. Wild type and Sp7 f/f mice received three tamoxifen doses at 10 weeks of age, followed by vehicle or iPTH for 4 weeks (Figure 8A).

Figure 8. — (A) Overview of lineage tracing strategy to study the relative contribution of Ebf3-lineage cells following 1 month of iPTH or vehicle treatment in male mice. (C) Representative fluorescent images of the proximal tibia from wild type and Sp7 f/f mice following 1 month of iPTH or vehicle treatment. Red = Ebf3-CreERT2-tdTomato; green = Bglap-GFP; blue = DAPI. White arrow: tdTomato+ cell on bone surface; yellow arrow: tdTomato+ osteocyte. Scale bar = 500 μm (lower magnification) and 50 μm (higher magnification). (C) Quantification of tdTomato+ cells on metaphyseal trabecular and (D) endocortical bone surfaces, (E) tdTomato+ osteocytes in the cortical bone of wild type or Sp7 f/f mice following 1 month of iPTH or vehicle treatment. (F) Distribution (area fraction) and (G) cell morphology (aspect ratio) analysis of tdTomato+ cells relative to defined distances from endocortical bone surface in the proximal tibia of wild type and Sp7 f/f mice following 1 month of iPTH or vehicle treatment. Levels of serum bone formation marker (H) P1NP and resorption marker (I) CTX from wild type and Sp7 f/f mice following 1 month of iPTH or vehicle treatment. (J) Representative sagittal femur microCT images and quantification of (K) trabecular and (L) primary spongiosa bone volume/total volume (BV/TV) in wild type and Sp7 f/f mice following 1 month of iPTH or vehicle treatment. Scale bar = 1 mm.Statistical test: Two-way ANOVA with treatment and genotype (**C, D, E, H, I, K and L**), or treatment and specified distance bins (**F and G**) as factors followed by Šidák’s multiple comparisons test. Data is expressed as mean ± SD, n = 6–8 mice/group.

Sp7 f/f mice showed fewer labeled cells on trabecular and endocortical surfaces at both metaphyseal and diaphyseal sites compared to wild type controls (Figures 8B–D, Supplementary Figure 14A). iPTH increased labeled progenitors on metaphyseal trabecular but not endocortical surfaces in Sp7 f/f mice, and Sp7 deletion blocked iPTH-induced labeled osteocyte accumulation in cortical bone (Figure 8E, Supplementary Figure 14B). Central marrow and periosteal cell numbers were unaffected by genotype or treatment (Supplementary Figures 14C–F). Spatially, the iPTH-induced increase in labeled cell density within 30 μm of endocortical bone was impaired in Sp7 f/f mice, and iPTH-stimulated cell elongation within 0–10 μm was abolished. Flow cytometry showed iPTH increased overall labeled stromal cell numbers in both genotypes; however, the iPTH-stimulated 6C3+ expansion seen in wild type mice was absent in Sp7 f/f mice (Supplementary Figures 14G, H).

As expected, iPTH elevated serum P1NP and CTX in wild type mice; Sp7 deletion blocked the P1NP increase and prevented a significant rise in CTX (Figures 8H, I). Femur microCT showed iPTH-induced gains in BV/TV in the secondary and primary spongiosa of wild type mice that were absent in Sp7 f/f mice (Figures 8J, K), and iPTH-stimulated increases in trabecular thickness and primary spongiosa BMD were similarly blocked (Supplementary Figures 15A, B). Cortical bone parameters were unaffected by genotype or treatment (Supplementary Figures 15C, D). Together, these findings establish Ebf3-lineage cells as major contributors to iPTH-induced bone anabolism, accounting for the bulk of trabecular bone gains at this treatment duration.

Ebf3-lineage cells continue to participate in the anabolic action of PTH in the setting of ovariectomy

To determine whether Ebf3-lineage progenitors contribute to iPTH-induced bone anabolism under estrogen-deficient conditions, we performed OVX in Ebf3-lineage tracing mice. All CAR and osteoblast clusters express Esr1 (estrogen receptor alpha), with highest levels in CAR6 and pre-osteoblasts (Supplementary Figure 16A), supporting the relevance of this model. OVX was performed at 12 weeks of age; 8 weeks later, mice received tamoxifen followed by vehicle, iPTH, OPG-Fc, or combined iPTH and OPG-Fc (Supplementary Figure 16B).

Consistent with findings in eugonadal mice, iPTH increased serum P1NP and CTX, OPG-Fc suppressed both markers, and all treatments increased vertebral trabecular bone mass (Supplemental Figures 17A-C). iPTH-stimulated recruitment of labeled cells to bone surfaces was preserved in older OVX mice, and osteoclast depletion with OPG-Fc again blocked this recruitment and prevented iPTH-induced cell morphology changes (Supplementary Figures 16C-G).

Flow cytometry revealed no difference in overall Lineage–tdTomato+ cell numbers across treatment groups (Supplementary Figures 16H, I). Within this population, however, iPTH-induced subset dynamics differed markedly from young eugonadal mice: rather than expanding the 6C3+ subset, iPTH decreased 6C3+ cell numbers in older OVX mice; OPG-Fc co-treatment also reduced 6C3+ cells, though to a lesser degree (Supplementary Figure 16J). Thus, while iPTH-induced labeled cell recruitment and osteoblastic differentiation are broadly shared between young eugonadal and older OVX mice, age- and/or estrogen-dependent differences may impact CAR cell subset composition and its response to iPTH.

CAR cells in teriparatide-treated postmenopausal women show reduced expression of lineage-enforcing transcription factors and increased TGFß-responsive genes

We next investigated whether the transcriptomic changes by PTH on Ebf3-lineage cells also occur in human CAR cells. We performed single cell RNA sequencing on FACS-isolated non-hematopoietic (CD45–CD235a–CD14–) stromal cells obtained from bone marrow aspirates of postmenopausal women receiving teriparatide, a PTH analogue, for 3 months⁵⁹. Using Harmony, we integrated these data with a single cell atlas composed of multiple recently published datasets derived from drug-naïve human bone marrow (Figure 9A)⁶⁰. After unsupervised clustering on this integrated dataset, 3 CAR and 1 pre-osteoblast clusters were identified (Figure 9B). CAR clusters were identified based on the highest relative expression of CXCL12, LEPR, EBF3, FOXC1, and KITLG. Pre-osteoblasts were identified based on the highest levels of LIMCH1, SP7, RUNX2, SPP1, IBSP, COL1A1, BGLAP, and PTH1R (Figure 9C and D).

Figure 9. — **(A)** UMAP visualization of human bone marrow single cell RNA-sequencing data following (B) Harmony-integration of bone marrow aspirate samples from postmenopausal women treated with teriparatide for 3 months (11,090 cells; n = 2 patient samples) and a naïve human bone marrow atlas. (C) Feature plots showing the expression CAR cell marker genes. (D) Violin plots showing mRNA expression levels of CAR cell and pre-osteoblast genes. (E) Volcano plot showing the total number of significant differentially expressed genes between each teriparatide-treated vs atlas CAR clusters (adjusted P-value < 0.05 and |log2FC| > 0.5). (F) Violin plots showing the expression levels of CAR cell genes, (G) adipogenic genes, (H) and osteogenic and bone matrix genes, across each teriparatide-treated versus atlas CAR clusters. (I) Gene set enrichment analysis of all up- and down-regulated differentially expressed genes identified in teriparatide-treated CAR clusters using the MSigDB Hallmark 2020 database. (J) Violin plots of TGFß-responsive module scores of teriparatide-treated versus atlas CAR clusters. Cliff’s delta (δ) = non-parametric effect size used to measure differences between module score distributions. (K) Violin plots showing the expression levels of TGFß-responsive genes between teriparatide-treated vs atlas CAR clusters. Differential gene expression between teriparatide-treated and atlas cells within each CAR cluster was assessed using the MAST hurdle model. Asterisk (*) denotes significance (adjusted P-value < 0.05 and |log2FC| > 0.5).

To assess the transcriptomic effects of teriparatide treatment in CAR cells, we performed cluster-specific DEG analysis (teriparatide versus atlas) in each CAR cluster (Figure 9E). Like the effects of iPTH in Ebf3 lineage cells in mice, teriparatide treatment in women led to marked decreases in the expression of hematopoiesis supportive genes KITLG and CXCL12, and CAR cell-enforcing transcription factors EBF3 and FOXC1 (Figure 9F). Furthermore, the expression of adipogenesis-related genes such as PPARG, ADIPOQ, APOE, and PLIN2 were reduced (Figure 9G) while osteogenic genes RUNX2, RUNX3, COL1A1, and COL1A2 were increased in teriparatide-treated CAR cells (Figure 9H). GSEA on all the teriparatide-upregulated DEGs identified the ‘Epithelial Mesenchymal Transition’ gene set as most enriched (Figure 9I). Within this gene set, several canonical TGFß-signaling target genes such as SERPINE1, TGFBI, CCN2, CCN1, INHBA, POSTN, and TGFß-responsive genes FN1, ACTA2, TAGLN, and PLOD2 were increased in teriparatide-treated CAR cells (Figure 9J). Consistent with this finding, increased TGFß-responsive module scores were noted in teriparatide-treated CAR cells versus naïve atlas populations (Figure 9K).

Taken together, these results establish CAR cells as a critical cellular target of iPTH. iPTH promotes osteoblastic differentiation of CAR cells through coupled cell-intrinsic and cell-extrinsic mechanisms. PTH directly acts on CAR cells to suppress the expression of critical transcription factors responsible for suppressing osteogenic differentiation, thereby priming these cells for osteogenic differentiation. Cell-extrinsically, PTH stimulates osteoclastic bone resorption, likely driven by RANKL induction by mature osteoblasts and osteocytes, resulting in enhanced TGFß signaling that may promote recruitment of these primed progenitors to bone surfaces. Once localized to bone surfaces, PTH again acts directly on CAR cell-derived pre-osteoblasts to drive their maturation into osteoblasts and osteocytes (Figure 10).

Figure 10. — CAR cells express the transcription factors Ebf1, Ebf3, and Foxc1, which maintain their hematopoiesis-supportive phenotype. iPTH directly acts on CAR cells via Pth1r to suppress CAR cell-maintaining transcriptional programs, thereby priming osteoblastic differentiation and inhibiting adipogenic fate (cell-autonomous effect). In parallel, iPTH stimulates osteoclastic bone resorption, through induction of RANKL in multiple cell types including mature osteoblasts, osteocytes, and CAR cells. Bone resorption releases matrix-derived growth factors, such as TGFβ, which recruit CAR cells to bone surfaces (non-cell-autonomous effect). Once localized to bone surfaces, iPTH again directly targets CAR cell-derived pre-osteoblasts via Pth1r to promote Sp7-dependent maturation into osteoblasts and terminally differentiated, matrix-embedded osteocytes (cell-autonomous effect). Solid lines indicate cell-autonomous mechanisms used by PTH to drive Ebf3 lineage cell osteoblast differentiation, and dashed lines indicate non-cell-autonomous mechanisms employed by PTH to recruit CAR cells to bone surfaces.

Discussion

This study identifies Ebf3-lineage CAR cells as major cellular contributor in the bone anabolic actions of iPTH at skeletal maturity. In contrast to growth-associated progenitors residing near the growth plate, Ebf3-lineage cells are distributed throughout bone and can give rise to osteoblasts on trabecular and endocortical surfaces. Our studies demonstrate that iPTH promotes the osteoblastic differentiation of CAR cells. As early response to PTH, the CAR cell phenotype is destabilized in a cell-autonomous manner through loss of the expression of crucial CAR-cell maintaining transcription factors Ebf3, Ebf1, and Foxc1, thereby permitting their osteogenic-commitment while reducing adipogenesis. iPTH also acts directly on Ebf3-lineage CAR cells to expand the 6C3+ CAR cell subpopulation to maintain the bone marrow niche. A key event in PTH-induced osteoblastic differentiation of CAR cells involves their recruitment to bone surfaces and morphologic change from a reticular to more elongated shape, a process that requires osteoclasts and may be mediated by TGFß signaling. Preventing Ebf3-lineage CAR cells from contributing to the osteoblast pool via Sp7 deletion severely blunts the PTH-induced gain in trabecular bone mass. Critically, the gene regulatory network controlled by iPTH in murine Ebf3-lineage CAR cells is conserved in human CAR cells from teriparatide-treated patients.

CAR cell-autonomous effects of PTH

Pth1r mediates the cell-autonomous effects of iPTH in Ebf3-lineage cells, driving 6C3+ CAR cell expansion, suppressing CAR cell-maintaining transcriptional programs, and promoting osteoblastic differentiation, but is dispensable for their recruitment onto bone surfaces. The long-term consequences of Pth1r deficiency in Ebf3-lineage on the bone marrow microenvironment, both at steady state and in response to iPTH, remain to be determined. CAR cells play a crucial role in supporting hematopoiesis. Deletion of Pth1r in Lepr-expressing cells and their progeny appeared to have only a modest effect on hematopoiesis⁶¹; however, the efficiency of Cre-dependent recombination of the Pth1r floxed allele in these cells was not assessed. Pth1r ablation in mesenchymal progenitors using Prx1-Cre⁶² resulted in a marked expansion of bone marrow adipocytes and increased bone resorption⁸. Given that iPTH-induced osteoblastic differentiation of CAR cells is blocked in the absence of Pth1r, whether this alters their adipogenic commitment or long-term hematopoietic supportive function following long-term iPTH administration warrants further study.

