Constitutive activation of TGF-β receptor type I promotes cortical and cancellous bone formation in adult mice
Tipthanan Chotipinit, Parichart Toejing, Chinnatam Phetkong, Somyoth Sridurongrit, Asada Leelahavanichkul, Julia F. Charles, Sutada Lotinun
TL;DR
Activating a specific TGF-β receptor in adult mice increases bone formation by boosting osteoblasts and reducing osteoclasts.
Contribution
Demonstrates that constitutive TGF-β receptor type I activation enhances bone formation in adult mice.
Findings
Mice with active TβRI showed increased cortical and cancellous bone volume in femurs and mandibles.
Osteoblast numbers increased and osteoclast numbers decreased in TβRI-activated mice.
TβRI activation upregulated osteoblast genes and downregulated osteoclast genes in both in vivo and in vitro models.
Abstract
Transforming growth factor β (TGF-β) is a pluripotent cytokine that plays a pivotal role in regulating bone remodeling. In this study, we investigated the skeletal phenotype of 24-week-old transgenic female mice expressing a constitutively active TGF-β receptor type I (TβRI) under the control of inducible Mx1-Cre promoter. Poly(I:C) injection was used to induce expression of TβRI to generate Mx1;TβRICA mice. In Mx1;TβRICA mice, serum calcium levels were increased, while parathyroid hormone (PTH) levels were decreased. Micro-computed tomography (μCT) analysis revealed a significant increase in cancellous and cortical bone volume in femurs and mandibles of Mx1;TβRICA mice compared to wild type mice. Histomorphometric analysis confirmed that this enhanced bone volume was associated with an increased number of osteoblasts and a reduced number of osteoclasts. Constitutive TβRI activation…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Fig 1
Fig 2
Fig 3
Fig 4
Fig 5
Fig 6
Fig 7
Fig 8
Fig 9
Fig 10
Fig 11
Fig 12- —The Ratchadaphiseksomphot Endowment Fund, Chulalongkorn University (The Exchange Faculty Travel Grant)
- —Thailand Science Research and Innovation Fund
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsTGF-β signaling in diseases · Bone health and osteoporosis research · Bone Metabolism and Diseases
Introduction
Transforming growth factor beta (TGF-β) is a growth factor associated with several important cellular functions, including cell proliferation, cell survival, cell differentiation, and cell migration. TGF-β is present in the extracellular matrix as a latent complex associated with its pro-domain [1]. It comprises three isoforms, TGF-β1, TGF-β2, and TGF-β3, which are synthesized as inactive precursors that require activation before binding to a receptor system composed of type I and type II transmembrane serine/threonine kinase receptors, TGF-β receptor type I (TβRI) and type II (TβRII) [1]. In the absence of ligand, both TβRI and TβRII exist as homodimers. TGF-β signal transduction involves both Smad-dependent and Smad-independent pathways. In the Smad-dependent pathway, upon ligand binding, TβRII recruits TβRI to form a heterotetrameric TβRII/TβRI complex. The constitutively active TβRII kinase then transphosphorylates the GS domain of TβRI, leading to its activation. Activated TβRI subsequently phosphorylates the transcription factors Smad2 and Smad3. The phosphorylated Smad2 and Smad3 form a complex with Smad4, which translocates to the nucleus to regulate gene expression [1]. In Smad-independent pathways, TGF-β can activate other signaling cascades, including MAPK (p38, JNK, and ERK), PI3K/Akt, Rho GTPase, Wnt, and Notch pathways, to regulate cellular functions [1].
All three isoforms of TGF-β were found in bone, but TGF-β1 is the most abundant [2]. TGF-β1 is expressed in bone marrow cells, chondrocytes, and cartilaginous matrix of both neonatal and adult mice [3]. It plays an essential role in bone development and homeostasis [4]. Previous studies have reported that TGF-β1 affects both osteoblasts and osteoclasts. It is associated with every stage of bone formation through recruiting osteoblast progenitors and stimulating their proliferation, as well as activating the early stages of differentiation. However, it inhibited later stages of differentiation and mineralization [4]. TGF-β1 exerted biphasic effects on osteoclast differentiation in a dose-dependent manner. Low doses of TGF-β1 induce the migration of macrophages to bone resorption sites, whereas high doses inhibit the migration of osteoclast precursors [5]. Additionally, low doses of TGF-β1 stimulate osteoclast differentiation by enhancing the expression of Csf1 and Tnfsf11 and bone resorption activity, while high doses suppress these effects. Furthermore, TGF-β1 binds to its receptor on osteoclasts, promoting osteoclast differentiation via the TRAF6–TAB1–TAK1 signaling complex [6].
Many skeletal disorders are linked to mutations in genes involved in TGF-β signaling, such as Camurati-Englemann and Loeys-Dietz syndrome [4]. In mice, Tgfb1^−/−^Rag2^−/−^ mice exhibited a loss of trabecular bone volume and thickness, along with increased trabecular bone separation. Moreover, immunostaining for Runx2 was not detected in femurs of Tgfb1^−/−^Rag2^−/−^ mice, suggesting a deficiency of osteoblasts on bone surfaces [7]. Similarly, deletion of TβRI in skeletal progenitor cells reduced trabecular bone and decreased the number of osteoblasts, indicating that TGF-β signaling via TβRI promotes pre-osteoblast commitment and early differentiation [8]. Likewise, deletion of TβRII in mesenchyme inhibited TGF-β signaling and resulted in severe calvarial bone defects, suggesting impaired intramembranous bone formation. This study also showed limb shortening due to decreased chondrocyte proliferation [9]. Moreover, delayed skull development following deletion of TβRII was accompanied by downregulation of essential osteoblast transcription factors, including Runx2 and Sp7, both of which are critical for osteoblast maturation and function [10].
