Enhancing Effect of Tob Deficiency on Bone Formation Is Specific to Bone Morphogenetic Protein-Induced Osteogenesis


  • The authors have no conflict of interest


Tob is a recently reported novel bone morphogenetic protein (BMP) inhibitor, which originally was identified by West-Western procedure using ErbB2 as a probe and contains a nuclear localization signal. To further characterize the effects of Tob deficiency on BMP-induced new bone (NB) formation, we examined microcomputed tomography (μCT) on the cross-section of the bone induced by daily injection with BMP onto the calvariae of newborn mice. The calvariae of the saline-injected Tob-deficient (TD) mice were similar to those of the saline-injected or untreated wild-type (WT) mice. BMP injection locally produced NB on the calvaria in WT mice as known previously. In contrast to WT mice, BMP injection onto the calvariae of TD mice produced a calcified area in the cross-section of NB, which was more than that produced by BMP in the WT calvariae. In addition, the horizontal width and the vertical height of the NB induced by BMP in TD mice were several-fold more than those in WT mice. The effect of Tob deficiency on bone-forming activity was selective to the response to the injection with BMP because the levels of injury-induced NB formation examined by μCT 10 days after bone marrow ablation in the femora were similar between the TD and WT mice. These data indicate that Tob acts as a novel specific antagonist against bone formation induced by BMP treatment in bone.


TOB IS A MEMBER of the antiproliferative gene family consisting of BTG1, PC3/PIS21/BTG2, ANA/BTG3, and Tob2.(1–6) All of these family members contain a Tob homology domain in their N-terminal ends. The Tob gene encodes a 45-kD protein, which was first discovered by West-Western procedure using ErbB2 as a probe. Antiproliferative action of this gene product was shown by its overexpression in NIH3T3 cells. However, the in vivo activity of this molecule was not clear until the production of Tob knockout mice.(7) Tob knockout mice were born normally and grew similarly to the wild-type (WT) mice and were fertile.(7) Although no apparent morphological changes were observed in their skeleton and other tissues until several months after birth, the bones in the Tob knockout mice were found to be osteosclerotic later in their lives.(7) Analysis of the bones of Tob-deficient (TD) mice revealed high levels of bone volume without any major alteration in osteoclastic parameters including osteoclast surface. Therefore, it was assumed that TD did not affect bone resorption, but rather it enhanced bone formation.(7)

Analysis of Tob actions on one of the most potent bone-forming cytokines, bone morphogenetic protein (BMP), indicated that the presence of Tob could suppress BMP actions on osteoblastic cell differentiation monitored by its enhancing activity on the expression of alkaline phosphatase in calvaria-derived osteoblastic cells.(7) Furthermore, Tob was found to bind to Smad proteins and to suppress their action to mediate BMP signaling in the nuclei.(7)

BMP activities have been known to be suppressed by the presence of extracellular peptide molecules such as noggin, chordin, and follistatin.(8–10) In addition, intracellular inhibitors such as Smad7 and Smad6,(11,12) or recently found Smurf(13) are also involved in suppression of BMP activities via inactivation or ubiquitination of Smad proteins(14) as well as BMP receptors.(15) Tob does not belong to any of these classes of inhibitors and hence a novel type of inhibitory molecule.

In our initial experiments, the effect of Tob deficiency on bone formation in vivo has been shown by injection of BMP onto the calvaria by analyzing the horizontal bone area and its density.(7) However, the relationship between the newly formed bones on the calvaria to the preexisting calvarial bones was not known in that experiment. In addition, specificity of the action of Tob against the BMP-induced bone formation compared with the other type of stimulation of bone formation in vivo was not known either. Such other stimuli would include bone formation after bone injury, which is one of the most well-known bone-forming activities in the adult animals. To see whether Tob action can be observed in both the BMP-induced bone formation and the other types of bone formation in vivo, we compared bone formation response to two stimuli, BMP injection and bone marrow ablation, in Tob knockout mice. The data indicated that although Tob deficiency enhanced BMP injection-induced bone formation, it did not enhance injury-induced bone formation observed after bone marrow ablation, indicating the specificity of Tob action against BMP treatment-induced bone formation.