CAR cell-extrinsic effects of PTH

PTH stimulates osteoclastic bone resorption, leading to enhanced TGFß signaling which may promote the recruitment of CAR cells to bone surfaces. This bone resorptive effect is likely driven by RANKL induction by osteoblasts and osteocytes, which are sources that are spatially localized at bone surfaces and therefore optimally positioned to promote osteoclast activity. CAR cells have been shown to be major producers of RANKL for osteoclast development^24,63,64, and our data indicate that PTH also directly induces RANKL expression in CAR cells in a Pth1r-dependent manner (Figure 5L). However, ablation of Pth1r in CAR cells did not impair their recruitment to bone surfaces in response to PTH, suggesting that CAR cell-derived RANKL is not the predominant driver of this osteoclast-dependent recruitment process. Bone matrix is abundant in growth factors including TGFß, IGFs, BMPs, WNTs which can be released during osteoclastic resorption⁶⁵. Out of these factors, TGFß1 has been implicated as a chemoattractant involved in mesenchymal stromal progenitor cell recruitment to bone surfaces during physiologic and stimulated bone remodeling^21,66,67. Defining specific TGFß ligand and receptor isoforms responsible for osteogenic CAR cell recruitment to bone surfaces represents an important priority for future research. Furthermore, whether this migratory process is a mediated directly by osteoclasts or indirectly through their bone resorptive activity requires further investigation. Our data suggests that modulating the TGFß-responsiveness of CAR cells to enhance their recruitment to bone surfaces may be a potential therapeutic strategy to promote the recruitment of these critical osteoblast progenitors to bone surfaces.

Relative contribution of CAR cells to the bone anabolism of iPTH

CAR cells, and not growth plate-associated progenitors, are the major osteoblast progenitor population that contributes to the bone anabolic actions of iPTH. Despite blocking the CAR cell contribution to the osteoblast pool via Sp7 deletion with Ebf3-CreERT2, we note that Bglap-GFP+ osteoblasts persist on trabecular and endocortical bone surfaces and are increased with iPTH treatment (Figure 8B), thus indicating that other sources of osteoblast progenitors are involved. Quiescent bone lining cells can be stimulated by iPTH to become active osteoblasts⁶⁸. However, from these studies it appears that CAR cell-derived osteoblasts are critical to iPTH-induced bone gains at this treatment duration (Figure 8H-L).

Functional heterogeneity among CAR cells

At present, we do not know why some CAR cells in the central marrow remain unrecruited to bone surfaces in response to PTH. CAR cell proliferation is low (approximately 5% EdU+; Supplementary Figure 6B), consistent with previous reports^5,11, and is unaffected by iPTH. Through single cell RNA-sequencing analysis of Ebf3-lineage cells, we identified transcriptomic heterogeneity amongst Ebf3-lineage CAR cells, despite all subsets expressing high levels of Cxcl12 and other canonical markers such as Lepr and Ebf3. The most notable contrast is between CAR1 and CAR6 which have transcriptomic profiles that are consistent with previous reports of an adipogenic and osteogenic-primed subsets, respectively^28,29,41. Even though all subsets exhibit PTH-induced increases in target genes Tnfsf11, Mmp13, and Vdr, CAR1 showed the lowest increase in PTH-response module score and no changes in adipogenic genes (Adipoq and Apoe) compared to all other CAR subsets (Figure 3). Thus, it may be that CAR1 cells represents a CAR subset that is not recruited to bone surfaces following iPTH. In addition, CAR1 is the least enriched in the expression of actin cytoskeleton remodeling or migration-related genes (Dock1, Trio, Arhgap35 (Supplementary Figure 8C)), further supporting the notion that this CAR subset may not be migratory. Unfortunately, differential gene expression analysis between the 6 CAR subsets did not reveal specific/exclusive single markers that can be used to identify spatially where these CAR subsets reside in the marrow. Further investigation using spatial transcriptomic approaches may shed additional light onto anatomic and functional differences between CAR cell subsets.

Clinical data for OPG-Fc + PTH

Clinical data from post-menopausal women showed that concomitant intermittent teriparatide and denosumab (akin to OPG-Fc) therapy is more effective at increasing bone mineral density than either medication alone⁶⁹. Our in vivo results are consistent with this (Supplementary Figure 13 and 17). iPTH acts to stimulate both modeling- (osteoclast-independent) and remodeling-based (osteoclast-dependent) bone formation, whereas OPG-Fc inhibits osteoclastic bone resorption. Therefore, changes in osteoblastic activity observed with OPG-Fc plus iPTH combination treatment likely reflect modeling-based bone formation. Clinical data showing that serum osteocalcin was less suppressed in combination group compared to denosumab monotherapy⁶⁹ support this notion. Notably, we observe that some tdTomato+ cells persist on endocortical surfaces at the mid-diaphysis with OPG-Fc treatment and that tdTomato+ cell numbers increase modestly with concurrent iPTH treatment compared to OPG-Fc alone (Figure 6H and Supplementary Figure 16E). Future studies are needed to address the mechanisms dictating how CAR cells are recruited to bone surfaces that are undergoing modeling versus remodeling.

Future Directions

This present study focuses primarily on the bone anabolic actions of iPTH on CAR cell dynamics, given its translational relevance to understanding the mechanisms of PTH-based osteoporosis therapy. The response of CAR cells to continuous (rather than intermittent) hyperparathyroidism remains to be determined. The transcription factors Runx1 and Runx2 play crucial regulatory roles in preventing the fibrotic conversion of CAR cells⁷⁰. The CAR cell-intrinsic or extrinsic mechanism underlying this phenomenon are unclear. Given that bone marrow fibrosis occurs in the setting of chronic hyperparathyroidism and the potential fibrotic fate of CAR cells, the role of CAR cells in this disease warrants further investigation.

Ebf3-CreERT2 also marks a PTH-responsive subset of periosteal cells (Figure 3F). Periosteal Ebf3-lineage cells do not differentiate into Bglap+ osteoblasts during the length (2 months) of tracing these cells at baseline or in response to iPTH. Osteocytes descended from Ebf3-lineage cells are preferentially localized on the endosteal half of cortical bone rather than periosteal side, further supporting the notion that Ebf3-lineage periosteal cells are not osteogenic-committed. The role of these periosteal Ebf3-expressing cells remains to be determined. Ebf3-lineage periosteal cells were identified based on expression of Prg4, Mfap5, and Clec3b, the latter two identify periosteal cells²⁸. Given the established role of the periosteum in bone repair and regeneration, whether these periosteal cells respond to perturbations, such as bone fracture or mechanical stress, warrants future investigation.

Study Limitations

First, although PTH acts directly on CAR cells to suppress expression of critical transcription factors (Ebf3, Ebf1, and Foxc1), the specific signaling mechanisms that mediate these gene expression changes⁷¹ are not known. Second, we employed two different lineage tracing systems, Ebf3- and Sox9-CreERT2, to study the effects of iPTH on two major osteoblast progenitor populations. In comparable tamoxifen-pulse chase time points in 10-week-old mice, we observed that Ebf3-CreERT2 labeled cells more efficiently gave rise to osteoblasts (Figure 1B) than Sox9-CreERT2 (Supplementary Figure 11F) after 10 days of iPTH. However, we were unable to directly assess the relative contribution between these two osteoblast progenitor populations in the same animal due to lack of appropriate dual recombinase-driven lineage tracing systems. Third, estrogen deficiency (induced by ovariectomy) led to a reduction of the 6C3+ subpopulation within Ebf3-lineage cells following iPTH treatment. This finding lies in contrast to our results in younger/eugonadal mice in which iPTH expands this subpopulation. The 6C3+ subpopulation represents the parental CAR cell population (Cxcl12-GFP⁺Ebf3-tdTomato⁺; Supplementary Figure 5 and 8); this finding suggests that the CAR cell pool may become depleted over time with prolonged iPTH treatment. If true, this may be a potential mechanism that could explain why the bone anabolic efficacy of iPTH-based therapies slowly wane over time^72,73. To rigorously test this hypothesis, future experiments controlled with sham-operated animals should be included to examine the effects of ovariectomy on CAR cell dynamics. Finally, due to the clinical study (DATA-Biopsy) design in which marrow aspirates were collected from teriparatide-treated post-menopausal women with osteoporosis, a control (placebo) treatment group was not included. Thus, we employed a reference “atlas” dataset of similarly-processed marrow cells as an aggregate untreated control population. Future prospective studies are needed to examine the effects of teriparatide (versus baseline or no treatment) on human CAR cells over time.

Summary

We establish that CAR cells are a critical subset of osteoblast progenitors that drive iPTH-induced bone anabolism through both cell-autonomous and non-autonomous mechanisms. We report key transcriptional mechanisms through which iPTH directs murine and human CAR cell fate, suppressing adipogenic programming in favor of osteogenic commitment. We demonstrate that osteoclastic bone resorption is required for iPTH-stimulated recruitment of CAR cells to bone surfaces, suggesting that TGFß-responsiveness may have an important role in this process. Together, these findings define a unified mechanism of bone anabolic PTH action in which intrinsic transcriptional priming of CAR cells towards osteogenic fate converges with osteoclast-mediated ‘niche activation’ to drive osteoblast differentiation. Targeting CAR cell TGFß-responsiveness may represent a therapeutic strategy to enhance recruitment of this critical class of osteoblast progenitors to bone surfaces and improve the efficacy of anabolic treatment for osteoporosis.

Methods

Mouse strains

Mice in this study were generated from Ebf3-CreERT2¹¹, Sox9-CreERT2⁷⁵, Bglap-GFP⁷⁶, Cxcl12-GFP⁷⁷, Pth1r-floxed⁷⁸, Sp7-floxed⁷⁹, Rosa26-loxP-stop-loxP-tdTomato (Ai14; JAX007914). All mice were backcrossed to C57BL/6J background for at least 10 generations. Mice were genotyped by PCR. Both male and female mice were used in all experiments unless otherwise indicated in the figure legends. Mice were housed in a pathogen-free, temperature- and humidity-controlled facility with husbandry maintained and provided by the Center of Comparative Medicine at the Massachusetts General Hospital. All animal experiments were approved and carried out in accordance with the guidelines issued by the Institutional Animal Care and Use Committees at the Center for Comparative Medicine at Massachusetts General Hospital under approved animal use protocol (2004N000176). At sacrifice, animals were euthanized by CO₂ inhalation, followed by cervical dislocation.

Tamoxifen administration

Tamoxifen (Sigma-Aldrich, T5648) was first dissolved in 100% ethanol, then an equal volume of sunflower seed oil (Sigma-Aldrich, S5007) was added and vigorously mixed. The tamoxifen-ethanol-oil mixture was incubated at 65°C until ethanol was completely evaporated and then stored at 37°C until use. Mice were given one dose of tamoxifen (2 mg) via intraperitoneal injection to induce Cre-dependent recombination in labeled cells for lineage tracing experiments, except for experiments in Figure 8, mice were given 3 doses of tamoxifen (2 mg) every other day to increase labelling and Sp7 deletion efficiency. For no tamoxifen controls, the same volume of sunflower oil (without tamoxifen) was injected.

Parathyroid hormone and OPG-Fc administration

Mice were given daily subcutaneous injections of vehicle (10 mM citric acid, 150 nM NaCl, 0.05% Tween-80) or human PTH 1–34 (100 μg/kg). For OPG-Fc treatment⁸⁰, mice were given once per week subcutaneous injections of vehicle (phosphate buffered saline; PBS) or OPG-Fc (5 mg/kg).

Histology

Harvested bone samples were immediately fixed with 4% paraformaldehyde for 24 hours at 4°C, then decalcified in 15% EDTA solution (changed every other day) for 14 days. Decalcified samples were cryoprotected in 30% sucrose/PBS solution for 24 hours at 4°C, followed by a 30% sucrose/PBS:OCT (1:1) solution for 24 hours. Samples were embedded in OCT compound (TissueTek, Sakura), and frozen on a block of dry ice. Frozen sample blocks were cryosectioned (Leica CM1850) serially at 10 μm thickness and adhered to positive charged glass slides (Fisherbrand Superfrost plus).

Cell quantification at distinct skeletal compartments of the tibia was performed on fluorescent images. At each skeletal site, a 500 μm x 750 μm region of interest (ROI) was defined. At the proximal metaphysis, the ROI for central marrow and cortical bone analysis was positioned 500 μm below the growth plate. At the midshaft diaphysis, the ROI for central marrow and cortical bone analysis was positioned mid-diaphysis. Cortical bone parameters (endosteal surface/periosteal surface/osteocyte number) were assessed at the compression side of the tibia. Additional details of these regions are depicted in Supplementary Figure 1C. Two serial sagittal sections per mouse corresponding to the central sagittal plane of the tibia were quantified by two different scientists blinded to the condition. Mean values per mouse were used in statistical analysis.

MATLAB image analysis

Fluorescence images of the metaphysis of the tibia were analyzed using a custom pipeline implemented in MATLAB to quantify the spatial distribution of Ebf3-lineage tdTomato+ cells relative to bone surfaces. Multi-channel images were imported and separated into nuclei (DAPI), tdTomato+ signal (distance source), and bone surfaces (distance targets). The nuclei channel was used to generate a binary mask defining the bone marrow space and excluding bone and tissue artifacts. Bone surfaces were identified from the edges of the marrow mask and classified into epiphyseal/growth plate, endocortical bone, and trabecular bone regions. TdTomato+ cells were segmented from the fluorescence channel following background correction and size/morphology filtering. Distances between tdTomato+ cells and endocortical bone surface were computed using Euclidean distance maps constrained to the marrow mask. All spatial measurements were converted to micrometers (1.8872 μm per pixel) based on microscope calibration. Each segmented tdTomato+ cell was assigned to the nearest endocortical bone surface, and assignment quality was evaluated using a clustering-based score. tdTomato+ cell distributions (area fraction) and morphological features (aspect ratio) relative to specified distance bins from the endocortical bone surface were quantified and exported for downstream statistical analysis.