In this study, we investigated the skeletal phenotype of 24-week-old female Mx1;TβRI^CA^ mice with constitutively active TβRI. Unlike the osteopenic phenotype observed at earlier time points (e.g., 9 weeks), our results demonstrated that constitutive activation of TβRI led to enhanced bone formation in the femurs and mandibles. Mx1;TβRI^CA^ osteoblasts had increased alkaline phosphatase and mineralization whereas Mx1;TβRI^CA^ osteoclast cultures showed decreased TRAP positive cells. In addition, osteoblast-related genes, including Runx2, Sp7, Alpl, Bglap, Col1a1, and Ptch2, were upregulated, while osteoclast-related genes, including Ctsk and Acp5, were downregulated in femurs of Mx1;TβRI^CA^ mice. Consistently, osteoblast-related genes, including Runx2, Sp7, Alpl, Col1a1, Ptch1, and Ptch2, were upregulated, while osteoclast-related genes, including Csf1, Ctsk, and Acp5, were downregulated in mandibles of Mx1;TβRI^CA^ mice. Likewise, the expression of osteoblast-related genes was enhanced in Mx1;TβRI^CA^ osteoblasts, whereas the expression of osteoclast-related genes was reduced in Mx1;TβRI^CA^ osteoclasts. This was accompanied by increased serum calcium levels and decreased serum PTH levels in Mx1;TβRI^CA^ mice*.*
Materials and methods
Mice
TβRI^CA^ mice provided by Dr. Laurent Bartholin were obtained from the Department of Anatomy, Faculty of Science, Mahidol University, Bangkok, Thailand. The development of TβRI^CA^ mice was conducted in accordance with the methods described by Vincent et al. [11]. Briefly, a constitutively active TGFβ type I receptor was knocked-in into the hypoxanthine phosphoribosyl-transferase (Hprt) locus on the X chromosome. Thus, TβRI^CA^ is X-linked and only female mice were studied. TβRI^CA^ mice were crossed with mice expressing Cre recombinase under the control of the interferon-inducible Mx1 promoter to generate Mx1;TβRI^CA^ mice. Activation of Mx1-Cre was performed by i.p. injection of 3 doses of 10 μg/g body weight polyinosinic-polycytidylic acid (poly(I:C)) every other day at 3 weeks of age. PCR genotyping of tail samples from 3-week-old mice was performed as in a previous study [12]. The animal protocol was approved by the Institutional Animal Care and Use Committee (IACUC) at the Faculty of Medicine, Chulalongkorn University (Protocol number: 006/2566), and followed the guidelines outlined in the Animal Research: Reporting In Vivo Experiments (ARRIVE, Singapore). Mice were housed under controlled conditions (25 ± 2°C) with a 12-hour light/dark cycle and had free access to standard rodent chow (C.P. Mice Feed, Perfect Companion Group Co., Ltd., Bangkok, Thailand) and water ad libitum at the Faculty of Medicine, Chulalongkorn University.
All experimental procedures were similar to previous studies [12]. 24-week-old female Mx1;TβRI^CA^ mice and their wild type (WT) mice were used in this study. 20 mg/kg calcein and 40 mg/kg demeclocycline were subcutaneously injected on days 9 and 2 before sacrifice, respectively. Upon completion of the experiment, mice were anesthetized with isoflurane and sacrificed by cardiac puncture. Blood samples were then collected, and serum was stored at −80°C for subsequent measurement of serum chemistries. The left mandibles and femurs were excised and preserved in 70% alcohol for micro-computed tomography (μCT), and the left femur was also used for histomorphometric analyses, while the right mandibles and femurs were snap-frozen in liquid nitrogen and stored at −80°C for RNA extraction and qPCR assessment.
Measurement of bone length
Bone length was measured using vernier calipers. Femur length was measured from the proximal to the distal ends, while mandible length was measured from the most anterior point of the dentary to the most posterior point of the articular condyle [13,14].
µCT analysis
μCT analysis of femurs and mandibles was conducted in accordance with recommended guidelines and scanned by a desktop μCT35 (Scanco Medical AG, Bassersdorf, Switzerland) [15] at 70 kV, 113 μA, integration time of 800 ms with nominal voxel size of 7 μm for femurs and 12 μm for mandibles. The region of interest (ROI) for femoral cortical bone analysis was defined using the automated midpoint finder in the Scanco software; a 0.6 mm length of diaphysis centered on the midpoint of the femur was analyzed. Cancellous bone was assessed in the distal metaphysis of the femur starting 0.2 mm below the growth plate and extending 2.1 mm proximal. The total number of slices containing trabecular bone was first determined in the group exhibiting the most extensive trabecular bone (Mx1;TβRI^CA^). Based on this reference, a standardized volume corresponding to 60% and 100% of the total trabecular region in Mx1;TβRI^CA^ and WT mice, respectively was selected and consistently applied to all experimental groups for quantitative analysis. The ROI for mandibular analysis was specified as bucco-lingual sections through the furcation area of the first molar, located between the mesial and distal roots. Scans were subjected to Gaussian filtration and segmentation using a bone/marrow cut of threshold at 190 mgHA/cm^3^ for femoral cancellous bone and 350 mgHA/cm^3^ for cortical bone, while those for mandibles were 220 mgHA/cm^3^ for cancellous bone and 260 mgHA/cm^3^ for cortical bone. Quantitative parameters included bone volume (BV/TV, %), trabecular thickness (Tb.Th, mm), trabecular number (Tb.N,/mm), trabecular separation (Tb.Sp, mm), connectivity density (Conn.D/mm³), structural model index (SMI, -), cortical volume (mm^3^), cortical thickness (mm), and bone mineral density (BMD, mgHA/cm³).
Histomorphometry
Undecalcified femurs were processed and embedded in methyl methacrylate. For mandibles, tissues were decalcified in 10% EDTA prepared in 0.05 M Tris buffer (pH 7.3) for 7 days, followed by dehydration through a graded ethanol series, and xylene, and embedded in paraffin. The 5 µm thick sections were cut by a Leica RM2255 microtome (Leica Biosystems, Nussloch, Germany). Undecalcified femur sections were stained with toluidine blue, whereas mandibular sections were stained with hematoxylin and eosin.
All measurements were performed in the distal metaphysis of the femurs, specifically 350 µm below the growth plate and mandible, using a semi-automatic image analysis system to determine osteoblast surface per bone surface (Ob.S/BS, %), osteoblast number per bone perimeter (N.Ob/B.Pm,/mm), osteoblast number per tissue area (N.Ob/T.Ar, mm²), osteoclast surface per bone surface (Oc.S/BS, %), osteoclast number per bone perimeter (N.Oc/B.Pm,/mm), osteoclast number per tissue area (N.Oc/T.Ar,/mm²), and eroded surface per bone surface (ES/BS, %). Additionally, cancellous bone volume per tissue volume (BV/TV, %), trabecular thickness (Tb.Th, μm), trabecular separation (Tb.Sp, μm), and trabecular number (Tb.N,/mm) were analyzed.