Newborn Tob knockout mice as well as WT mice were used and BMP injection was conducted according to the method described previously.(16) For BMP injection, newborn Tob knockout mice were divided into two groups, and one group received daily saline injection for 10 days, the other group of mice received 5-μg BMP-2 (human recombinant) injection daily for 10 days and the mice were killed on the 11th day of the experiment. Bone marrow ablation was conducted as described previously.(17,18) Twelve-week-old Tob knockout mice and WT mice were subjected to the bone marrow ablation operation in the right femur, and the left femur was left intact as a control. After the bone marrow ablation, bone volume was evaluated 10 days later.

Histological examination

Bones were first fixed in paraformaldehyde overnight and then were decalcified in EDTA for several days. After dehydration through the graded alcohol series, bones were embedded in paraffin and sections were made and stained for hematoxylin and eosion.

Microcomputed tomography analysis

Calvaria bones and femora were subjected to microcomputed tomography (μCT) analysis. The total area of new bone (NB) formation on top of the calvaria was evaluated by using the tangential sections, and the height and width of the NB formed in response to BMP were evaluated. μCT analysis of the marrow-ablated femora was conducted in the midsaggital sections of the femora, and the newly formed bone observed in the ablated marrow was examined to obtain quantitative values. Quantification of the trabecular bone volume per tissue volume (BV/TV) in the femora of WT and TD mice after bone marrow ablation was conducted within bone marrow in a rectangular region (0.7 mm × 2.1 mm, total of 1.47 mm2) at 0.2 mm away from the growth plate. This region included most of the area of bone marrow ablation where the preexisting trabecular bones were removed.

Bone marrow ablation

The right femora of the mice was used for the bone marrow ablation as described previously. First, a drill hole was made in the intercondylar region of the knee, and then a Kirschner wire with a diameter of 0.6 mm was inserted into the bone marrow retrogradely from the distal end of the femur toward the proximal end of the femur to make sure that bone marrow was ablated totally. X-ray pictures were taken with the inserted Kirschner wires.(17) No bone destruction except for the bone marrow removal was observed based on this X-ray. Kirschner wires were removed and the wound was closed. Animals were allowed to move freely after the operation.

Bone marrow cell culture

Bone marrow cells were obtained from the femora of 12-week-old WT or TD mice. The cells were cultured for 3 days in α-minimum essential medium (α-MEM) supplemented with 5% fetal bovine serum and then replated into 96-well plates and subsequently treated with BMP at 500 ng/ml or vehicle for 3 days. At the end of the cultures, the cells were subjected to 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide (MTT) assay(19) or to alkaline phosphatase assay.(20)

Statistical analysis

Data were expressed as mean ± SE. Significance of the difference was evaluated by using the Fisher's test.


We examined the newly formed bone in the tangential cross-section through the frontal plane in the calvaria. As shown in Fig. 1A, the calvaria injected with saline did not show any major response in the frontal section of the μCT picture. BMP injection resulted in the formation of a dome-like NB on top of the calvaria (Fig. 1B). In the case of Tob knockout mice (Fig. 1C), the thickness of the calvaria itself was similar to that in WT mice (Fig. 1A), and saline injection per se did not induce any major response on the calvaria (Fig. 1C). In contrast to WT mice, Tob knockout mice responded to BMP injection by forming a larger domelike bone mass (Fig. 1D) on top of the calvaria than that in BMP-injected WT mice (Fig. 1B).

Figure FIG. 1..

Frontal section of the calvaria after treatment with BMP. (A and B) Newborn WT or (C and D) TD mice were injected daily with (A and C) saline or (B and D) 5 μg of BMP-2 for 10 days and were killed on day 11. For saline injection experiments, six WT and seven TD mice were used, and for BMP injection experiments five WT and five TD mice were used. Calvariae were subjected to μCT analyses in the frontal section. Arrow head indicates the presence of a thin noncalcified layer separating the newly formed bone from the calvaria.