Microscopy

Multicolor fluorescent images were captured using the Keyence BZ-X700 fluorescent microscope, with preconfigured filter settings for DAPI/FITC/Texas Red/Cy5. Custom filter for ELF 97 (excitation/emission = 350 nm/540 nm) was used to detect fluorescent TRAP staining. Z-stack confocal images were acquired using Zeiss LSM 800 Airyscan confocal microscope with 4 lasers and corresponding filters for DAPI/GFP/tdTomato/Cy5. Representative images of at least three independent biological samples are shown in the figures. Quantification of fluorescent images was performed using Image J software (National Institute of Health).

Immunohistochemistry

Immunohistochemical staining of mouse tibia, tissue section underwent antigen retrieval using proteinase K solution (Dako, S3020) for 5 minutes at room temperature, permeabilized with 0.2% TritonX/PBS for 30 minutes at room temperature, blocked with 3% bovine serum albumin (BSA) in 0.2% TritonX/PBS solution for 1 hour at room temperature, followed by overnight incubation at 4°C with rat anti-endomucin (Emcn) monoclonal antibody (1:100, Santa Cruz Biotechnology, sc65495), rat anti-CD31 monoclonal antibody (1:100, Bio-Rad, MCA2388), rabbit anti-perilipin A/B polyclonal antibody (1:100, Sigma-Aldrich, P1873), or rabbit anti-phospho-SMAD2/SMAD3 polyclonal antibody (1:100, ThermoFisher, PA5–110155). Negative controls consisted of isotype-matched primary antibodies at the same protein concentration. The next day, sections were incubated with Alexa Fluor 647-conjugated goat anti-rat (Invitrogen, A-21245) or anti-rabbit (Invitrogen, A-21247) secondary antibodies at room temperature for 3 hours. Nuclei were stained with DAPI (4′,6-diamidino-2-phenylindole, 1 μg/ml, Invitrogen, D1306) for 5 minutes at room temperature, then cover slipped and mounted using Fluoromount (SouthernBiotech, OB100–01).

Bone single cell isolation

A single-cell suspension of bone cells was prepared using a slightly modified version of previously described protocol²⁸. Briefly, femurs and tibias from a single mouse were harvested and stripped of muscle tissue. Whole bones from each mouse were placed in a mortar containing 1 ml of Hank’s Balanced Salt Solution (HBSS, Sigma-Aldrich, H6648) and mechanically homogenized using a pestle. Thereafter, this bone-cell solution was transferred into a single well of a 12-well culture dish containing HBSS and 2 Wünsch units of Liberase^™ (Roche, 5401127001) and placed on a shaking incubator (ThermoMixer C, Eppendorf) at 37°C for 20 minutes. After each round of enzymatic digestion, cells were mechanically triturated using a P1000 pipette with tip end cut off, then cell supernatant was collected and filtered through a 70 μm cell strainer (Falcon, 352350) into a 50 mL tube on ice. The bone fragments were digested a second time with Liberase^™ and the released cells were collected as described above. After this, the bone fragments were incubated in PBS containing 5 mM EDTA and 0.1% BSA for 20 minutes at 37°C for 20 minutes, and then collected the released cells as described above. Remaining bone fragments were subjected to alternating exposures of either Liberase^™ or 5 mM EDTA solution to yield a total of 4 Liberase^™ digestions and 3 EDTA incubations. After 7 rounds of digestion, the collected cells were pelleted and resuspended in sorting buffer (2% fetal bovine serum in HBSS; FACS buffer).

Fluorescence activated cell sorting (FACS)

For FACS isolation of mouse tdTomato+ stromal cells for single cell RNA-sequencing experiments, mouse bone single cell suspension collected after 7 rounds of serial enzymatic digestion underwent magnetic activated cell depletion of hematopoietic cells using mouse lineage depletion kit (Miltenyi Biotec, 130–110–470) combined with CD45 microbeads (Miltenyi Biotec, 130–052–301) according to the manufacturer’s instructions. After magnetic depletion, cells were incubated in the dark for 30 minutes on ice with the prepared primary antibody cocktail. Cells were washed 2 times and resuspended in FACS buffer with DAPI (1 μg/ml, Invitrogen, D1306). FACS was performed using BD FACSAria^™ II equipped with 5 lasers (BD bioscience). Stromal tdTomato+ fraction identified as CD45–Ter119–tdTomato+ was collected for 10X single cell RNA-sequencing.

For FACS isolation of human bone marrow stromal cells for single cell RNA-sequencing experiments, human bone marrow cells were incubated with blocking buffer (0.5% BSA in PBS; 1:50, BD bioscience, 564220) for 10 minutes on ice, then incubated in the dark for 30 minutes on ice with primary antibody cocktail with cell viability dye Calcein AM (5 μM; ThemoFisher, C3099). Cells were washed 2 times and resuspended in sorting buffer (0.5% BSA in PBS). FACS was performed using Bigfoot Spectral Cell Sorter equipped with 7 lasers (Invitrogen). Stromal fraction identified as CD45–CD235a–CD14– was collected for 10X single cell RNA-sequencing.

FACS analysis for mouse skeletal stem and progenitor subpopulations was performed on mouse bone single cell suspension collected after 7 rounds of serial enzymatic digestion. Primary antibody dilutions were prepared in Brilliant Stain buffer (BD bioscience, 563794), and cells were incubated in the dark for 30 minutes on ice with the prepared primary antibody cocktail. Cells were washed 2 times and resuspended in FACS buffer with DAPI (1 μg/ml, Invitrogen, D1306). FACS was performed using Bigfoot Spectral Cell Sorter equipped with 7 lasers (Invitrogen). Mouse bone marrow cells and UltraComp eBeads (Invitrogen, 01–333–42) were used to set initial fluorochrome compensation. Fluorescence minus one (FMO) controls were used for additional compensation and to assess background levels for each stain. Gates were drawn as determined by internal FMO controls to define positive and negative populations for each cell surface marker. Per biological sample, 5–8 million events were recorded and analyzed using FlowJo (10.1v). Full details of the FACS gating strategies shown in Supplementary Figure 4A.

FACS antibodies

Antibodies for FACS of mouse samples included CD45 (eBioscience, 17–5921–82; BD bioscience, 568344), Ter119 (eBioscience, 17–0451–82; BD bioscience, 741736), CD31 (BD bioscience, 551262), CD200 (BD bioscience, 565547), CD105 (BioLegend, 120410), Thy1 (BD bioscience, 565257), 6C3 (eBioscience, 46–5891–82), CD51 (BD bioscience, 740946). Antibodies for FACS of human samples included CD235ab (BioLegend, 306614), CD14(BioLegend, 301824), CD45 (BioLegend, 304052), CD31 (BioLegend, 303122).

Cell proliferation and apoptosis flow cytometric assay

To evaluate cell proliferation, 1.5 mg of 5-ethynyl-2’-deoxyuridine (EdU) (Invitrogen, A10044) was dissolved in PBS and administered via intraperitoneal injection to mice 12 hours before sacrifice. Bone cells were isolated as described above in the bone single cell isolation section. Proliferative cells (EdU+) were detected using the Click-iT^™ Alexa Fluor 488 flow cytometry assay kit (Invitrogen, C10425). Apoptotic cells were detected using the FITC Annexin V Apoptosis detection kit (BioLegend, 640922). Both kits were used following the manufacturer’s instructions.

In vitro culture of Ebf3-lineage cells

For intracellular cAMP quantification, FACS-isolated mouse tdTomato+ stromal cells were seeded in tissue culture 96-well plate at a density of 1 × 10⁴ cells per well, with 3 technical replicates per mouse sample. For phospho-CREB immunocytochemistry, FACS-isolated stromal cells were seeded in an 8-well cell culture chamber slide at a density of 5 × 10² cells per well. For phospho-CREB flow cytometry analysis, FACS-isolated stromal cells per mouse sample were seeded in a 24-well plate at a density of 5 × 10⁴ cells per well. Cells were cultured at 37°C and 5% CO₂ in low-glucose DMEM with GlutaMAX supplement (Gibco, 10567022) and 10% mesenchymal stem cell-qualified FBS (Gibco, 12662029) for 7 days before PTH stimulation experimentation.

Intracellular cAMP assay

Prior to stimulation, cells were washed HB solution (1M HEPES, 7.5% BSA in HBSS). Cells were then treated for 30 minutes at room temperature with either vehicle, human PTH 1–34 (100 nM), or forskolin (30 μM) in a IBMX buffer solution (0.2M IBMX and 5N NaOH in HB solution). After, IBMX buffer solution was aspirated and 50 mM HCL solution was added to the cells. Cell solution was used for intracellular cAMP quantification by cAMP Competitive ELISA Assay Kit (Abcam, ab290713) following manufacturer’s protocol. The signal was read on the EnVision plate reader (Perkin Elmer, 2104) at 405 nm.

Phospho-CREB assays

For phospho-CREB immunocytochemical staining, post-stimulated cells were fixed with 4% paraformaldehyde solution at room temperature for 10 minutes and then washed 3 times with PBS. Cells were permeabilized with 0.2% Triton-X/PBS for 15 minutes, then blocked with 3% BSA in 0.2% TritonX/PBS at room temperature for 1 hour, followed by overnight incubation at 4°C with rabbit anti-phospho-CREB (1:200, Cell Signaling Technology, 9198S). Negative controls consisted of isotype-matched primary antibodies at the same protein concentration. The next day, cells were incubated with Alexa Fluor 488-conjugated goat anti-rabbit (Invitrogen, A-11008) at room temperature for 3 hours. Nuclei were stained with DAPI (4′,6-diamidino-2-phenylindole, 1 μg/ml, Invitrogen, D1306) for 5 minutes at room temperature, then cover slipped and mounted using Fluoromount (SouthernBiotech, OB100–01).

For phospho-CREB flow cytometric analysis, post-stimulated cells were trypsinized and immediately fixed with 4% paraformaldehyde solution at room temperature for 10 minutes. Cells washed with PBS and pelleted, and then permeabilized with ice-cold 90% methanol/PBS on ice for 10 minutes. After, cells were washed with PBS and incubated with anti-phospho-CREB-Alexa 488 conjugated antibody (1:50, Cell Signaling Technology, 9187S) at room temperature for 1 hour. Negative controls consisted of isotype-matched primary antibodies at the same protein concentration. Cells were washed 3 times with PBS and resuspended in PBS for flow cytometry.

Single cell RNA-sequencing analysis

Cell viability and numbers were assessed using a Luna FX7 fluorescent cell counter (Logos Biosystem) before loading onto the Chromium GEM-X Single Cell 3’ Chip (10X Genomics). Single cell RNA-sequencing cDNA libraries were constructed using the Chromium GEM-X Single Cell 3’ Reagent Kit according to the manufacturer’s instructions, followed by sequencing of the cDNA libraries using the Illumina NextSeq 2000 instrument. Sequencing depth of all libraries ranged from 400–800 million reads to reach a target of 25,000 mean reads per cell. Raw reads obtained from scRNA-seq experiments were demultiplexed, aligned to the custom mouse mm10 genome (with tdTomato gene inserted) using Cellranger toolkit (10X Genomics). Analyses were performed using Seurat⁸¹ in R and Scanpy⁸² (scVelo⁸³ and CellRank⁴²) in Python. Cells were filtered out based the following criteria: less than 200 genes and more than 4,000 genes per cell and with more than 20% mitochondrial read content. After quality control filtering of low quality cells from each library, all libraries were then integrated using Harmony²⁷ to correct for batch effects, followed by standard single cell processing steps including normalization, identification of highly variable genes across single cells, scaling based on number of Unique Molecular Identifier (UMI), dimensionality reduction (PCA and UMAP), unsupervised clustering, and identifying cluster-specific differentially expressed genes using the nonparametric Wilcoxon rank sum test included in Seurat. Afterwards, contaminating hematopoietic cell clusters with low or no tdTomato mRNA gene expression were filtered out and the remaining clusters were re-clustered by performing the following single cell processing steps: identifying highly variable genes, scaling, dimensionality reduction, and unsupervised clustering at 0.2 resolution parameter, resulting in the final UMAP shown in the figures. Clustree⁸⁴ was used to assess cluster stability across 0.2 to 1 resolution parameter to guide the selection of a biologically meaningful clustering resolution. Within each cluster, cells from the vehicle- and PTH-treated libraries were randomly down-sampled to equal numbers prior to performing differential gene expression analysis using the MAST hurdle model⁷⁴. The list of significant differentially expressed genes (adjusted P-value < 0.05 and |log2FC| > 0.5) for each cluster was then used for gene set enrichment analysis (GSEA) by Enrichr⁸⁵ (https://maayanlab.cloud/Enrichr/) to identify significantly enriched gene sets (adjusted P-value < 0.05).

For the RNA velocity-based fate probability analysis, spliced and unspliced RNA matrices (loom files) for each library were generated using velocyto⁸⁶ according to the standard protocol (https://velocyto.org/velocyto.py/tutorial/index.html). Next, RNA velocity matrices were then computed using the standard scVelo⁸³ workflow with default parameters as described in its documentation (https://scvelo.readthedocs.io/en/stable/). Cell-cell transition matrices and cell fate probabilities were then inferred using the CellRank⁴² framework following the recommended pipeline (https://cellrank.readthedocs.io/en/latest/notebooks/tutorials/estimators/700_fate_probabilities.html).

To assess the transcriptomic effects of PTH on human CAR cells, single cell RNA-sequencing libraries from FACS-isolated human bone marrow stromal cells (described below) were integrated using Harmony with a single cell atlas composed of multiple recently published datasets derived from naïve human bone marrow⁶⁰. Thereafter, the integrated dataset underwent quality control and single cell processing workflow (described above) for downstream analysis.

RNA isolation and RT-qPCR

Total RNA was isolated from FACS-isolated tdTomato+ cells using PicoPure RNA isolation Kit (Applied Biosystems, KIT0204) according to manufacturer’s protocol. cDNA was prepared with 200 ng RNA and synthesized using the Primescript RT kit (Takara Inc). RT-qPCR was performed using PerfeCTa SYBR Green FastMix (Quanta bio) in the StepOnePlus^™ Real-Time PCR system (Applied Biosystems). Glyceraldehyde-3-phosphate dehydrogenase (Gapdh) was used as a control housekeeping gene. Relative expression was calculated using the 2^–deltaCT method. Primers used for RT-qPCR are listed in Supplementary Table 1.