Unstained sections were used for determining the dynamic parameters in femurs, including the percentage of mineralizing surface per bone surface (MS/BS, %), mineral apposition rate (MAR, μm/day), bone formation rate per bone surface (BFR/BS, μm^3^/μm^2^/year), bone formation rate per bone volume (BFR/BV, %/year), and bone formation rate per tissue volume (BFR/TV, %/year). All histomorphometric measurements were performed using the OsteoMeasure system version 4.2.0.1 (OsteoMetric Inc., Decatur, GA, USA).following standardized nomenclature [16].
Microindentation analysis
The femurs were processed and embedded without demineralization in methyl methacrylate. The samples were then polished using 600-, 800-, and 1200-grit abrasive sandpaper, followed by 0.5 µm de-agglomerated alumina powder. A microhardness tester, FM-810 (Future-Tech Corp., Kawasaki, Japan) was used to assess the hardness of the cortical femoral bone at the midshaft. Data from five indentations with a Vickers diamond indenter applying a load of 25 gf for 10 seconds, each spaced 50 µm apart, were collected. The Vickers hardness (HV) was calculated using the equation HV = 1.854 × L/D², where D is the average length of the two diagonals (in millimeters), and L is the applied load (in kilograms).
Osteoblast culture
Primary osteoblasts were isolated following the protocol described by Chevalier et al. [17]. Briefly, bone marrow was flushed from long bones and mandibles. The remaining bone fragments were then minced into small pieces and digested in α-MEM containing 1 mg/mL collagenase type II (Worthington Biochemical Corporation, Lakewood, NJ, USA) for 2 h at 37°C with gentle agitation. The cell suspension was centrifuged at 5,000 rpm for 10 min. Bone fragments were then transferred to 75-cm² culture flasks containing α-MEM supplemented with 20% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin and maintained until reaching confluence. Subsequently, cells were cultured in α-MEM supplemented with 20% FBS, 5 mM β-glycerophosphate, 10 μM dexamethasone, and 50 μg/mL ascorbic acid for 7 days. Cells were isolated for qPCR analysis. In addition, osteoblasts were fixed in 3.7% formaldehyde and subjected to Fast Blue alkaline phosphatase and 2% alizarin red staining (Sigma, St. Louis, MO, USA).
Osteoclast culture
Bone marrow cells were harvested from long bones and mandibles by flushing with α-MEM. The cell suspension was passed through a 40-µm filter and cultured in α-MEM supplemented with 10% FBS, 100 U/mL penicillin, and 100 µg/mL streptomycin for 24 h. Non-adherent cells were subsequently collected and cultured in α-MEM containing 20 ng/mL M-CSF (R&D Systems, Minneapolis, MN, USA) for 2 days to generate bone marrow–derived macrophages (BMMs). Thereafter, BMMs were cultured in α-MEM supplemented with 20 ng/mL M-CSF and 3.3 ng/mL RANKL (R&D Systems, Minneapolis, MN, USA) for 6 days. Cells were then harvested for quantitative PCR analysis. In addition, TRAP-positive osteoclast number per total area (N.Oc/Ar) was quantified using OsteoMeasure software.
qPCR analysis
The femoral metaphysis and mandibles were snap-frozen in liquid nitrogen. Subsequently, the bones were placed in a pre-cooled mortar and grounded into a fine powder using a pestle, while maintaining the material in liquid nitrogen. RNA extraction was performed using TRIzol reagent (Invitrogen, Waltham, MA, USA), following the manufacturer’s procedure, and was followed by RNA cleanup using the RNeasy Mini Kit (Qiagen, Germantown, MD, USA). The RNA concentration was then measured using a NanoDrop 1000 (Thermo Fisher Scientific, Waltham, MA, USA). A total of 1 μg of RNA was reverse transcribed using the SuperScript VILO cDNA synthesis kit (Invitrogen, Carlsbad, CA, USA). qPCR was carried out with the Luna Universal qPCR master mix (New England Biolabs, Ipswich, MA, USA) using the CFX96™ Optics Module (Bio-Rad, Hercules, CA, USA). The qPCR program was set to 57°C for 39 cycles. Gene expression was normalized to Gapdh expression. Relative gene expression compared to average of WT was calculated using the CT method. All primer sequences are shown in Table 1.
Table 1: Oligonucleotide primers used for qPCR analysis.
Serum chemistry
Serum phosphorus and calcium levels were measured according to the manufacturer’s instructions (Stanbio Laboratory, Boerne, TX, USA; catalog numbers 0830−125 and 0155−225, respectively). Parathyroid hormone (PTH) levels were quantified using a commercially available ELISA kit (Quidel, San Diego, CA, USA; catalog number 60−2305), following the manufacturer’s protocol. The levels of cytokines IL-23, IL-1α, TNF-α, IFN-γ, MCP-1, IL-12p70, IL-1β, IL-10, IL-6, IL-27, IL-17A, IFN-β, and GM-CSF were measured by a multiplex bead-based assay (LEGENDplex^TM^) following the manufacturer’s instructions (LEGENDplex™; BioLegend, San Diego, CA, USA; catalog number 740446).
Statistical analysis
Data are represented as mean ± SEM. Data normality was assessed using the Shapiro–Wilk test. To compare differences between two groups, an unpaired t-test was used and analyzed using SPSS 24 (IBM, Armonk, NY, USA). A p-value of < 0.05 was considered statistically significant.
Results
Mx1;TβRICA mice increase serum calcium and decrease serum PTH levels
Body weight and bone length in Mx1;TβRI^CA^ mice were similar to those in WT mice (Fig 1A, 1B and 1C). To determine the changes in serum chemistries in Mx1;TβRI^CA^ mice, serum phosphorus, calcium, and PTH were studied. There were no significant differences in serum phosphorus levels between Mx1;TβRI^CA^ and WT mice (Fig 1D). Serum calcium levels were increased, while serum PTH levels were decreased in Mx1;TβRI^CA^ mice compared to WT mice (Fig 1E and 1F).
*Mx1;TβRICA mice alter serum calcium and PTH levels.(A) Body weight (n = 9–11) and (B) Femoral and (C) mandibular bone lengths of Mx1;TβRICA and WT mice (n = 6). (D) Serum phosphorus, (E) calcium, and (F) PTH levels of Mx1; TβRICA mice compared to WT mice (n = 6-8). Data are presented as mean ± SEM. p < 0.05 compared to WT mice.