Examination of the interface between such domelike NB and calvaria in WT mice revealed that the newly formed bone was separated by a thin uncalcified layer from the calvaria bone (Fig. 1B, arrow head). This separation was observed also in the case of Tob knockout mice (Fig. 1D, arrow head). Therefore, the cells responding to BMP treatment in Tob knockout mice and WT mice were mainly those in the periosteal cells sitting close to but not directly touching on the top of calvaria bone.

To evaluate the vertical size of the calcified area in the dome-like NBs in WT and Tob knockout mice formed in response to the BMP treatment, we measured the total area of calcified mass in the frontal tangential cross-section on the μCT pictures using an image analyzer. The data indicated that BMP increased the cross-sectional calcified mass in the newly formed bone in WT mice as shown in Fig. 2A (second column from the left). Again, in Tob knockout mice, BMP treatment increased the calcified mass in the cross-section of the newly formed bone over fivefold more (Fig. 2A, right-most column) than that induced by BMP in WT mice (p < 0.05). The analysis of the width of the base of the NBs also showed twofold more enhancement in BMP-induced NB in Tob knockout mice compared with WT (Fig. 2B; p < 0.05). In addition, the height of the BMP-induced dome-shaped NB was threefold more in Tob knockout mice compared with that of the BMP-induced bone in WT mice (Fig. 2C; p < 0.05). These observations on the geometrical parameters of the BMP-induced NB formation indicated that BMP-induced formation of the three-dimensional calcified mass was enhanced in the absence of Tob.

Figure FIG. 2..

Quantification of the newly formed bone on the calvaria after treatment with BMP. Quantification of the (A) area, (B) base width, and (C) height of the newly formed bone on the calvariae induced by BMP treatment was conducted by measuring the calcified area shown in Fig. 1. *Statistically significant difference (p < 0.05).

Histology of the newly formed bone indicated the presence of woven bone being formed by the osteoblastic cells in both WT and knockout mice. The shape of osteoblastic cells located on top of the newly formed woven bone matrix basically was similar between WT and Tob knockout mice. The gross morphology of the BMP-induced woven bone also was similar between WT and Tob knockout mice (Figs. 3A and 3B), suggesting that the accelerated bone formation in the Tob knockout mice in response to BMP was caused by the increase in production of bone matrix that was histologically normal in quality.

Figure FIG. 3..

Histology of the newly formed bone on the calvariae. New born (A) WT or (B) TD (TOBKO) mice were injected daily with 5 μg of BMP-2 for 10 days and were killed on day 11. Calvariae were subjected to histological analyses after Hematoxilin and eosin staining.

Body weights of the WT mice were increasing from 2 to 7 g during the 10-day period of saline injection, and this increase in the body weight was similar to that in the BMP-injected WT mice. In the case of Tob knockout mice, the body weight again was increasing similarly from 2 to 7 g during the 10-day period regardless of the injection of saline or BMP (data not shown). Therefore, although a relatively high amount of BMP was injected into the newborn mice, it did not affect the growth of the animals regardless of the genotype. Thus, Tob deficiency-induced increase in the NB formation was not the result of the systemic effects of BMP on the size or weight of the animals, but that of the local effects of BMP on the sites of injection.

Based on the establishment of the enhanced bone formation in response to BMP in Tob knockout mice, we further asked whether the response to such bone formation stimuli was observed not only in the case of the BMP injection but also in the case of any other stimuli to enhance bone formation in Tob knockout mice. To answer this question, we adopted a bone marrow ablation system. It is known that in the bone marrow ablation system, bone formation can occur within 10 days after ablation, and the formation of the bone can be quantified by using μCT analysis.(18) Compared with the control side intact bone marrow (no ablation) in WT (Fig. 4A), Tob knockout mice (Fig. 4C) indicated slightly more trabecular bones in the control-side intact bone marrow (no ablation) as reported previously. In WT mice, bone marrow ablation resulted in formation of new cancellous bones, which were filling the bone marrow cavity ablated by the insertion of the Kirschner wire as shown in Fig. 4B. In Tob knockout mice, bone marrow ablation also resulted in the formation of the cancellous bone as shown in Fig. 4D.