Tartrate-resistant acid phosphatase (TRAP) staining

Sections were incubated in TRAP buffer (0.92% sodium acetate anhydrous, 1.14% L-(+) tartaric acid, 1% glacial acetic acid, pH 4.1–4.3) at room temperature for 10 minutes, then incubate with ELF97 phosphatase substrate (125 μM, Invitrogen, E6588) in TRAP buffer under UV light at room temperature for 15 minutes. Sections were washed 2 times with PBS, followed by 2 more washes with TRAP washing buffer (20 mM EDTA, 5 mM levamisole hydrochloride, pH 9). Nuclei were stained with DRAQ5 (1:1000 in PBS, Thermo Scientific, 62251) at room temperature for 5 minutes, washed with PBS, then cover slipped and mounted using Fluoromount (SouthernBiotech, OB100–01).

Ovariectomy

Twelve-week-old female C57/BL6J mice were subjected to either bilateral ovariectomy (OVX) or sham surgery. For analgesia, carprofen (10 mg/kg) was injected subcutaneously before the operation, followed by 1 dose of carprofen every day for 3 days after the operation. Mice were anesthetized and maintained with 3% inhalant isoflurane delivered with 100% O2 (1 liter/min) via face mask. Eye lubricant was used to protect the eyes during surgery. The lower back was shaved and prepared with a triple application of betadine disinfectant alternating with isopropyl alcohol. Mice were then placed on a warm circulating water pad to maintain the body temperature during the operation. A sterile drape was fitted over the entire animal and only surgical site was exposed. The surgery began with a 1 cm dorsal midline skin incision caudal to the posterior border of the ribs. Skin was separated from the muscle layer and a small incision was made on the muscle layer to enter the abdominal cavity. Using sterile forceps, the peri-ovarian fat is gently pulled out to lift and visualize the ovary. The ovary is resected, and the uterine horn returned into the abdomen. The same process is repeated to remove the other ovary. Muscle layer was closed with a continuous suture using 6–0 absorbable Vicryl (Ethicon), and the skin closed with interrupted sutures. Wound clips were applied to reinforce wound closure. Sham operations involved the same incision approach and visualization of the ovary, but without its removal. After surgery, mice were returned to their home cage and allowed unrestricted movement and access to food and water. Eight weeks post-surgery (20 weeks old), mice were euthanized by CO2 inhalation followed by cervical dislocation. Uterine tissues were weighed to confirm successful ovariectomy.

MicroCT analysis

A high-resolution desktop micro-tomographic imaging system (μCT40, Scanco Medical AG, Brüttisellen, Switzerland) was used to assess the trabecular bone microarchitecture, volume, and mineral density in the primary and secondary spongiosa of the distal femur and the L5 vertebral body of each mouse. Scans were acquired using a 10 μm3 isotropic voxel size, 70 kVp peak x-ray tube intensity, 114 mA x-ray tube current, and 200 ms integration time, and were subjected to Gaussian filtration and segmentation. The primary spongiosa region of interest started at the peak of the distal growth plate and extended proximally for 500 μm (50 transverse slices). The secondary spongiosa region of interest started immediately superior to the primary spongiosa region and extended 1500 μm (150 transverse slices) proximally. In both regions, bone was segmented from soft-tissue using a mineral density threshold of 400 mgHA/cm3. All analyses were performed using the Scanco Evaluation software. Trabecular bone in the endocortical area of the secondary spongiosa region was analyzed for bone volume fraction (Tb.BV/TV, %), trabecular thickness (Tb.Th, mm), trabecular number (Tb.N, mm-1), trabecular separation (Tb.Sp, mm), connectivity density (Conn.D, mm-3), and trabecular bone mineral density (Tb.BMD, mgHA/cm3). The bone in the primary spongiosa was analyzed for bone volume, total volume, bone volume fraction, and bone and tissue mineral densities. Cortical bone architecture and mineral density were analyzed at the femoral mid-diaphysis. Cortical bone was segmented with the mineral density of 700 mgHA/cm3 and then analyzed using the Scanco mid-shaft evaluation script to measure total cross-sectional area (bone + medullary area) (Tt.Ar, mm2), cortical bone area (Ct.Ar, mm2), medullary area (Ma.Ar, mm2), bone area fraction (Ct.Ar/Tt.Ar, %), cortical tissue mineral density (Ct.TMD, mgHA/cm3), cortical thickness (Ct.Th, mm), cortical porosity (%), as well as maximum, minimum and polar moments of inertia (Imax, Imin, and J, mm4). For L5 trabecular analysis, bone was evaluated in a cylindrical region of interest. The cylindrical region was a 2 mm (200 transverse slices) long cylinder (diameter = 750 μm, 75 voxels) that was centralized on the vertebral body. Bone was segmented from soft tissue using a mineral density threshold of 400 mgHA/cm3 and the following architectural parameters were measured using the Scanco Trabecular Bone Morphometry evaluation script: trabecular bone volume fraction (Tb.BV/TV, %), trabecular bone mineral density (Tb. BMD, mgHA/cm³), trabecular thickness (Tb.Th, mm), trabecular number (Tb.N, mm⁻¹), and trabecular separation (Tb.Sp, mm), and connectivity density (Conn.D, 1/mm³).

Serum bone turnover marker ELISAs

Twelve hour fasting serum was collected from mice just prior to sacrifice by retro-orbital bleed. Serum measurements for P1NP (IDS Immunodiagnostic Systems, AC-33F1) and CTX (IDS Immunodiagnostic Systems, AC-06F1) were performed using commercial bioassay detection kits following the manufacturer’s instructions.

Human bone marrow stromal cells

DATA-Biopsy study design and participant characteristics are described previously^59,69. Briefly, bone marrow aspirates were obtained from postmenopausal women receiving teriparatide (20 μg/day) subcutaneously for 3 months. Marrow aspirates were transferred into EDTA-coated tubes (Becton Dickinson, 366643) and placed on a rocker to prevent clotting. Bone marrow stromal cells were extracted by Ficoll-Paque gradient separation and prepared for FACS as described above.

Statistical analysis

Statistical analyses were performed by GraphPad Prism 10 (GraphPad Software Inc, USA). For comparison between two experimental groups, Student’s two-tailed t tests was used. For more than two groups, one-way or two-way ANOVA followed by Šidák’s multiple comparisons test were used. Data is expressed as mean ± SD. Statistical significance was assigned at P < 0.05 with P values indicated in the figures. The number of mice per experiment is indicated in the figure legend.

Supplementary Material

Supplement 1

NIHPP2026.05.21.726951v1-supplement-1.pdf^{(1.7MB, pdf)}

Acknowledgements

We thank Drs. Roland Baron, Francesca Gori, Yi-Hsiang Hsu, Douglas Kiel, Matthew Greenblatt, Christa Maes, Sundeep Khosla, Natalie Sims, Dolores Shoback, and all members of the Wein laboratory and MGH Endocrine Unit for valuable feedback. We thank Dr. Marie Demay for providing guidance and methods for isolating human bone marrow cells. BSKC is supported by Canadian Institutes of Health Research Postdoctoral Fellowship (MFE-201024). MNW acknowledges support from NIAMS (P50AR080596) and NIDDK (R01DK116716) and a Chen Institute MGH Research Scholar award. GP acknowledges support from NIAMS Clinical Investigator Award (K08AR084618). Bone MicroCT work was supported by the NIAMS-funded Center for Musculoskeletal Research (P30AR075042). Flow cytometry work was supported with expert technical assistance by the HSCI-CRM Flow Cytometry facility at Massachusetts General Hospital. Confocal imaging was performed in the Microscopy Core of the Program in Membrane Biology, which is partially supported by a Centre for the Study of Inflammatory Bowel Disease Grant DK043351 and a Boston Area Diabetes and Endocrinology Research Center (BADERC) Award DK135043. The Zeiss confocal system is supported by grant S10OD021577-01.

Footnotes

Declaration of interests

MNW receives research funding from Biomarin, Angitia Biopharmaceuticals, and Bayer. MNW received research funding from Radius Health and is a coinventor on a pending patent (US patent application 16/333,546) regarding the use of SIK inhibitors for osteoporosis. MNW holds equity in and is a scientific advisory board member for Relation Therapeutics.