Mx1;TβRICA mice have increased cancellous and cortical bone volume in femurs
To assess the impact of Mx1-Cre-induced constitutive activation of TβRI on femoral cancellous bone in 24-week-old mice, µCT analysis was performed. 3D reconstructed images of distal femoral trabecular and midshaft cortical bone demonstrated increased trabecular and cortical bone in Mx1;TβRI^CA^ mice compared to WT mice (Fig 2A). Cancellous bone volume, trabecular thickness, connectivity density, and bone mineral density were significantly higher in femoral bones of Mx1;TβRI^CA^ mice compared to WT mice (Fig 2B). In contrast, trends in trabecular number, trabecular separation, and structural model index comparing Mx1;TβRI^CA^ and WT mice were congruent with changes in BV/TV and Tb.Th but differences were not statistically significant. Additionally, cortical volume, cortical thickness, and bone mineral density were significantly increased in the femurs of Mx1;TβRI^CA^ mice (Fig 2C). These findings suggest that Mx1-Cre-induced constitutive activation of TβRI increases femoral bone in both cancellous and cortical compartments in mice.
*Mx1;TβRICA mice enhance femoral cancellous and cortical bone.(A) Representative 3D images of cross section of cancellous and cortical bone of femurs in Mx1;TβRICA and WT mice. (B) Cancellous analysis, BV/TV, Tb.Th, Tb.N, Tb.Sp, Conn.D, SMI and BMD in Mx1; TβRICA mice compared to WT mice (n = 6–7), Cortical analysis, cortical volume, cortical thickness, and BMD in Mx1; TβRICA mice compared to WT mice (n = 7). Data are shown as mean ± SEM. *p < 0.05, and *p < 0.01 compared to WT mice.
Mx1;TβRICA mice have increased cancellous and cortical bone volume in mandibles
To study the impact of constitutive activation of TβRI on mandibular cancellous bone in 24-week-old mice, µCT analysis was performed. Representative µCT images of WT and Mx1;TβRI^CA^ mandibles are shown (Fig 3A). Cancellous bone volume, trabecular thickness, and bone mineral density were significantly increased in the mandibles of Mx1;TβRI^CA^ mice compared to WT mice (Fig 3B). In contrast, trabecular number, trabecular separation, connectivity density, and structural model index were similar between the mandibular bones of Mx1;TβRI^CA^ and WT mice (Fig 3B). Furthermore, cortical volume, cortical thickness, and bone mineral density were significantly increased in the mandibles of Mx1;TβRI^CA^ mice (Fig 3C). Mx1-Cre-induced constitutive activation of TβRI resulted in increased cancellous and cortical compartments at multiple anatomic sites.
*Mx1;TβRICA mice increase mandibular cancellous and cortical bone.(A) Representative cross section through 3D reconstruction of cancellous and cortical mandibular bone in Mx1;TβRICA and WT mice. (B) Cancellous analysis, BV/TV, Tb.Th, Tb.N, Tb.Sp, Conn.D, SMI and BMD in Mx1; TβRICA mice compared to WT mice (n = 6). Cortical analysis, cortical volume, cortical thickness, and BMD in Mx1; TβRICA mice compared to WT mice (n = 7-10). Data are presented as mean ± SEM. *p < 0.05, **p < 0.01, and **p < 0.001 compared to WT mice.
Mx1;TβRICA mice have increased osteoblast and decreased osteoclast in femur and mandible
Histomorphometric analysis of femurs confirmed the µCT findings, showing increased cancellous bone in Mx1;TβRI^CA^ mice compared to WT mice. As shown in Table 2, Mx1;TβRI^CA^ mice showed an increase in cancellous bone volume, trabecular thickness, and trabecular number, along with a reduction in trabecular separation compared to WT mice. Additionally, osteoblast surface relative to bone surface, number of osteoblasts per bone perimeter, and number of osteoblasts per tissue area were all higher in Mx1;TβRI^CA^ mice. Representative histological sections showed increased cancellous bone volume and osteoblasts aligned along bone surface in Mx1;TβRI^CA^ mice compared to WT mice (Fig 4). Moreover, Mx1;TβRI^CA^ mice showed a reduction in osteoclast surface per bone surface, osteoclast number per bone perimeter, osteoclast number per tissue area, and eroded surface. Moreover, mineral apposition rate, bone formation rate per bone surface and bone formation rate per tissue volume were significantly increased in Mx1;TβRI^CA^ mice compared to WT mice. Therefore, our data suggested that Mx1;TβRI^CA^ mice exhibit alterations in bone turnover, characterized by increased bone formation and decreased bone resorption, leading to overall increased bone volume.
Table 2: Histomorphometric analysis of femoral and mandibular bone in Mx1;TβRICA and WT mice.
Mx1; TβRICA mice displayed increased femoral trabecular bone and osteoblasts.Representative images of toluidine blue staining of femoral trabecular bone from WT mice and Mx1;TβRICA mice. Red arrows indicate osteoblasts along trabecular bone surface. scale bar = 200 µm.
Histomorphometric analysis of the mandible further supported the µCT findings, revealing an increase in cancellous bone in Mx1;TβRI^CA^ mice compared with WT mice. As summarized in Table 2, Mx1;TβRI^CA^ mice exhibited significantly greater cancellous bone volume, trabecular thickness, along a decrease in trabecular separation. In addition, osteoblast surface relative to bone surface, number of osteoblasts per bone perimeter, and number of osteoblasts per tissue area were all elevated in Mx1;TβRI^CA^ mice. Furthermore, Mx1;TβRI^CA^ mice showed reduced osteoclast surface per bone surface, osteoclast number per bone perimeter, osteoclast number per tissue area, and eroded surface.
Mx1;TβRICA mice have increased femoral bone hardness
Mx1;TβRI^CA^ mice exhibited increased bone formation. In addition to the increase in bone volume, we determined whether the mechanical properties of the cortical bone were affected. We performed microindentation hardness testing at the midshaft of the femur. The femoral cortical bone was assessed using five indentations (Fig 5A), and hardness was subsequently measured. Mx1;TβRI^CA^ mice showed significantly greater cortical bone hardness compared to WT mice (Fig 5B).
*Mx1;TβRICA mice have increased cortical bone hardness in the femurs.(A) Representative image showing five indentation points on femoral cortical midshaft. (B) Quantification of femoral cortical bone hardness in Mx1;TβRICA mice compared to WT mice (n = 5-6). Results are presented as mean ± SEM. p < 0.05 compared to WT mice. HV, Vickers hardness.