Figure FIG. 4..

Bone marrow ablation induced bone formation in the femora. Bone marrow was left (A and C) intact or (B and D) ablated in (A and B) WT or (C and D) TD mice. Ablation was conducted in the bone marrow (arrows) as described in the Materials and Methods section using 12-week-old WT or TD mice. We used six WT mice and six TD mice. μCT analysis was conduced in the femora after 10 days. (E) Quantification of the fractional bone volume was conducted by using the μCT-based pictures shown in panels A-D. *Difference between BV/TV in intact femora of WT and that in intact femora of TD mice is statistically significant (p < 0.05).

Quantification of the bone volume of the cancellous bones in the intact (no ablation) femora indicated that statistically higher levels in the intact-control femora of Tob knockout mice (BV/TV, ∼15%) compared with the intact-control femora of WT mice (BV/TV, ∼10%; Fig. 4E; p < 0.05) as reported previously. Because there was virtually no cancellous bone just after bone marrow ablation in both WT and TD mice, the balance between the levels at the starting point and the final levels of bone volume indicate the net NB formation without any influence of the difference in the preexisting basal bone volume. The volume of net NB formed in response to bone marrow ablation was ∼30% in WT mice and a similar level of newly formed bone was observed in Tob knockout mice as well (Fig. 4E). These data indicated that the enhancement of the NB formation in the absence of Tob is specific to the BMP-injection experiments and not a general phenomenon as shown by the similar levels of injury-induced NB formation in response to bone marrow ablation.

Because calvaria-derived bone cells from TD mice were reported to respond more to BMP treatment than those cells from WT mice(7) and bone marrow ablation was reported to induce BMP expression,(18) the similar levels of the bone formation after bone marrow ablation between the mice with two genotypes were unexpected. We postulated that bone marrow osteogenic cells in TD mice still could respond more to BMP than those cells in WT mice if they were not in an injury situation in which numerous cytokines are present and activated at the same time in the cases such as bone marrow ablation. To test this hypothesis, we further investigated responses to BMP of bone marrow cells obtained from WT and TD mice. We examined the effects of BMP on cell proliferation estimated by MTT assay and also the effects of BMP on alkaline phosphatase expression in these cells. As shown in Fig. 5A, BMP-2 enhancement of the MTT levels in TD bone marrow cells was more than that in WT bone marrow cells. BMP enhancement of alkaline phosphatase expression in TD bone marrow cells also was more than that in the WT bone marrow cells (Fig. 5B). These data are similar to those observed in terms of the BMP response profile in calvaria cells obtained from TD mice, which we have described previously.(7)

Figure FIG. 5..

BMP effects on proliferation and alkaline phosphatase expression in bone marrow derived cells. Bone marrow cells were obtained as described in the Materials and Methods section. The cells were treated with 500 ng/ml BMP for 3 days. (A) MTT assay or (B) alkaline phosphatase assay was conducted in WT or TD cells. Each group consists of six wells. *The difference is statistically significant (p < 0.05).


In this study, we asked whether the enhanced bone formation observed in TD mice in response to external stimuli is specific to BMP treatment. Our data indicated that Tob deficiency enhanced BMP-injection-induced NB formation in the calvariae in vivo and the levels of NB formation induced by bone marrow ablation in TD mice were similar to those observed in WT mice, suggesting that enhanced bone formation observed in the case of TD mice treated with BMP is specific to the stimuli elicited by the application of this single cytokine.