References

1.Morrison S.J., and Spradling A.C. (2008). Stem cells and niches: mechanisms that promote stem cell maintenance throughout life. Cell 132, 598–611. 10.1016/j.cell.2008.01.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Schofield R. (1978). The relationship between the spleen colony-forming cell and the haemopoietic stem cell. Blood Cells 4, 7–25. [PubMed] [Google Scholar]
3.Méndez-Ferrer S., Michurina T.V., Ferraro F., Mazloom A.R., Macarthur B.D., Lira S.A., Scadden D.T., Ma’ayan A., Enikolopov G.N., and Frenette P.S. (2010). Mesenchymal and haematopoietic stem cells form a unique bone marrow niche. Nature 466, 829–834. 10.1038/nature09262. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Park D., Spencer J.A., Koh B.I., Kobayashi T., Fujisaki J., Clemens T.L., Lin C.P., Kronenberg H.M., and Scadden D.T. (2012). Endogenous bone marrow MSCs are dynamic, fate-restricted participants in bone maintenance and regeneration. Cell Stem Cell 10, 259–272. 10.1016/j.stem.2012.02.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Zhou B.O., Yue R., Murphy M.M., Peyer J.G., and Morrison S.J. (2014). Leptin-Receptor-Expressing Mesenchymal Stromal Cells Represent the Main Source of Bone Formed by Adult Bone Marrow. Cell Stem Cell 15, 154–168. 10.1016/j.stem.2014.06.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Ono N., Ono W., Nagasawa T., and Kronenberg H.M. (2014). A subset of chondrogenic cells provides early mesenchymal progenitors in growing bones. Nat Cell Biol 16, 1157–1167. 10.1038/ncb3067. [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Worthley D.L., Churchill M., Compton J.T., Tailor Y., Rao M., Si Y., Levin D., Schwartz M.G., Uygur A., Hayakawa Y., et al. (2015). Gremlin 1 identifies a skeletal stem cell with bone, cartilage, and reticular stromal potential. Cell 160, 269–284. 10.1016/j.cell.2014.11.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Fan Y., Hanai J., Le P.T., Bi R., Maridas D., DeMambro V., Figueroa C.A., Kir S., Zhou X., Mannstadt M., et al. (2017). Parathyroid Hormone Directs Bone Marrow Mesenchymal Cell Fate. Cell Metab 25, 661–672. 10.1016/j.cmet.2017.01.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Omatsu Y., Sugiyama T., Kohara H., Kondoh G., Fujii N., Kohno K., and Nagasawa T. (2010). The Essential Functions of Adipo-osteogenic Progenitors as the Hematopoietic Stem and Progenitor Cell Niche. Immunity 33, 387–399. 10.1016/j.immuni.2010.08.017. [DOI] [PubMed] [Google Scholar]
10.Omatsu Y., Seike M., Sugiyama T., Kume T., and Nagasawa T. (2014). Foxc1 is a critical regulator of haematopoietic stem/progenitor cell niche formation. Nature 508, 536–540. 10.1038/nature13071. [DOI] [PubMed] [Google Scholar]
11.Seike M., Omatsu Y., Watanabe H., Kondoh G., and Nagasawa T. (2018). Stem cell niche-specific Ebf3 maintains the bone marrow cavity. Genes Dev 32, 359–372. 10.1101/gad.311068.117. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Shu H.S., Liu Y.L., Tang X.T., Zhang X.S., Zhou B., Zou W., and Zhou B.O. (2021). Tracing the skeletal progenitor transition during postnatal bone formation. Cell Stem Cell 28, 2122–2136.e3. 10.1016/j.stem.2021.08.010. [DOI] [PubMed] [Google Scholar]
13.Yang L., Tsang K.Y., Tang H.C., Chan D., and Cheah K.S.E. (2014). Hypertrophic chondrocytes can become osteoblasts and osteocytes in endochondral bone formation. Proc Natl Acad Sci U S A 111, 12097–12102. 10.1073/pnas.1302703111. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Balani D.H., Ono N., and Kronenberg H.M. (2017). Parathyroid hormone regulates fates of murine osteoblast precursors in vivo. J Clin Invest 127, 3327–3338. 10.1172/JCI91699. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Yang M., Arai A., Udagawa N., Hiraga T., Lijuan Z., Ito S., Komori T., Moriishi T., Matsuo K., Shimoda K., et al. (2017). Osteogenic Factor Runx2 Marks a Subset of Leptin Receptor-Positive Cells that Sit Atop the Bone Marrow Stromal Cell Hierarchy. Sci Rep 7, 4928. 10.1038/s41598-017-05401-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Yang M., Arai A., Udagawa N., Zhao L., Nishida D., Murakami K., Hiraga T., Takao-Kawabata R., Matsuo K., Komori T., et al. Parathyroid Hormone Shifts Cell Fate of a Leptin Receptor‐Marked Stromal Population from Adipogenic to Osteoblastic Lineage. 10.1002/jbmr.3811. [DOI] [PubMed] [Google Scholar]
17.Caire R., Roche B., Picot T., Aanei C.-M., He Z., Campos L., Thomas M., Malaval L., Vico L., and Lafage-Proust M.-H. (2019). Parathyroid Hormone Remodels Bone Transitional Vessels and the Leptin Receptor-Positive Pericyte Network in Mice. J Bone Miner Res 34, 1487–1501. 10.1002/jbmr.3728. [DOI] [PubMed] [Google Scholar]
18.Zhang J., Cohen A., Shen B., Du L., Tasdogan A., Zhao Z., Shane E.J., and Morrison S.J. (2021). The effect of parathyroid hormone on osteogenesis is mediated partly by osteolectin. Proceedings of the National Academy of Sciences 118, e2026176118. 10.1073/pnas.2026176118. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Nishida S., Yamaguchi A., Tanizawa T., Endo N., Mashiba T., Uchiyama Y., Suda T., Yoshiki S., and Takahashi H.E. (1994). Increased bone formation by intermittent parathyroid hormone administration is due to the stimulation of proliferation and differentiation of osteoprogenitor cells in bone marrow. Bone 15, 717–723. 10.1016/8756-3282(94)90322-0. [DOI] [PubMed] [Google Scholar]
20.Xian L., Wu X., Pang L., Lou M., Rosen C.J., Qiu T., Crane J., Frassica F., Zhang L., Rodriguez J.P., et al. (2012). Matrix IGF-1 maintains bone mass by activation of mTOR in mesenchymal stem cells. Nat Med 18, 1095–1101. 10.1038/nm.2793. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Tang Y., Wu X., Lei W., Pang L., Wan C., Shi Z., Zhao L., Nagy T.R., Peng X., Hu J., et al. (2009). TGF-beta1-induced migration of bone mesenchymal stem cells couples bone resorption with formation. Nat Med 15, 757–765. 10.1038/nm.1979. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Charles J.F., and Aliprantis A.O. (2014). Osteoclasts: more than ‘bone eaters.’ Trends Mol Med 20, 449–459. 10.1016/j.molmed.2014.06.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Keller H., and Kneissel M. (2005). SOST is a target gene for PTH in bone. Bone 37, 148–158. 10.1016/j.bone.2005.03.018. [DOI] [PubMed] [Google Scholar]
24.Yu W., Zhong L., Yao L., Wei Y., Gui T., Li Z., Kim H., Holdreith N., Jiang X., Tong W., et al. Bone marrow adipogenic lineage precursors promote osteoclastogenesis in bone remodeling and pathologic bone loss. J Clin Invest 131, e140214. 10.1172/JCI140214. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Chan C.K.F., Seo E.Y., Chen J.Y., Lo D., McArdle A., Sinha R., Tevlin R., Seita J., Vincent-Tompkins J., Wearda T., et al. (2015). Identification and Specification of the Mouse Skeletal Stem Cell. Cell 160, 285–298. 10.1016/j.cell.2014.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Jilka R.L., Weinstein R.S., Bellido T., Roberson P., Parfitt A.M., and Manolagas S.C. (1999). Increased bone formation by prevention of osteoblast apoptosis with parathyroid hormone. J Clin Invest 104, 439–446. 10.1172/JCI6610. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Korsunsky I., Millard N., Fan J., Slowikowski K., Zhang F., Wei K., Baglaenko Y., Brenner M., Loh P., and Raychaudhuri S. (2019). Fast, sensitive and accurate integration of single-cell data with Harmony. Nat Methods 16, 1289–1296. 10.1038/s41592-019-0619-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Nookaew I., Xiong J., Onal M., Bustamante-Gomez C., Wanchai V., Fu Q., Kim H.-N., Almeida M., and O’Brien C.A. (2024). Refining the identity of mesenchymal cell types associated with murine periosteal and endosteal bone. J Biol Chem 300, 107158. 10.1016/j.jbc.2024.107158. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Baccin C., Al-Sabah J., Velten L., Helbling P.M., Grünschläger F., Hernández-Malmierca P., Nombela-Arrieta C., Steinmetz L.M., Trumpp A., and Haas S. (2020). Combined single-cell and spatial transcriptomics reveal the molecular, cellular and spatial bone marrow niche organization. Nat Cell Biol 22, 38–48. 10.1038/s41556-019-0439-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Translocation of proteins across the endoplasmic reticulum. I. Signal recognition protein (SRP) binds to in-vitro-assembled polysomes synthesizing secretory protein (1981). J Cell Biol 91, 545–550. 10.1083/jcb.91.2.545. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Kuge O., Dascher C., Orci L., Rowe T., Amherdt M., Plutner H., Ravazzola M., Tanigawa G., Rothman J.E., and Balch W.E. (1994). Sar1 promotes vesicle budding from the endoplasmic reticulum but not Golgi compartments. J Cell Biol 125, 51–65. 10.1083/jcb.125.1.51. [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Zhong L., Yao L., Tower R.J., Wei Y., Miao Z., Park J., Shrestha R., Wang L., Yu W., Holdreith N., et al. (2020). Single cell transcriptomics identifies a unique adipose lineage cell population that regulates bone marrow environment. eLife 9, e54695. 10.7554/eLife.54695. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Laurin M., and Côté J.-F. (2014). Insights into the biological functions of Dock family guanine nucleotide exchange factors. Genes Dev 28, 533–547. 10.1101/gad.236349.113. [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Bellanger J.M., Astier C., Sardet C., Ohta Y., Stossel T.P., and Debant A. (2000). The Rac1- and RhoG-specific GEF domain of Trio targets filamin to remodel cytoskeletal actin. Nat Cell Biol 2, 888–892. 10.1038/35046533. [DOI] [PubMed] [Google Scholar]
35.Héraud C., Pinault M., Lagrée V., and Moreau V. (2019). p190RhoGAPs, the ARHGAP35- and ARHGAP5-Encoded Proteins, in Health and Disease. Cells 8, 351. 10.3390/cells8040351. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Ding X., Wu J., and Jiang C. (2010). ABCG2: a potential marker of stem cells and novel target in stem cell and cancer therapy. Life Sci 86, 631–637. 10.1016/j.lfs.2010.02.012. [DOI] [PubMed] [Google Scholar]
37.Wei Q., Nakahara F., Asada N., Zhang D., Gao X., Xu C., Alfieri A., Brodin N.P., Zimmerman S.E., Mar J.C., et al. (2020). Snai2 Maintains Bone Marrow Niche Cells by Repressing Osteopontin Expression. Developmental Cell 53, 503–513.e5. 10.1016/j.devcel.2020.04.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Kalucka J., de Rooij L.P.M.H., Goveia J., Rohlenova K., Dumas S.J., Meta E., Conchinha N.V., Taverna F., Teuwen L.-A., Veys K., et al. (2020). Single-Cell Transcriptome Atlas of Murine Endothelial Cells. Cell 180, 764–779.e20. 10.1016/j.cell.2020.01.015. [DOI] [PubMed] [Google Scholar]
39.Williams M., Lyu M.S., Yang Y.L., Lin E.P., Dunbrack R., Birren B., Cunningham J., and Hunter K. (1999). Ier5, a novel member of the slow-kinetics immediate-early genes. Genomics 55, 327–334. 10.1006/geno.1998.5679. [DOI] [PubMed] [Google Scholar]
40.Yu H., Lin L., Zhang Z., Zhang H., and Hu H. (2020). Targeting NF-κB pathway for the therapy of diseases: mechanism and clinical study. Sig Transduct Target Ther 5, 209. 10.1038/s41392-020-00312-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Tikhonova A.N., Dolgalev I., Hu H., Sivaraj K.K., Hoxha E., Cuesta-Domínguez Á., Pinho S., Akhmetzyanova I., Gao J., Witkowski M., et al. (2019). The bone marrow microenvironment at single-cell resolution. Nature 569, 222–228. 10.1038/s41586-019-1104-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Lange M., Bergen V., Klein M., Setty M., Reuter B., Bakhti M., Lickert H., Ansari M., Schniering J., Schiller H.B., et al. (2022). CellRank for directed single-cell fate mapping. Nat Methods 19, 159–170. 10.1038/s41592-021-01346-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Gupta R.K., Mepani R.J., Kleiner S., Lo J.C., Khandekar M.J., Cohen P., Frontini A., Bhowmick D.C., Ye L., Cinti S., et al. (2012). Zfp423 Expression Identifies Committed Preadipocytes and Localizes to Adipose Endothelial and Perivascular Cells. Cell Metabolism 15, 230–239. 10.1016/j.cmet.2012.01.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Shao M., Ishibashi J., Kusminski C.M., Wang Q.A., Hepler C., Vishvanath L., MacPherson K.A., Spurgin S.B., Sun K., Holland W.L., et al. (2016). Zfp423 Maintains White Adipocyte Identity through Suppression of the Beige Cell Thermogenic Gene Program. Cell Metab 23, 1167–1184. 10.1016/j.cmet.2016.04.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Hong S.H.H., Lu X., Nanes M.S., and Mitchell J. (2009). Regulation of osterix (Osx, Sp7) and the Osx promoter by parathyroid hormone in osteoblasts. Journal of Molecular Endocrinology 43, 197–207. 10.1677/JME-09-0012. [DOI] [PubMed] [Google Scholar]
46.Zhou B.O., Ding L., and Morrison S.J. Hematopoietic stem and progenitor cells regulate the regeneration of their niche by secreting Angiopoietin-1. eLife 4, e05521. 10.7554/eLife.05521. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Pinho S., Wei Q., Maryanovich M., Zhang D., Balandrán J.C., Pierce H., Nakahara F., Di Staulo A., Bartholdy B.A., Xu J., et al. (2022). VCAM1 confers innate immune tolerance on haematopoietic and leukaemic stem cells. Nat Cell Biol 24, 290–298. 10.1038/s41556-022-00849-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Prisby R., Guignandon A., Vanden-Bossche A., Mac-Way F., Linossier M.-T., Thomas M., Laroche N., Malaval L., Langer M., Peter Z.-A., et al. (2011). Intermittent PTH(1–84) is osteoanabolic but not osteoangiogenic and relocates bone marrow blood vessels closer to bone-forming sites. J Bone Miner Res 26, 2583–2596. 10.1002/jbmr.459. [DOI] [PubMed] [Google Scholar]
49.Mohanakrishnan V., Sivaraj K.K., Jeong H.-W., Bovay E., Dharmalingam B., Bixel M.G., Dinh V.V., Petkova M., Paredes Ugarte I., Kuo Y.-T., et al. (2024). Specialized post-arterial capillaries facilitate adult bone remodelling. Nat Cell Biol 26, 2020–2034. 10.1038/s41556-024-01545-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Smith N.P., Janton C., Alcantara A.C.M., George M., Mankongtreecheep K., Mao G., Villani A.-C., and Kronenberg H.M. (2025). Single cell RNA sequencing shows that cells expressing Sox9 postnatally populate most skeletal lineages in mouse bone. J Bone Miner Res 40, 799–812. 10.1093/jbmr/zjaf043. [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Nakashima K., Zhou X., Kunkel G., Zhang Z., Deng J.M., Behringer R.R., and de Crombrugghe B. (2002). The novel zinc finger-containing transcription factor osterix is required for osteoblast differentiation and bone formation. Cell 108, 17–29. 10.1016/s0092-8674(01)00622-5. [DOI] [PubMed] [Google Scholar]
52.Wein M.N., and Kronenberg H.M. (2018). Regulation of Bone Remodeling by Parathyroid Hormone. Cold Spring Harb Perspect Med 8, a031237. 10.1101/cshperspect.a031237. [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Katsuno Y., and Derynck R. (2021). Epithelial plasticity, epithelial-mesenchymal transition, and the TGF-β family. Developmental Cell 56, 726–746. 10.1016/j.devcel.2021.02.028. [DOI] [PubMed] [Google Scholar]
54.Fournier P.G.J., Juárez P., Jiang G., Clines G.A., Niewolna M., Kim H.S., Walton H.W., Peng X.H., Liu Y., Mohammad K.S., et al. (2015). The TGF-β Signaling Regulator PMEPA1 Suppresses Prostate Cancer Metastases to Bone. Cancer Cell 27, 809–821. 10.1016/j.ccell.2015.04.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Xu F., Zhang J., Hu G., Liu L., and Liang W. (2017). Hypoxia and TGF-β1 induced PLOD2 expression improve the migration and invasion of cervical cancer cells by promoting epithelial-to-mesenchymal transition (EMT) and focal adhesion formation. Cancer Cell Int 17, 54. 10.1186/s12935-017-0420-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Gjaltema R.A.F., de Rond S., Rots M.G., and Bank R.A. (2015). Procollagen Lysyl Hydroxylase 2 Expression Is Regulated by an Alternative Downstream Transforming Growth Factor β−1 Activation Mechanism. J Biol Chem 290, 28465–28476. 10.1074/jbc.M114.634311. [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Ranganathan P., Agrawal A., Bhushan R., Chavalmane A.K., Kalathur R.K.R., Takahashi T., and Kondaiah P. (2007). Expression profiling of genes regulated by TGF-beta: Differential regulation in normal and tumour cells. BMC Genomics 8, 98. 10.1186/1471-2164-8-98. [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Inhibition of Sca-1-positive skeletal stem cell recruitment by alendronate blunts the anabolic effects of parathyroid hormone on bone remodeling - PubMed https://pubmed.ncbi.nlm.nih.gov/21040899/. [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Kobelski M.M., Ramchand S.K., Tsai J.N., Leder B.Z., and Demay M.B. (2024). Evaluation of Osteogenic Phenotype in Postmenopausal Women Receiving Anabolic and Antiresorptive Osteoporosis Therapies. Endocrinology 165, bqae142. 10.1210/endocr/bqae142. [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Chu N.T.L., Dregval O., Chang Y.-W., Kriukov E., Tian X., Liu X., Trompet D., Zhang M.S., Li L., Li Z., et al. (2025). Quantitative Multicolored Deep Imaging of Human Bones Reveals a Composite Osteo-Sinusoidal Niche for Mesenchymal Stromal Cells. Preprint at bioRxiv, 10.1101/2025.10.07.680053 https://doi.org/10.1101/2025.10.07.680053. [DOI] [Google Scholar]
61.Kimura T., Panaroni C., Rankin E., Purton L.E., and Wu J.Y. (2022). Loss of parathyroid hormone receptor signaling in osteoprogenitors is associated with accumulation of multiple hematopoietic lineages in the bone marrow. J Bone Miner Res 37, 1321–1334. 10.1002/jbmr.4568. [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Greenbaum A., Hsu Y.-M.S., Day R.B., Schuettpelz L.G., Christopher M.J., Borgerding J.N., Nagasawa T., and Link D.C. (2013). CXCL12 in early mesenchymal progenitors is required for haematopoietic stem-cell maintenance. Nature 495, 227–230. 10.1038/nature11926. [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Kaneko T., Yari S., Kikuta J., Omatsu Y., Seno S., Kikuchi S., Sato K., Fujii K., Sudo T., Hasegawa T., et al. (2025). The RANK/RANKL axis controls vascular dynamics in the bone marrow. Proceedings of the National Academy of Sciences 122, e2425366122. 10.1073/pnas.2425366122. [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Lu J., He Q., Wang H., Yao L., Duffy M., Guo H., Braun C., Zhou Y., Liang Q., Lin Y., et al. (2025). Bone marrow adipogenic lineage precursors are the major regulator of bone resorption in adult mice. Bone Res 13, 39. 10.1038/s41413-025-00405-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
65.Linkhart T.A., Mohan S., and Baylink D.J. (1996). Growth factors for bone growth and repair: IGF, TGFβ and BMP. Bone 19, S1–S12. 10.1016/S8756-3282(96)00138-X. [DOI] [PubMed] [Google Scholar]
66.Wu X., Pang L., Lei W., Lu W., Li J., Li Z., Frassica F.J., Chen X., Wan M., and Cao X. (2010). Inhibition of Sca-1-positive skeletal stem cell recruitment by alendronate blunts the anabolic effects of parathyroid hormone on bone remodeling. Cell Stem Cell 7, 571–580. 10.1016/j.stem.2010.09.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Crane J.L., and Cao X. (2014). Bone marrow mesenchymal stem cells and TGF-β signaling in bone remodeling. J Clin Invest 124, 466–472. 10.1172/JCI70050. [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Kim S.W., Pajevic P.D., Selig M., Barry K.J., Yang J.-Y., Shin C.S., Baek W.-Y., Kim J.-E., and Kronenberg H.M. Intermittent parathyroid hormone administration converts quiescent lining cells to active osteoblasts. 10.1002/jbmr.1665. [DOI] [PMC free article] [PubMed] [Google Scholar]
69.Leder B.Z., Tsai J.N., Uihlein A.V., Burnett-Bowie S.-A.M., Zhu Y., Foley K., Lee H., and Neer R.M. (2013). Two Years of Denosumab and Teriparatide Administration in Postmenopausal Women With Osteoporosis (The DATA Extension Study): A Randomized Controlled Trial. Lancet 382, 50–56. 10.1016/S0140-6736(13)60856-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Omatsu Y., Aiba S., Maeta T., Higaki K., Aoki K., Watanabe H., Kondoh G., Nishimura R., Takeda S., Chung U., et al. (2022). Runx1 and Runx2 inhibit fibrotic conversion of cellular niches for hematopoietic stem cells. Nat Commun 13, 2654. 10.1038/s41467-022-30266-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
71.Wein M.N., Liang Y., Goransson O., Sundberg T.B., Wang J., Williams E.A., O’Meara M.J., Govea N., Beqo B., Nishimori S., et al. (2016). SIKs control osteocyte responses to parathyroid hormone. Nat Commun 7, 13176. 10.1038/ncomms13176. [DOI] [PMC free article] [PubMed] [Google Scholar]
72.Lindsay R., Nieves J., Formica C., Henneman E., Woelfert L., Shen V., Dempster D., and Cosman F. (1997). Randomised controlled study of effect of parathyroid hormone on vertebral-bone mass and fracture incidence among postmenopausal women on oestrogen with osteoporosis. Lancet 350, 550–555. 10.1016/S0140-6736(97)02342-8. [DOI] [PubMed] [Google Scholar]
73.Gatti D., Viapiana O., Idolazzi L., Fracassi E., Rossini M., and Adami S. (2011). The Waning of Teriparatide Effect on Bone Formation Markers in Postmenopausal Osteoporosis Is Associated with Increasing Serum Levels of DKK1. J Clin Endocrinol Metab 96, 1555– 1559. 10.1210/jc.2010-2552. [DOI] [PubMed] [Google Scholar]
74.Finak G., McDavid A., Yajima M., Deng J., Gersuk V., Shalek A.K., Slichter C.K., Miller H.W., McElrath M.J., Prlic M., et al. (2015). MAST: a flexible statistical framework for assessing transcriptional changes and characterizing heterogeneity in single-cell RNA sequencing data. Genome Biol 16, 278. 10.1186/s13059-015-0844-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
75.Soeda T., Deng J.M., de Crombrugghe B., Behringer R.R., Nakamura T., and Akiyama H. (2010). Sox9-expressing precursors are the cellular origin of the cruciate ligament of the knee joint and the limb tendons. Genesis 48, 635–644. 10.1002/dvg.20667. [DOI] [PMC free article] [PubMed] [Google Scholar]
76.Kalajzic Z., Liu P., Kalajzic I., Du Z., Braut A., Mina M., Canalis E., and Rowe D.W. (2002). Directing the expression of a green fluorescent protein transgene in differentiated osteoblasts: comparison between rat type I collagen and rat osteocalcin promoters. Bone 31, 654–660. 10.1016/s8756-3282(02)00912-2. [DOI] [PubMed] [Google Scholar]
77.Ara T., Tokoyoda K., Sugiyama T., Egawa T., Kawabata K., and Nagasawa T. (2003). Long-term hematopoietic stem cells require stromal cell-derived factor-1 for colonizing bone marrow during ontogeny. Immunity 19, 257–267. 10.1016/s1074-7613(03)00201-2. [DOI] [PubMed] [Google Scholar]
78.Kobayashi T., Chung U., Schipani E., Starbuck M., Karsenty G., Katagiri T., Goad D.L., Lanske B., and Kronenberg H.M. (2002). PTHrP and Indian hedgehog control differentiation of growth plate chondrocytes at multiple steps. Development 129, 2977–2986. 10.1242/dev.129.12.2977. [DOI] [PubMed] [Google Scholar]
79.Akiyama H., Kim J.-E., Nakashima K., Balmes G., Iwai N., Deng J.M., Zhang Z., Martin J.F., Behringer R.R., Nakamura T., et al. (2005). Osteo-chondroprogenitor cells are derived from Sox9 expressing precursors. Proc Natl Acad Sci U S A 102, 14665–14670. 10.1073/pnas.0504750102. [DOI] [PMC free article] [PubMed] [Google Scholar]
80.Yoon S.-H., Meyer M.B., Arevalo C., Tekguc M., Zhang C., Wang J.S., Castro Andrade C.D., Strauss K., Sato T., Benkusky N.A., et al. (2023). A parathyroid hormone/salt-inducible kinase signaling axis controls renal vitamin D activation and organismal calcium homeostasis. J Clin Invest 133, e163627. 10.1172/JCI163627. [DOI] [PMC free article] [PubMed] [Google Scholar]
81.Hao Y., Stuart T., Kowalski M.H., Choudhary S., Hoffman P., Hartman A., Srivastava A., Molla G., Madad S., Fernandez-Granda C., et al. (2024). Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nat Biotechnol 42, 293–304. 10.1038/s41587-023-01767-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
82.Wolf F.A., Angerer P., and Theis F.J. (2018). SCANPY: large-scale single-cell gene expression data analysis. Genome Biol 19, 15. 10.1186/s13059-017-1382-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
83.Bergen V., Lange M., Peidli S., Wolf F.A., and Theis F.J. (2020). Generalizing RNA velocity to transient cell states through dynamical modeling. Nat Biotechnol 38, 1408–1414. 10.1038/s41587-020-0591-3. [DOI] [PubMed] [Google Scholar]
84.Zappia L., and Oshlack A. (2018). Clustering trees: a visualization for evaluating clusterings at multiple resolutions. Gigascience 7, giy083. 10.1093/gigascience/giy083. [DOI] [PMC free article] [PubMed] [Google Scholar]
85.Xie Z., Bailey A., Kuleshov M.V., Clarke D.J.B., Evangelista J.E., Jenkins S.L., Lachmann A., Wojciechowicz M.L., Kropiwnicki E., Jagodnik K.M., et al. (2021). Gene Set Knowledge Discovery with Enrichr. Curr Protoc 1, e90. 10.1002/cpz1.90. [DOI] [PMC free article] [PubMed] [Google Scholar]
86.La Manno G., Soldatov R., Zeisel A., Braun E., Hochgerner H., Petukhov V., Lidschreiber K., Kastriti M.E., Lönnerberg P., Furlan A., et al. (2018). RNA velocity of single cells. Nature 560, 494–498. 10.1038/s41586-018-0414-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1