Mx1;TβRICA mice have increased osteoblast and decreased osteoclast differentiation
Histomorphometric analysis showed that constitutive activation of TβRI was associated with increased osteoblast and decreased osteoclast number. To further examine the effect of sustained TβRI activation on osteoblast differentiation, in vitro assays were performed using primary osteoblast progenitors isolated from long bones and mandibles. Osteoblasts derived from both long bones and mandibles of Mx1;TβRI^CA^ mice exhibited increased ALP staining and activity and mineralization, indicating that expression of the constitutively active TβRI promotes osteoblast differentiation (Fig 6).
*Constitutive activation of TβRI increased osteoblast differentiation.(A) Representative ALP staining of osteoblasts derived from long bones and mandibles of Mx1;TβRICA mice and WT mice. Images were acquired at 4 × magnification; scale bar = 200 μm. (B) ALP activity (U/mL) in osteoblasts derived from long bones and mandibles of Mx1;TβRICA mice and WT mice (n = 4-5). (C) Representative alizarin red staining of osteoblasts derived from long bones and mandibles of Mx1;TβRICA mice and WT mice. Images were acquired at 4 × magnification; scale bar = 200 μm. (D) Mineralization (mM, n = 4-5), Data are presented as mean ± SEM. *p < 0.05 and *p < 0.01 compared to WT mice.
We examined whether Mx1;TβRI^CA^ mice exhibited alterations in osteoclast differentiation. In vitro osteoclasts were conducted using osteoclast precursors derived from long bones and mandibles. Consistent with the reduced bone resorption observed in vivo, constitutive activation of TβRI significantly decreased the number of TRAP-positive osteoclasts derived from BMMs of Mx1;TβRI^CA^ mice compared with WT mice (Fig 7). These findings indicate that TβRI signaling plays a critical role in regulating both osteoblast and osteoclast differentiation.
*Constitutive activation of TβRI suppresses osteoclast differentiation.(A) Representative TRAP staining of osteoclasts derived from long bones and mandibles of Mx1;TβRICA and WT mice (10 × magnification; scale bar = 100 μm). (B) Quantification of osteoclast number per area (N.Oc/Ar) in long bones and mandibles from Mx1;TβRICA mice compared with WT mice (n = 4-6). Data are presented as mean ± SEM. *p < 0.05 and *p < 0.01 compared to WT mice.
Mx1;TβRICA mice have increased osteoblast- and decreased osteoclast-related gene expression in femurs
qPCR was conducted to assess the expression of osteoblast-and osteoclast-associated transcripts in femurs in order to investigate the mechanisms by which constitutive activation of TβRI increased bone formation and decreased bone resorption. Mx1;TβRI^CA^ femurs showed higher levels of Runx2 and Sp7, key transcription factors for osteoblast differentiation, compared to WT controls. Additionally, key markers of bone formation, including Alpl, Bglap, Col1a1, and Ptch2, were significantly elevated in Mx1;TβRI^CA^ femurs (Fig 8). In contrast, Ctsk and Acp5, regulators of bone resorption, were significantly decreased in the femoral bone of Mx1;TβRI^CA^ mice (Fig 8). These findings suggest that Mx1;TβRI^CA^ mice enhanced femoral bone formation and suppressed bone resorption, characterized by increased osteoblast-related gene expression and decreased osteoclast-related gene expression.
*Mx1; TβRICA mice increase osteoblast-related and decrease osteoclast-related gene expression in femurs.qPCR analysis of osteoblast-related and osteoclast-related gene expression in femurs from Mx1; TβRICA mice (n = 5–7) compared to WT mice (n = 6–8). Results are presented as mean ± SEM. *p < 0.05, and *p < 0.01 compared to WT mice.
To validate the gene expression changes observed in the femur, primary osteoblasts and osteoclasts isolated from long bones were analyzed for the expression of osteoblast- and osteoclast-related genes using qPCR. Compared with WT mice, osteoblasts from Mx1;TβRI^CA^ mice exhibited significantly higher expression of Runx2 and Sp7, key transcription factors involved in osteoblast differentiation (Fig 9). In addition, the expression of bone formation–related markers, including Alpl, Bglap, Col1a1, and Ptch1, was markedly increased in Mx1;TβRI^CA^ osteoblasts. In contrast, the expression of osteoclast-associated genes, including Sufu, Ctsk, and Acp5, was significantly reduced in osteoclasts from Mx1;TβRI^CA^ mice (Fig 9).
*Mx1; TβRICA mice increase osteoblast-related and decrease osteoclast-related gene expression in primary bone cell cultures.qPCR analysis of osteoblast-related and osteoclast-related gene expression in vitro derived from long bone of Mx1; TβRICA mice compared to WT mice (n = 4-6). Results are presented as mean ± SEM. *p < 0.05, and *p < 0.01 compared to WT mice.
Mx1;TβRICA mice have increased osteoblast and decreased osteoclast-related gene expression in mandibles
qPCR was performed to examine the expression of osteoblast-and osteoclast-associated transcripts in mandibles to determine the mechanisms by which constitutive activation of TβRI increases bone formation and decreases bone resorption. Mx1;TβRI^CA^ mandibles showed higher levels of Runx2 and Sp7 compared to WT mice (Fig 10). Additionally, Alpl, Col1a1, Ptch1 and Ptch2 were significantly increased in Mx1;TβRI^CA^ mandibles. In contrast*, Csf1, Ctsk,* and Acp5 were significantly decreased in the mandibular bone of Mx1;TβRI^CA^ mice (Fig 10). These findings indicated that Mx1;TβRI^CA^ mice exhibited increased mandibular bone formation and decreased bone resorption determined by increased osteoblast-related gene expression and decreased osteoclast-related gene expression.
*Mx1; TβRICA mice increase osteoblast-related and decrease osteoclast-related gene expression in mandibles.qPCR analysis of osteoblast-related and osteoclast-related gene expression in mandibles from Mx1; T βRICA mice (n = 5–6) compared to WT mice (n = 5-6). Results are presented as mean ± SEM. *p < 0.05, and *p < 0.01 compared to WT mice.
To confirm the gene expression changes observed in the mandible, primary osteoblasts and osteoclasts isolated from mandibular bone were analyzed for the expression of osteoblast- and osteoclast-related genes using qPCR. Compared to WT mice, osteoblasts from Mx1;TβRI^CA^ mice exhibited significantly higher expression of Runx2 and Sp7, key transcription factors involved in osteoblast differentiation (Fig 11). In addition, the expression of bone formation–related markers, including Alpl, Col1a1, Ibsp, Ptch1, and Ptch2, was markedly increased in Mx1;TβRI^CA^ osteoblasts. In contrast, the expression of Ctsk was significantly reduced in osteoclasts from Mx1;TβRI^CA^ mice.