Our analysis on the newly formed bone in TD mice in response to BMP injection indicated that the quality of the bone per se appears to be similar between the WT and Tob knockout mice, suggesting that measures to suppress Tob activity may facilitate to increase bone volume without affecting the quality of the bone, and this avenue would be important to contemplate possible application of the finding on Tob to design treatment strategies for curing osteopenic diseases such as osteoporosis.

μCT analysis in the frontal tangential plane and histological examination indicated that regardless of the genotype, newly formed bone on top of the calvaria did not have a direct connection to the calvaria bone because they were separated by a thin uncalcified layer. This is interesting in that BMP-responsive cell population could be in a subset of cells in the periostea, which could give rise to distinct a domelike bone separated from the calvaria at its base. Similarity in terms of the presence of this separation between BMP-treated WT and BMP-treated Tob-knockout mice further suggests that a certain subset of cells targeted by BMP could be influenced by the absence of Tob. In addition, preexisting calvaria bone underneath the locally injected BMP did not show any major difference even after 10 days of daily injection regardless of the presence or absence of Tob. This indicated that bone cells inside the calvarial plates or very close to the outer surface and inner surface of the calvaria did not respond to BMP even when the topical application was repeated at the very proximity to these cells and Tob did not affect these cells that did not respond to BMP. It is not known whether the expression of Tob could be different between the subset of osteoblastic cells in the inner periosteal layer (directly contacting calvarial bone surface) or the outer periosteal layer and/or the cells inside the calvaria. The presence of several Tob-related genes may be responsible in a certain subset of cells to compensate the absence of Tob. In fact, Tob is known to be expressed in osteoblasts and Tob2 is expressed reciprocally in osteoclasts.(7)

Stimuli to periosteum by the local injection of BMP without any additional cytokines would be different from those induced by marrow ablation where injury-dependent activation of the repair process involving many local factors such as platelet-derived growth factor (PDGF) and transforming growth factor (TGF) β would be dominant. In such a case, even when coexpression of BMP might exist, its contribution to the bone formation during the whole repair process would be interfered significantly by the presence of many other cytokines and growth factors. Therefore, Tob deficiency would not be as potently influential on the bone formation after bone marrow ablation as its effects in the cases of simple BMP-injection experiments. Based on this hypothesis, we conducted in vitro experiments and showed that TD enhanced the responses of the bone marrow cells to BMP treatment when no other additional cytokines were present. Our data indicated that bone formation in response to BMP is affected by the absence of Tob, and the bone formation in response to injury by bone marrow ablation was not. These observations are compatible with those that TGF-β as well as fibroblast growth factor (FGF) signalings were not affected by the deficiency of Tob as reported in our previous paper.(7) Because the two in vivo bone formation systems we used are different in terms of the ages of the mice, we do not exclude the possibility that the age difference also may influence the different responses in BMP injection and bone marrow ablation.

The BV/TV levels after bone marrow ablation were statistically similar between WT and TD mice. This quantification was conducted within the area where bone marrow was mostly ablated. Therefore, at the start point (just after the ablation), there was virtually no bone (i.e., no difference in BV/TV) in the ablated bone marrows of WT and TD mice because cancellous bone was removed mostly in the area of BV/TV measurement. Thus, the net bone formation levels after marrow ablation, which is the balance between the starting point levels and ending point levels, could be interpreted to be similar regardless of the genotypes of the mice.

Overall, our data indicated that the signaling in bone in vivo elicited by the treatment with BMP was enhanced specifically by the absence of Tob. Thus, Tob is an important inhibitory regulator of bone formation and its presence should be taken into consideration when one contemplates to promote bone formation by using the osteogenic activity of BMP in vivo.


This research was supported by the grants-in-aid received from the Japanese Ministry of Education (12557123, 13045011, 13216034, 01142, and 0930734) grants from Core Research for Evolutional Science and Technology (CREST) of Japan Science and Technology Corporation (JST), a grant from the “Research for the Future” Program of the Japan Society for the Promotion of Science (JSPS; 96100205), and grants from National Space Development Agency of Japan (NASDA) and Tokyo Biochemistry Research Foundation.