NIHPP2026.05.21.726951v1-supplement-1.pdf^{(1.7MB, pdf)}

[R1] 1.Morrison S.J., and Spradling A.C. (2008). Stem cells and niches: mechanisms that promote stem cell maintenance throughout life. Cell 132, 598–611. 10.1016/j.cell.2008.01.038. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R2] 2.Schofield R. (1978). The relationship between the spleen colony-forming cell and the haemopoietic stem cell. Blood Cells 4, 7–25. [PubMed] [Google Scholar]

[R3] 3.Méndez-Ferrer S., Michurina T.V., Ferraro F., Mazloom A.R., Macarthur B.D., Lira S.A., Scadden D.T., Ma’ayan A., Enikolopov G.N., and Frenette P.S. (2010). Mesenchymal and haematopoietic stem cells form a unique bone marrow niche. Nature 466, 829–834. 10.1038/nature09262. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R4] 4.Park D., Spencer J.A., Koh B.I., Kobayashi T., Fujisaki J., Clemens T.L., Lin C.P., Kronenberg H.M., and Scadden D.T. (2012). Endogenous bone marrow MSCs are dynamic, fate-restricted participants in bone maintenance and regeneration. Cell Stem Cell 10, 259–272. 10.1016/j.stem.2012.02.003. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R5] 5.Zhou B.O., Yue R., Murphy M.M., Peyer J.G., and Morrison S.J. (2014). Leptin-Receptor-Expressing Mesenchymal Stromal Cells Represent the Main Source of Bone Formed by Adult Bone Marrow. Cell Stem Cell 15, 154–168. 10.1016/j.stem.2014.06.008. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R6] 6.Ono N., Ono W., Nagasawa T., and Kronenberg H.M. (2014). A subset of chondrogenic cells provides early mesenchymal progenitors in growing bones. Nat Cell Biol 16, 1157–1167. 10.1038/ncb3067. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R7] 7.Worthley D.L., Churchill M., Compton J.T., Tailor Y., Rao M., Si Y., Levin D., Schwartz M.G., Uygur A., Hayakawa Y., et al. (2015). Gremlin 1 identifies a skeletal stem cell with bone, cartilage, and reticular stromal potential. Cell 160, 269–284. 10.1016/j.cell.2014.11.042. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R8] 8.Fan Y., Hanai J., Le P.T., Bi R., Maridas D., DeMambro V., Figueroa C.A., Kir S., Zhou X., Mannstadt M., et al. (2017). Parathyroid Hormone Directs Bone Marrow Mesenchymal Cell Fate. Cell Metab 25, 661–672. 10.1016/j.cmet.2017.01.001. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R9] 9.Omatsu Y., Sugiyama T., Kohara H., Kondoh G., Fujii N., Kohno K., and Nagasawa T. (2010). The Essential Functions of Adipo-osteogenic Progenitors as the Hematopoietic Stem and Progenitor Cell Niche. Immunity 33, 387–399. 10.1016/j.immuni.2010.08.017. [DOI] [PubMed] [Google Scholar]

[R10] 10.Omatsu Y., Seike M., Sugiyama T., Kume T., and Nagasawa T. (2014). Foxc1 is a critical regulator of haematopoietic stem/progenitor cell niche formation. Nature 508, 536–540. 10.1038/nature13071. [DOI] [PubMed] [Google Scholar]

[R11] 11.Seike M., Omatsu Y., Watanabe H., Kondoh G., and Nagasawa T. (2018). Stem cell niche-specific Ebf3 maintains the bone marrow cavity. Genes Dev 32, 359–372. 10.1101/gad.311068.117. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R12] 12.Shu H.S., Liu Y.L., Tang X.T., Zhang X.S., Zhou B., Zou W., and Zhou B.O. (2021). Tracing the skeletal progenitor transition during postnatal bone formation. Cell Stem Cell 28, 2122–2136.e3. 10.1016/j.stem.2021.08.010. [DOI] [PubMed] [Google Scholar]

[R13] 13.Yang L., Tsang K.Y., Tang H.C., Chan D., and Cheah K.S.E. (2014). Hypertrophic chondrocytes can become osteoblasts and osteocytes in endochondral bone formation. Proc Natl Acad Sci U S A 111, 12097–12102. 10.1073/pnas.1302703111. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R14] 14.Balani D.H., Ono N., and Kronenberg H.M. (2017). Parathyroid hormone regulates fates of murine osteoblast precursors in vivo. J Clin Invest 127, 3327–3338. 10.1172/JCI91699. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R15] 15.Yang M., Arai A., Udagawa N., Hiraga T., Lijuan Z., Ito S., Komori T., Moriishi T., Matsuo K., Shimoda K., et al. (2017). Osteogenic Factor Runx2 Marks a Subset of Leptin Receptor-Positive Cells that Sit Atop the Bone Marrow Stromal Cell Hierarchy. Sci Rep 7, 4928. 10.1038/s41598-017-05401-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R16] 16.Yang M., Arai A., Udagawa N., Zhao L., Nishida D., Murakami K., Hiraga T., Takao-Kawabata R., Matsuo K., Komori T., et al. Parathyroid Hormone Shifts Cell Fate of a Leptin Receptor‐Marked Stromal Population from Adipogenic to Osteoblastic Lineage. 10.1002/jbmr.3811. [DOI] [PubMed] [Google Scholar]