*Mx1; TβRICAmice increase osteoblast-related and decrease osteoclast-related gene expression in vitro derived from mandibles.qPCR analysis of osteoblast-related and osteoclast-related gene expression in vitro derived from mandibles of Mx1; TβRICA mice compared to WT mice (n = 4-6). Results are presented as mean ± SEM. *p < 0.05, **p < 0.01 and **p < 0.001 compared to WT mice.
Mx1;TβRICA mice have decreased pro-inflammatory cytokine and increased anti-inflammatory cytokine levels related to bone formation
Cytokines implicated in regulating bone turnover were assessed by measuring serum levels of pro-inflammatory and anti-inflammatory cytokines in Mx1;TβRI^CA^ and WT mice. Mx1;TβRI^CA^ mice showed significantly decreased serum concentrations of the pro-inflammatory cytokines, GM-CSF, IL-1α, IL-1β, IL-6, and IL-23 (Fig 12). In contrast, the anti-inflammatory cytokines, IL-10, IL-27, and IFN-β levels were significantly increased compared to WT mice. These findings suggest that enhanced bone formation in Mx1;TβRI^CA^ mice may be mediated by an increase in circulating anti-inflammatory cytokines and a decrease in pro-inflammatory cytokines.
*Mx1;TβRICA mice exhibited reduced pro-inflammatory cytokines and elevated anti-inflammatory cytokines.Flow cytometry analysis of serum levels of cytokines in Mx1;TβRICA mice (n = 5-6) compared to WT mice (n = 5-7). Data are presented as mean ± SEM. *p < 0.05, and *p < 0.01 compared to WT mice.
Discussion
TGF-β plays an essential role in the regulation of the differentiation and function of osteoblasts and osteoclasts, skeletal development, and homeostasis. It binds to its receptor, which is expressed on both osteoblasts and osteoclasts, leading to the stimulation of downstream signaling pathways. Physiologically, TGF-β signaling plays a tightly regulated and context-dependent role in bone remodeling. Transient activation of TGF-β promotes osteoblast recruitment and early-stage differentiation. In addition, TGF-β1 exerts dose-dependent effects on osteoclasts, stimulating osteoclast differentiation and migration at low concentrations while suppressing these processes at higher levels. These tightly controlled temporal and quantitative effects highlight the importance of balanced TGF-β signaling in maintaining skeletal homeostasis. [4]. Although the effects of TβRII on bone and cartilage function have been well reported, the function of TβRI in bone homeostasis is less well understood. In the present study, we investigated the effects of the constitutive TβRI activation on the skeletal phenotype in transgenic 24-week-old female Mx1;TβRI^CA^ mice with inducible expression of a constitutively active TβRI in hematopoietic stem cells (HSCs). In normal mice, Mx1-Cre is silent. It can be activated in HSCs after injection with poly (I:C), and also activates in monocytes and macrophages [18,19]. We verified the genotypes of the study mice by PCR, demonstrating that mice exhibited the Mx1-Cre band after poly (I:C) injection, confirming successful induction of Cre recombinase activity. In contrast to physiological condition, constitutive activation of TβRI in 24-week-old Mx1;TβRI^CA^ mice represents a state of persistent signaling associated with increased bone mass. This phenotype is characterized by enhanced osteoblast activity, potentially mediated by Hedgehog pathway activation, together with reduced bone resorption, likely resulting from impaired osteoclast differentiation.
Our results showed that Mx1;TβRI^CA^ mice exhibited increased bone mass at 24 weeks. This is in contrast to our previous study in young Mx1;TβRI^CA^ mice at 9 weeks of age, which exhibited reduced bone volume, as well as decreased cortical and cancellous thickness in both femurs and mandibles by reducing osteoblast differentiation and increasing osteoclast differentiation, possibly by suppressing Hedgehog signaling pathways [12]. In addition, it was reported that Lilrb4a, which is involved in osteoclast inhibition, and Hdac7, a Runx2 corepressor involved in bone turnover, were implicated in young female Mx1;TβRI^CA^ mice. Silencing Lilrb4a led to an increase in osteoclast numbers and upregulation of osteoclast marker genes, while knockdown of Hdac7 resulted in enhanced ALP activity, mineralization, and increased expression of osteoblast marker genes [20]. We also found that calcium levels were increased in Mx1;TβRI^CA^ mice, accompanied by decreased serum PTH. PTH controls calcium homeostasis, which is related to its actions in stimulating bone remodeling. Low calcium levels promote PTH secretion, which induces bone resorption, while high calcium levels have the opposite effect [21]. In addition, PTH inhibited phosphorylation of Smad2 and its subsequent association with Smad4, suppressing transcriptional activation in response to TGF-β signaling [22]. The discrepancy between the skeletal phenotype at 9 and 24 weeks of age may be due to age-dependent physiological changes or hormonal feedback regulation. Indeed, in our prior study we found that 9-week-old Mx1;TβRI^CA^ mice had higher PTH levels. This suggested a complex interaction between age-related changes, hormonal regulation, and TGF-β signaling in maintaining skeletal homeostasis. Furthermore, previous evidence found that TGF-β1 signaling was linked to calcium homeostasis. In particular, reduced TGF-β1 mRNA expression in mouse bone has been associated with decreased calcium levels, indicating that TGF-β1 may play a regulatory role in maintaining systemic calcium balance [23]. Previous studies have shown that TGF-β signaling can stimulate FGF23 production in osteoblastic cells. FGF23, in turn, promotes renal calcium reabsorption by increasing TRPV5 channel abundance in the distal tubules, thereby contributing to the maintenance of systemic calcium homeostasis [24,25].
Our results revealed that the constitutive activation of TβRI in 24-week-old Mx1;TβRI^CA^ mice led to increased bone mass by µCT analysis, as evidenced by enhanced cancellous bone volume, cortical bone volume, trabecular thickness, cortical thickness, and BMD in mandibles and femurs, compared to WT controls. In addition, microindentation testing revealed that Mx1;TβRI^CA^ mice had increased bone hardness. These findings were further supported by histomorphometric analysis. Importantly, histomorphometric analysis revealed a significant increase in BFR/BS, BFR/TV and osteoblast numbers and a decrease in osteoclast numbers suggesting increased bone formation and decreased resorption in 24-week-old Mx1;TβRI^CA^ mice. Numerous investigations have shown that TGF-β signaling is associated with bone formation [26]. Intravenous injection of TGF-β1 at a dose of 1000 µg/kg increased new endosteal bone formation in rats and rabbits [27]. In addition, TGF-β1-coated beta-tricalcium phosphate pellets enhanced new bone formation in a calvarial defect model [28]. Moreover, TGF-β signaling plays a role in early alveolar bone formation by stimulating the proliferation and differentiation of periodontal ligament mesenchymal stem cells [29]. Furthermore, deletion of TβRII in mesenchyme resulted in reduced parietal and frontal bones in the skull, indicating a defect in intramembranous bone formation, and led to limb shortening. Therefore, the absence of TβRII results in severe developmental defects in the limbs, skull, and axial bones [9]. Conditional knockout of TβRII in mice led to adverse effects on chondrogenic, and osteogenic proliferation, and differentiation, resulting in delayed and impaired endochondral bone formation with unclosed fracture lines [30]. In support of the essential function of TGF-β signaling in osteogenic differentiation, our findings suggested that TβRI activation signaling promoted bone formation in 24-week-old Mx1;TβRI^CA^ mice.
In this study, we explored the role of TGF-β signaling in regulating both osteoblast and osteoclast function. Osteoblasts and osteoclasts control bone homeostasis through bone formation and resorption, respectively. Osteoblasts derive from mesenchymal stem cells. The differentiation process is controlled by key transcription factors, such as Runx2, Osx, and Drosophila distal-less 5 (Dlx 5). Once Runx2 is activated, mesenchymal stem cells become pre-osteoblasts, which begin to produce early osteogenic markers, including ALP and collagen type I. These pre-osteoblasts then differentiate into mature osteoblasts, which produce bone matrix proteins, including osteopontin, osteocalcin, and bone sialoprotein. These proteins provide structural integrity and participate in the deposition of hydroxyapatite [31]. TGF-β has been shown to promote osteoblast differentiation during the early stages of development, while inhibit later stages of osteoblast maturation and mineralization [4,32]. In our study, we found that osteoblast differentiation was increased in primary osteoblasts derived from Mx1;TβRI^CA^ long bones and mandibles as indicated by increased alkaline phosphatase activity and mineralization. We also observed a marked upregulation of mRNA expression for key osteoblast transcription factors (Runx2 and Sp7) and osteogenic markers (Alpl, and Col1a1) in the femoral and mandibular bone in response to TβRI^CA^ expression. These results indicate that TGF-β signaling plays an important role in bone formation by promoting osteoblast proliferation and differentiation.
It is well established that Runx2 plays a role in the response to TGF-β1 in inducing osteogenic differentiation. A prior investigation indicated that Smad activated JunB, an upstream activator of Runx2 expression, and MAPK, are involved in the induction of Runx2 by TGF-β1 [33]. Our finding is consistent with previous studies showing that TGF-β1 promotes the proliferation and osteoblastic differentiation of marrow stromal cells, with increased total cell number, ALP activity, and osteocalcin [34]. Conversely, inhibition of TGF-β signaling using SB505124, a TβRII inhibitor, reduced osteoblast differentiation, as reflected by decreased ALP activity, mineralization, and downregulation of osteoblast-associated genes in human bone marrow mesenchymal stem cells [35]. Further, a study using a conditional knockout of the TβRI mice, in which TβRI was conditionally inactivated in skeletal progenitors during early bone development, showed that TGF-β signaling triggers osteoprogenitor proliferation, early differentiation, and commitment to the osteoblastic lineage via the MAPKs and Smad2/3 pathways [8]. In addition, previous studies have demonstrated that the Mx1-expressing cell population includes not only HSCs but also bone marrow mesenchymal stem cells, osteoprogenitors, and osteoblasts which were activated after poly (I:C) stimulation [16–18]. Our previous study in 9 weeks old mice found an increase in TβRI gene expression and ratios of pSmad2/total Smad2 and pSmad3/total Smad3 in osteoblasts isolated from Mx1;TβRI^CA^ mice compared to WT mice [11]. Therefore, the increase of osteoblast-related genes in 24-week-old Mx1;TβRI^CA^ mice might be through the activation of TβRI, which is associated with bone formation.
One feature of the bone phenotype of Mx1;TβRI^CA^ mice is altered expression of the genes Ptch1 and Ptch2, encoding multi-transmembrane proteins that bind to and act as receptors for Hedgehog ligands, thereby regulating the expression of downstream target genes [36]. In young Mx1;TβRI^CA^ mice, Ptch1/2 are downregulated, and bone formation is reduced, while in old Mx1;TβRI^CA^ mice, Ptch1/2 are upregulated, and bone formation is increased. The Hedgehog signaling pathway plays a critical role in the differentiation and development of osteoblasts. Hedgehog signaling pathway is required during intramembranous bone formation. Ihh^−/−^ mice had a reduction of cranial bone size [37], and activation of the Hedgehog signaling pathway promotes osteoblast differentiation by increasing the expression of Ptch1, BMP4, Runx2, and Opn under both normal and high-glucose conditions [38]. In addition, high expression of Ptch1 has been shown to enhance the early stages of the socket healing process after tooth extraction, facilitating trabecular bone formation [39]. Our findings suggest a potential crosstalk between the TGF-β and Hedgehog signaling pathways during TβRI activation, providing mechanistic insights into constitutive TβRI activation-induced bone formation.
In contrast to its role in osteoblast differentiation, a biphasic effect of TGF-β on osteoclasts has been reported, depending on its concentration [40,41]. M-CSF is closely associated with osteoclastogenesis. Hematopoietic cells are committed to macrophage colony-forming units (CFU-M) upon exposure to M-CSF. Once activated by the Tnfsf11-Tnfsf11A signaling pathway, CFU-M cells differentiate into mononucleated osteoclasts, which subsequently fuse to form multinucleated osteoclasts [42]. These multinucleated osteoclasts secrete acid and proteolytic enzymes, such as Ctsk and Acp5, to degrade bone during bone resorption [42]. Our results revealed that osteoclast differentiation was decreased in primary osteoclasts from Mx1;TβRI^CA^ long bones and mandibles. We also observed decreased expression of Ctsk and Acp5 in the femurs and reduced expression of Csf1, Ctsk, and Acp5 in the mandibles of Mx1;TβRI^CA^ mice compared to WT mice. In accordance with these findings, TGF-β signaling may modulate osteoclast activity, influencing bone resorption in these mice. Previous studies demonstrated that a low dose of TGF-β triggered osteoclast differentiation by enhancing the Tnfsf11/Tnfrsf1b ratio, while a high dose of TGF-β inhibited osteoclast differentiation. This biphasic effect is mediated by Smad1 and Smad3 signaling, respectively [43]. Therefore, transcriptional downregulation of osteoclast gene expression may be associated with the modulation of bone resorption in 24-week-old Mx1;TβRI^CA^ mice. The results of gene expression are similar in vitro compared to in vivo system.