[R17] 17.Caire R., Roche B., Picot T., Aanei C.-M., He Z., Campos L., Thomas M., Malaval L., Vico L., and Lafage-Proust M.-H. (2019). Parathyroid Hormone Remodels Bone Transitional Vessels and the Leptin Receptor-Positive Pericyte Network in Mice. J Bone Miner Res 34, 1487–1501. 10.1002/jbmr.3728. [DOI] [PubMed] [Google Scholar]

[R18] 18.Zhang J., Cohen A., Shen B., Du L., Tasdogan A., Zhao Z., Shane E.J., and Morrison S.J. (2021). The effect of parathyroid hormone on osteogenesis is mediated partly by osteolectin. Proceedings of the National Academy of Sciences 118, e2026176118. 10.1073/pnas.2026176118. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R19] 19.Nishida S., Yamaguchi A., Tanizawa T., Endo N., Mashiba T., Uchiyama Y., Suda T., Yoshiki S., and Takahashi H.E. (1994). Increased bone formation by intermittent parathyroid hormone administration is due to the stimulation of proliferation and differentiation of osteoprogenitor cells in bone marrow. Bone 15, 717–723. 10.1016/8756-3282(94)90322-0. [DOI] [PubMed] [Google Scholar]

[R20] 20.Xian L., Wu X., Pang L., Lou M., Rosen C.J., Qiu T., Crane J., Frassica F., Zhang L., Rodriguez J.P., et al. (2012). Matrix IGF-1 maintains bone mass by activation of mTOR in mesenchymal stem cells. Nat Med 18, 1095–1101. 10.1038/nm.2793. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R21] 21.Tang Y., Wu X., Lei W., Pang L., Wan C., Shi Z., Zhao L., Nagy T.R., Peng X., Hu J., et al. (2009). TGF-beta1-induced migration of bone mesenchymal stem cells couples bone resorption with formation. Nat Med 15, 757–765. 10.1038/nm.1979. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R22] 22.Charles J.F., and Aliprantis A.O. (2014). Osteoclasts: more than ‘bone eaters.’ Trends Mol Med 20, 449–459. 10.1016/j.molmed.2014.06.001. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R23] 23.Keller H., and Kneissel M. (2005). SOST is a target gene for PTH in bone. Bone 37, 148–158. 10.1016/j.bone.2005.03.018. [DOI] [PubMed] [Google Scholar]

[R24] 24.Yu W., Zhong L., Yao L., Wei Y., Gui T., Li Z., Kim H., Holdreith N., Jiang X., Tong W., et al. Bone marrow adipogenic lineage precursors promote osteoclastogenesis in bone remodeling and pathologic bone loss. J Clin Invest 131, e140214. 10.1172/JCI140214. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R25] 25.Chan C.K.F., Seo E.Y., Chen J.Y., Lo D., McArdle A., Sinha R., Tevlin R., Seita J., Vincent-Tompkins J., Wearda T., et al. (2015). Identification and Specification of the Mouse Skeletal Stem Cell. Cell 160, 285–298. 10.1016/j.cell.2014.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R26] 26.Jilka R.L., Weinstein R.S., Bellido T., Roberson P., Parfitt A.M., and Manolagas S.C. (1999). Increased bone formation by prevention of osteoblast apoptosis with parathyroid hormone. J Clin Invest 104, 439–446. 10.1172/JCI6610. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R27] 27.Korsunsky I., Millard N., Fan J., Slowikowski K., Zhang F., Wei K., Baglaenko Y., Brenner M., Loh P., and Raychaudhuri S. (2019). Fast, sensitive and accurate integration of single-cell data with Harmony. Nat Methods 16, 1289–1296. 10.1038/s41592-019-0619-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R28] 28.Nookaew I., Xiong J., Onal M., Bustamante-Gomez C., Wanchai V., Fu Q., Kim H.-N., Almeida M., and O’Brien C.A. (2024). Refining the identity of mesenchymal cell types associated with murine periosteal and endosteal bone. J Biol Chem 300, 107158. 10.1016/j.jbc.2024.107158. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R29] 29.Baccin C., Al-Sabah J., Velten L., Helbling P.M., Grünschläger F., Hernández-Malmierca P., Nombela-Arrieta C., Steinmetz L.M., Trumpp A., and Haas S. (2020). Combined single-cell and spatial transcriptomics reveal the molecular, cellular and spatial bone marrow niche organization. Nat Cell Biol 22, 38–48. 10.1038/s41556-019-0439-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R30] 30.Translocation of proteins across the endoplasmic reticulum. I. Signal recognition protein (SRP) binds to in-vitro-assembled polysomes synthesizing secretory protein (1981). J Cell Biol 91, 545–550. 10.1083/jcb.91.2.545. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R31] 31.Kuge O., Dascher C., Orci L., Rowe T., Amherdt M., Plutner H., Ravazzola M., Tanigawa G., Rothman J.E., and Balch W.E. (1994). Sar1 promotes vesicle budding from the endoplasmic reticulum but not Golgi compartments. J Cell Biol 125, 51–65. 10.1083/jcb.125.1.51. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R32] 32.Zhong L., Yao L., Tower R.J., Wei Y., Miao Z., Park J., Shrestha R., Wang L., Yu W., Holdreith N., et al. (2020). Single cell transcriptomics identifies a unique adipose lineage cell population that regulates bone marrow environment. eLife 9, e54695. 10.7554/eLife.54695. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R33] 33.Laurin M., and Côté J.-F. (2014). Insights into the biological functions of Dock family guanine nucleotide exchange factors. Genes Dev 28, 533–547. 10.1101/gad.236349.113. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R34] 34.Bellanger J.M., Astier C., Sardet C., Ohta Y., Stossel T.P., and Debant A. (2000). The Rac1- and RhoG-specific GEF domain of Trio targets filamin to remodel cytoskeletal actin. Nat Cell Biol 2, 888–892. 10.1038/35046533. [DOI] [PubMed] [Google Scholar]

[R35] 35.Héraud C., Pinault M., Lagrée V., and Moreau V. (2019). p190RhoGAPs, the ARHGAP35- and ARHGAP5-Encoded Proteins, in Health and Disease. Cells 8, 351. 10.3390/cells8040351. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R36] 36.Ding X., Wu J., and Jiang C. (2010). ABCG2: a potential marker of stem cells and novel target in stem cell and cancer therapy. Life Sci 86, 631–637. 10.1016/j.lfs.2010.02.012. [DOI] [PubMed] [Google Scholar]

[R37] 37.Wei Q., Nakahara F., Asada N., Zhang D., Gao X., Xu C., Alfieri A., Brodin N.P., Zimmerman S.E., Mar J.C., et al. (2020). Snai2 Maintains Bone Marrow Niche Cells by Repressing Osteopontin Expression. Developmental Cell 53, 503–513.e5. 10.1016/j.devcel.2020.04.012. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R38] 38.Kalucka J., de Rooij L.P.M.H., Goveia J., Rohlenova K., Dumas S.J., Meta E., Conchinha N.V., Taverna F., Teuwen L.-A., Veys K., et al. (2020). Single-Cell Transcriptome Atlas of Murine Endothelial Cells. Cell 180, 764–779.e20. 10.1016/j.cell.2020.01.015. [DOI] [PubMed] [Google Scholar]

[R39] 39.Williams M., Lyu M.S., Yang Y.L., Lin E.P., Dunbrack R., Birren B., Cunningham J., and Hunter K. (1999). Ier5, a novel member of the slow-kinetics immediate-early genes. Genomics 55, 327–334. 10.1006/geno.1998.5679. [DOI] [PubMed] [Google Scholar]

[R40] 40.Yu H., Lin L., Zhang Z., Zhang H., and Hu H. (2020). Targeting NF-κB pathway for the therapy of diseases: mechanism and clinical study. Sig Transduct Target Ther 5, 209. 10.1038/s41392-020-00312-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R41] 41.Tikhonova A.N., Dolgalev I., Hu H., Sivaraj K.K., Hoxha E., Cuesta-Domínguez Á., Pinho S., Akhmetzyanova I., Gao J., Witkowski M., et al. (2019). The bone marrow microenvironment at single-cell resolution. Nature 569, 222–228. 10.1038/s41586-019-1104-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R42] 42.Lange M., Bergen V., Klein M., Setty M., Reuter B., Bakhti M., Lickert H., Ansari M., Schniering J., Schiller H.B., et al. (2022). CellRank for directed single-cell fate mapping. Nat Methods 19, 159–170. 10.1038/s41592-021-01346-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R43] 43.Gupta R.K., Mepani R.J., Kleiner S., Lo J.C., Khandekar M.J., Cohen P., Frontini A., Bhowmick D.C., Ye L., Cinti S., et al. (2012). Zfp423 Expression Identifies Committed Preadipocytes and Localizes to Adipose Endothelial and Perivascular Cells. Cell Metabolism 15, 230–239. 10.1016/j.cmet.2012.01.010. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R44] 44.Shao M., Ishibashi J., Kusminski C.M., Wang Q.A., Hepler C., Vishvanath L., MacPherson K.A., Spurgin S.B., Sun K., Holland W.L., et al. (2016). Zfp423 Maintains White Adipocyte Identity through Suppression of the Beige Cell Thermogenic Gene Program. Cell Metab 23, 1167–1184. 10.1016/j.cmet.2016.04.023. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R45] 45.Hong S.H.H., Lu X., Nanes M.S., and Mitchell J. (2009). Regulation of osterix (Osx, Sp7) and the Osx promoter by parathyroid hormone in osteoblasts. Journal of Molecular Endocrinology 43, 197–207. 10.1677/JME-09-0012. [DOI] [PubMed] [Google Scholar]

[R46] 46.Zhou B.O., Ding L., and Morrison S.J. Hematopoietic stem and progenitor cells regulate the regeneration of their niche by secreting Angiopoietin-1. eLife 4, e05521. 10.7554/eLife.05521. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R47] 47.Pinho S., Wei Q., Maryanovich M., Zhang D., Balandrán J.C., Pierce H., Nakahara F., Di Staulo A., Bartholdy B.A., Xu J., et al. (2022). VCAM1 confers innate immune tolerance on haematopoietic and leukaemic stem cells. Nat Cell Biol 24, 290–298. 10.1038/s41556-022-00849-4. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R48] 48.Prisby R., Guignandon A., Vanden-Bossche A., Mac-Way F., Linossier M.-T., Thomas M., Laroche N., Malaval L., Langer M., Peter Z.-A., et al. (2011). Intermittent PTH(1–84) is osteoanabolic but not osteoangiogenic and relocates bone marrow blood vessels closer to bone-forming sites. J Bone Miner Res 26, 2583–2596. 10.1002/jbmr.459. [DOI] [PubMed] [Google Scholar]

[R49] 49.Mohanakrishnan V., Sivaraj K.K., Jeong H.-W., Bovay E., Dharmalingam B., Bixel M.G., Dinh V.V., Petkova M., Paredes Ugarte I., Kuo Y.-T., et al. (2024). Specialized post-arterial capillaries facilitate adult bone remodelling. Nat Cell Biol 26, 2020–2034. 10.1038/s41556-024-01545-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R50] 50.Smith N.P., Janton C., Alcantara A.C.M., George M., Mankongtreecheep K., Mao G., Villani A.-C., and Kronenberg H.M. (2025). Single cell RNA sequencing shows that cells expressing Sox9 postnatally populate most skeletal lineages in mouse bone. J Bone Miner Res 40, 799–812. 10.1093/jbmr/zjaf043. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R51] 51.Nakashima K., Zhou X., Kunkel G., Zhang Z., Deng J.M., Behringer R.R., and de Crombrugghe B. (2002). The novel zinc finger-containing transcription factor osterix is required for osteoblast differentiation and bone formation. Cell 108, 17–29. 10.1016/s0092-8674(01)00622-5. [DOI] [PubMed] [Google Scholar]

[R52] 52.Wein M.N., and Kronenberg H.M. (2018). Regulation of Bone Remodeling by Parathyroid Hormone. Cold Spring Harb Perspect Med 8, a031237. 10.1101/cshperspect.a031237. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R53] 53.Katsuno Y., and Derynck R. (2021). Epithelial plasticity, epithelial-mesenchymal transition, and the TGF-β family. Developmental Cell 56, 726–746. 10.1016/j.devcel.2021.02.028. [DOI] [PubMed] [Google Scholar]

[R54] 54.Fournier P.G.J., Juárez P., Jiang G., Clines G.A., Niewolna M., Kim H.S., Walton H.W., Peng X.H., Liu Y., Mohammad K.S., et al. (2015). The TGF-β Signaling Regulator PMEPA1 Suppresses Prostate Cancer Metastases to Bone. Cancer Cell 27, 809–821. 10.1016/j.ccell.2015.04.009. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R55] 55.Xu F., Zhang J., Hu G., Liu L., and Liang W. (2017). Hypoxia and TGF-β1 induced PLOD2 expression improve the migration and invasion of cervical cancer cells by promoting epithelial-to-mesenchymal transition (EMT) and focal adhesion formation. Cancer Cell Int 17, 54. 10.1186/s12935-017-0420-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R56] 56.Gjaltema R.A.F., de Rond S., Rots M.G., and Bank R.A. (2015). Procollagen Lysyl Hydroxylase 2 Expression Is Regulated by an Alternative Downstream Transforming Growth Factor β−1 Activation Mechanism. J Biol Chem 290, 28465–28476. 10.1074/jbc.M114.634311. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R57] 57.Ranganathan P., Agrawal A., Bhushan R., Chavalmane A.K., Kalathur R.K.R., Takahashi T., and Kondaiah P. (2007). Expression profiling of genes regulated by TGF-beta: Differential regulation in normal and tumour cells. BMC Genomics 8, 98. 10.1186/1471-2164-8-98. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R58] 58.Inhibition of Sca-1-positive skeletal stem cell recruitment by alendronate blunts the anabolic effects of parathyroid hormone on bone remodeling - PubMed https://pubmed.ncbi.nlm.nih.gov/21040899/. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R59] 59.Kobelski M.M., Ramchand S.K., Tsai J.N., Leder B.Z., and Demay M.B. (2024). Evaluation of Osteogenic Phenotype in Postmenopausal Women Receiving Anabolic and Antiresorptive Osteoporosis Therapies. Endocrinology 165, bqae142. 10.1210/endocr/bqae142. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R60] 60.Chu N.T.L., Dregval O., Chang Y.-W., Kriukov E., Tian X., Liu X., Trompet D., Zhang M.S., Li L., Li Z., et al. (2025). Quantitative Multicolored Deep Imaging of Human Bones Reveals a Composite Osteo-Sinusoidal Niche for Mesenchymal Stromal Cells. Preprint at bioRxiv, 10.1101/2025.10.07.680053 https://doi.org/10.1101/2025.10.07.680053. [DOI] [Google Scholar]