Many cytokines are implicated in modulating bone turnover. Elevated levels of pro-inflammatory cytokines can enhance osteoclastogenesis. GM-CSF, a pro-inflammatory cytokine produced by activated T cells, macrophages, endothelial cells, and fibroblasts, can induce osteoclastogenesis and trigger fusion [44]. IL-1 also plays a key role in inflammation-mediated bone loss, with both IL-1α and IL-1β promoting osteoclast differentiation [45]. In addition, IL-6 facilitates osteoclastogenesis by enhancing Tnfsf11 expression, while inhibition of IL-6 blocks osteoclast formation [46]. Similarly, IL-23 promotes osteoclastogenesis by upregulating Tnfsf11A expression [47]. Elevated IL-23 levels have been associated with reduced trabecular bone density and bone loss in ovariectomized mice [48]. Interestingly, our study demonstrated a significant downregulation of these pro-inflammatory cytokines in Mx1;TβRI^CA^ mice. In contrast, we observed an upregulation of anti-inflammatory cytokines, including IL-10, IL-27, and IFN-β, in these mice. IL-10 suppresses osteoclastogenesis by inhibiting NFATc1 expression and its nuclear translocation [49]. It can also reduce osteoclast formation indirectly by lowering the levels of TNF-α, IL-1, and IL-6 [50]. IL-27 has been shown to inhibit the expression of c-Fos and NFATc1 by blocking Tnfsf11A-mediated activation of the ERK, p38, and NF-κB signaling pathways in osteoclast precursors [51]. IFN-β has also been suggested as an anti-osteoclastogenic cytokine, inhibiting osteoclastogenesis by the regulation of the JAK1/STAT3/c-Fos signaling pathway [52]. Changes in cytokine profiles may arise not only from direct effects on bone cells but also from altered immune and stromal cell contributions to the bone marrow microenvironment. Indeed, innate and adaptive immune cells have well-documented roles in regulating osteoclastogenesis and bone remodeling through production of pro- and anti-inflammatory cytokines, highlighting the potential for immune–bone crosstalk to modulate the skeletal phenotype in this model. Therefore, our results revealed that decreased pro-inflammatory cytokine levels and increased anti-inflammatory cytokine levels were associated with bone formation in Mx1;TβRI^CA^ mice.
In summary, the present study demonstrated that constitutive activation of TβRI in the 24-week-old Mx1;TβRI^CA^ mice resulted in increased bone mass due to enhanced bone formation and reduced bone resorption. Enhanced bone formation was supported by increased osteoblast numbers, possibly through activation of the Hedgehog signaling pathway. Decreased bone resorption, indicated by decreased osteoclast numbers, could be mediated by the impact of TGF-β on transcription control of the osteoclast differentiation program, by reductions in pro-osteoclastogenic inflammatory cytokine levels and/or reductions in PTH. These findings suggested that TGF-β could serve as a key therapeutic target for regulating bone remodeling in conditions characterized by abnormal bone formation or resorption, such as osteoporosis and other metabolic bone diseases.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Janssens K, ten Dijke P, Janssens S, Van Hul W. Transforming growth factor-beta 1 to the bone. Endocr Rev. 2005;26(6):743–74. doi: 10.1210/er.2004-0001 15901668 · doi ↗ · pubmed ↗
- 2Hering S, Isken E, Knabbe C, Janott J, Jost C, Pommer A, et al. TG Fbeta 1 and TG Fbeta 2 m RNA and protein expression in human bone samples. Exp Clin Endocrinol Diabetes. 2001;109(4):217–26. doi: 10.1055/s-2001-15109 11453034 · doi ↗ · pubmed ↗
- 3Thompson NL, Flanders KC, Smith JM, Ellingsworth LR, Roberts AB, Sporn MB. Expression of transforming growth factor-beta 1 in specific cells and tissues of adult and neonatal mice. J Cell Biol. 1989;108(2):661–9. doi: 10.1083/jcb.108.2.661 2645303 PMC 2115426 · doi ↗ · pubmed ↗
- 4Wu M, Wu S, Chen W, Li Y-P. The roles and regulatory mechanisms of TGF-β and BMP signaling in bone and cartilage development, homeostasis and disease. Cell Res. 2024;34(2):101–23. doi: 10.1038/s 41422-023-00918-9 38267638 PMC 10837209 · doi ↗ · pubmed ↗
- 5Wu M, Chen G, Li Y-P. TGF-β and BMP signaling in osteoblast, skeletal development, and bone formation, homeostasis and disease. Bone Res. 2016;4:16009. doi: 10.1038/boneres.2016.9 27563484 PMC 4985055 · doi ↗ · pubmed ↗
- 6Yasui T, Kadono Y, Nakamura M, Oshima Y, Matsumoto T, Masuda H, et al. Regulation of RANKL-induced osteoclastogenesis by TGF-β through molecular interaction between Smad 3 and Traf 6. J Bone Miner Res. 2011;26(7):1447–56. doi: 10.1002/jbmr.357 21305609 · doi ↗ · pubmed ↗
- 7Tang Y, Wu X, Lei W, Pang L, Wan C, Shi Z, et al. TGF-beta 1-induced migration of bone mesenchymal stem cells couples bone resorption with formation. Nat Med. 2009;15(7):757–65. doi: 10.1038/nm.1979 19584867 PMC 2727637 · doi ↗ · pubmed ↗
- 8Matsunobu T, Torigoe K, Ishikawa M, de Vega S, Kulkarni AB, Iwamoto Y, et al. Critical roles of the TGF-beta type I receptor ALK 5 in perichondrial formation and function, cartilage integrity, and osteoblast differentiation during growth plate development. Dev Biol. 2009;332(2):325–38. doi: 10.1016/j.ydbio.2009.06.002 19501582 PMC 2716725 · doi ↗ · pubmed ↗