[R61] 61.Kimura T., Panaroni C., Rankin E., Purton L.E., and Wu J.Y. (2022). Loss of parathyroid hormone receptor signaling in osteoprogenitors is associated with accumulation of multiple hematopoietic lineages in the bone marrow. J Bone Miner Res 37, 1321–1334. 10.1002/jbmr.4568. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R62] 62.Greenbaum A., Hsu Y.-M.S., Day R.B., Schuettpelz L.G., Christopher M.J., Borgerding J.N., Nagasawa T., and Link D.C. (2013). CXCL12 in early mesenchymal progenitors is required for haematopoietic stem-cell maintenance. Nature 495, 227–230. 10.1038/nature11926. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R63] 63.Kaneko T., Yari S., Kikuta J., Omatsu Y., Seno S., Kikuchi S., Sato K., Fujii K., Sudo T., Hasegawa T., et al. (2025). The RANK/RANKL axis controls vascular dynamics in the bone marrow. Proceedings of the National Academy of Sciences 122, e2425366122. 10.1073/pnas.2425366122. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R64] 64.Lu J., He Q., Wang H., Yao L., Duffy M., Guo H., Braun C., Zhou Y., Liang Q., Lin Y., et al. (2025). Bone marrow adipogenic lineage precursors are the major regulator of bone resorption in adult mice. Bone Res 13, 39. 10.1038/s41413-025-00405-4. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R65] 65.Linkhart T.A., Mohan S., and Baylink D.J. (1996). Growth factors for bone growth and repair: IGF, TGFβ and BMP. Bone 19, S1–S12. 10.1016/S8756-3282(96)00138-X. [DOI] [PubMed] [Google Scholar]

[R66] 66.Wu X., Pang L., Lei W., Lu W., Li J., Li Z., Frassica F.J., Chen X., Wan M., and Cao X. (2010). Inhibition of Sca-1-positive skeletal stem cell recruitment by alendronate blunts the anabolic effects of parathyroid hormone on bone remodeling. Cell Stem Cell 7, 571–580. 10.1016/j.stem.2010.09.012. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R67] 67.Crane J.L., and Cao X. (2014). Bone marrow mesenchymal stem cells and TGF-β signaling in bone remodeling. J Clin Invest 124, 466–472. 10.1172/JCI70050. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R68] 68.Kim S.W., Pajevic P.D., Selig M., Barry K.J., Yang J.-Y., Shin C.S., Baek W.-Y., Kim J.-E., and Kronenberg H.M. Intermittent parathyroid hormone administration converts quiescent lining cells to active osteoblasts. 10.1002/jbmr.1665. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R69] 69.Leder B.Z., Tsai J.N., Uihlein A.V., Burnett-Bowie S.-A.M., Zhu Y., Foley K., Lee H., and Neer R.M. (2013). Two Years of Denosumab and Teriparatide Administration in Postmenopausal Women With Osteoporosis (The DATA Extension Study): A Randomized Controlled Trial. Lancet 382, 50–56. 10.1016/S0140-6736(13)60856-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R70] 70.Omatsu Y., Aiba S., Maeta T., Higaki K., Aoki K., Watanabe H., Kondoh G., Nishimura R., Takeda S., Chung U., et al. (2022). Runx1 and Runx2 inhibit fibrotic conversion of cellular niches for hematopoietic stem cells. Nat Commun 13, 2654. 10.1038/s41467-022-30266-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R71] 71.Wein M.N., Liang Y., Goransson O., Sundberg T.B., Wang J., Williams E.A., O’Meara M.J., Govea N., Beqo B., Nishimori S., et al. (2016). SIKs control osteocyte responses to parathyroid hormone. Nat Commun 7, 13176. 10.1038/ncomms13176. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R72] 72.Lindsay R., Nieves J., Formica C., Henneman E., Woelfert L., Shen V., Dempster D., and Cosman F. (1997). Randomised controlled study of effect of parathyroid hormone on vertebral-bone mass and fracture incidence among postmenopausal women on oestrogen with osteoporosis. Lancet 350, 550–555. 10.1016/S0140-6736(97)02342-8. [DOI] [PubMed] [Google Scholar]

[R73] 73.Gatti D., Viapiana O., Idolazzi L., Fracassi E., Rossini M., and Adami S. (2011). The Waning of Teriparatide Effect on Bone Formation Markers in Postmenopausal Osteoporosis Is Associated with Increasing Serum Levels of DKK1. J Clin Endocrinol Metab 96, 1555– 1559. 10.1210/jc.2010-2552. [DOI] [PubMed] [Google Scholar]

[R74] 74.Finak G., McDavid A., Yajima M., Deng J., Gersuk V., Shalek A.K., Slichter C.K., Miller H.W., McElrath M.J., Prlic M., et al. (2015). MAST: a flexible statistical framework for assessing transcriptional changes and characterizing heterogeneity in single-cell RNA sequencing data. Genome Biol 16, 278. 10.1186/s13059-015-0844-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R75] 75.Soeda T., Deng J.M., de Crombrugghe B., Behringer R.R., Nakamura T., and Akiyama H. (2010). Sox9-expressing precursors are the cellular origin of the cruciate ligament of the knee joint and the limb tendons. Genesis 48, 635–644. 10.1002/dvg.20667. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R76] 76.Kalajzic Z., Liu P., Kalajzic I., Du Z., Braut A., Mina M., Canalis E., and Rowe D.W. (2002). Directing the expression of a green fluorescent protein transgene in differentiated osteoblasts: comparison between rat type I collagen and rat osteocalcin promoters. Bone 31, 654–660. 10.1016/s8756-3282(02)00912-2. [DOI] [PubMed] [Google Scholar]

[R77] 77.Ara T., Tokoyoda K., Sugiyama T., Egawa T., Kawabata K., and Nagasawa T. (2003). Long-term hematopoietic stem cells require stromal cell-derived factor-1 for colonizing bone marrow during ontogeny. Immunity 19, 257–267. 10.1016/s1074-7613(03)00201-2. [DOI] [PubMed] [Google Scholar]

[R78] 78.Kobayashi T., Chung U., Schipani E., Starbuck M., Karsenty G., Katagiri T., Goad D.L., Lanske B., and Kronenberg H.M. (2002). PTHrP and Indian hedgehog control differentiation of growth plate chondrocytes at multiple steps. Development 129, 2977–2986. 10.1242/dev.129.12.2977. [DOI] [PubMed] [Google Scholar]

[R79] 79.Akiyama H., Kim J.-E., Nakashima K., Balmes G., Iwai N., Deng J.M., Zhang Z., Martin J.F., Behringer R.R., Nakamura T., et al. (2005). Osteo-chondroprogenitor cells are derived from Sox9 expressing precursors. Proc Natl Acad Sci U S A 102, 14665–14670. 10.1073/pnas.0504750102. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R80] 80.Yoon S.-H., Meyer M.B., Arevalo C., Tekguc M., Zhang C., Wang J.S., Castro Andrade C.D., Strauss K., Sato T., Benkusky N.A., et al. (2023). A parathyroid hormone/salt-inducible kinase signaling axis controls renal vitamin D activation and organismal calcium homeostasis. J Clin Invest 133, e163627. 10.1172/JCI163627. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R81] 81.Hao Y., Stuart T., Kowalski M.H., Choudhary S., Hoffman P., Hartman A., Srivastava A., Molla G., Madad S., Fernandez-Granda C., et al. (2024). Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nat Biotechnol 42, 293–304. 10.1038/s41587-023-01767-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R82] 82.Wolf F.A., Angerer P., and Theis F.J. (2018). SCANPY: large-scale single-cell gene expression data analysis. Genome Biol 19, 15. 10.1186/s13059-017-1382-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R83] 83.Bergen V., Lange M., Peidli S., Wolf F.A., and Theis F.J. (2020). Generalizing RNA velocity to transient cell states through dynamical modeling. Nat Biotechnol 38, 1408–1414. 10.1038/s41587-020-0591-3. [DOI] [PubMed] [Google Scholar]

[R84] 84.Zappia L., and Oshlack A. (2018). Clustering trees: a visualization for evaluating clusterings at multiple resolutions. Gigascience 7, giy083. 10.1093/gigascience/giy083. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R85] 85.Xie Z., Bailey A., Kuleshov M.V., Clarke D.J.B., Evangelista J.E., Jenkins S.L., Lachmann A., Wojciechowicz M.L., Kropiwnicki E., Jagodnik K.M., et al. (2021). Gene Set Knowledge Discovery with Enrichr. Curr Protoc 1, e90. 10.1002/cpz1.90. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R86] 86.La Manno G., Soldatov R., Zeisel A., Braun E., Hochgerner H., Petukhov V., Lidschreiber K., Kastriti M.E., Lönnerberg P., Furlan A., et al. (2018). RNA velocity of single cells. Nature 560, 494–498. 10.1038/s41586-018-0414-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

This is a preprint.

Intermittent parathyroid hormone employs autonomous and non-autonomous mechanisms to drive osteogenesis from Ebf3-expressing skeletal progenitor cells

Byron SK Chan

Bingzi Dong

Majd George

Juwell Wu

Birol Ay

Daniel J Brooks

Mary L Bouxsein

Benjamin Z Leder

Garyfallia Papaioannou

Karin Gustafsson

David T Scadden

Takashi Nagasawa

Henry M Kronenberg

Charles P Lin

Marc N Wein

Abstract

Introduction

Results

Intermittent PTH treatment recruits Ebf3-lineage cells to bone surfaces and induces their differentiation into mature osteoblasts

Figure 1. Intermittent parathyroid hormone treatment stimulates Ebf3-lineage cell recruitment to bone surfaces and promotes their osteoblastic differentiation.

Adipogenic differentiation of Ebf3-lineage cells is inhibited by PTH

Ebf3-lineage osteogenic/stromal subpopulations are increased by PTH

Figure 2. Ebf3-lineage cells can directly respond to iPTH.

Ebf3-lineage cells can respond directly to PTH

Transcriptomic effects of PTH on Ebf3-lineage cells

Figure 3. iPTH destabilizes the CAR cell phenotype.

A subset of Ebf3-lineage endothelial cells does not change in numbers in response to PTH

Sox9-lineage cells rarely differentiate into mature osteoblasts following PTH treatment in skeletally mature mice

PTH-induced osteoblastic differentiation of Ebf3-lineage cells is inhibited by Sp7 deletion

Figure 4. Sp7 is required for the iPTH-stimulated maturation of Ebf3-lineage cells into osteoblasts.

Non-cell-autonomous recruitment of Ebf3-lineage cells to bone surfaces

Figure 5. Pth1r is required for iPTH-mediated suppression of CAR cell-enforcing transcription factors in Ebf3-lineage cells and for their maturation into osteoblasts.

PTH-stimulated recruitment of Ebf3-lineage cells to bone surfaces requires osteoclasts and is associated with TGFß-responsiveness

Figure 6. Osteoclasts are required for iPTH-stimulated recruitment of Ebf3-lineage cells to bone surfaces.

Figure 7. Recruitment of Ebf3-lineage cells to bone surfaces in response to iPTH is associated with TGFß-response.

Ebf3-lineage cells are important contributors to overall bone anabolism of iPTH

Figure 8. Ebf3-lineage cells are substantial contributors to the overall bone anabolism of iPTH.

Ebf3-lineage cells continue to participate in the anabolic action of PTH in the setting of ovariectomy

CAR cells in teriparatide-treated postmenopausal women show reduced expression of lineage-enforcing transcription factors and increased TGFß-responsive genes

Figure 9. Destabilization of the CAR cell state by PTH occurs in human CAR cells.

Figure 10. Model of CAR cell autonomous and non-cell-autonomous mechanisms by which iPTH promotes CAR cell-mediated osteoblastogenesis.

Discussion

CAR cell-autonomous effects of PTH

CAR cell-extrinsic effects of PTH

Relative contribution of CAR cells to the bone anabolism of iPTH

Functional heterogeneity among CAR cells

Clinical data for OPG-Fc + PTH

Future Directions

Study Limitations

Summary

Methods

Mouse strains

Tamoxifen administration

Parathyroid hormone and OPG-Fc administration

Histology

MATLAB image analysis

Microscopy

Immunohistochemistry

Bone single cell isolation

Fluorescence activated cell sorting (FACS)

FACS antibodies

Cell proliferation and apoptosis flow cytometric assay

In vitro culture of Ebf3-lineage cells

Intracellular cAMP assay

Phospho-CREB assays

Single cell RNA-sequencing analysis

RNA isolation and RT-qPCR

Tartrate-resistant acid phosphatase (TRAP) staining

Ovariectomy

MicroCT analysis

Serum bone turnover marker ELISAs

Human bone marrow stromal cells

Statistical analysis

Supplementary Material

Acknowledgements

Footnotes

References