Bcl2l1 Deficiency in Osteoblasts Reduces the Trabecular Bone Due to Enhanced Osteoclastogenesis Likely through Osteoblast Apoptosis

Bcl2l1 (Bcl-XL) belongs to the Bcl-2 family, Bcl2 and Bcl2-XL are major anti-apoptotic proteins, and the apoptosis of osteoblasts is a key event for bone homeostasis. As the functions of Bcl2l1 in osteoblasts and bone homeostasis remain unclear, we generated osteoblast-specific Bcl2l1-deficient (Bcl2l1fl/flCre) mice using 2.3-kb Col1a1 Cre. Trabecular bone volume and the trabecular number were lower in Bcl2l1fl/flCre mice of both sexes than in Bcl2l1fl/fl mice. In bone histomorphometric analysis, osteoclast parameters were increased in Bcl2l1fl/flCre mice, whereas osteoblast parameters and the bone formation rate were similar to those in Bcl2l1fl/fl mice. TUNEL-positive osteoblastic cells and serum TRAP5b levels were increased in Bcl2l1fl/flCre mice. The deletion of Bcl2l1 in osteoblasts induced Tnfsf11 expression, whereas the overexpression of Bcl-XL had no effect. In a co-culture of Bcl2l1-deficient primary osteoblasts and wild-type bone-marrow-derived monocyte/macrophage lineage cells, the numbers of multinucleated TRAP-positive cells and resorption pits increased. Furthermore, serum deprivation or the deletion of Bcl2l1 in primary osteoblasts increased apoptosis and ATP levels in the medium. Therefore, the reduction in trabecular bone in Bcl2l1fl/flCre mice may be due to enhanced bone resorption through osteoblast apoptosis and the release of ATP from apoptotic osteoblasts, and Bcl2l1 may inhibit bone resorption by preventing osteoblast apoptosis.


Introduction
Osteoblast apoptosis plays an important role in bone development and maintenance.Between 60 and 80% of osteoblasts that originally assemble at the resorption pit are estimated to die by apoptosis.Furthermore, bone loss caused by sex steroid deficiency, glucocorticoid excess, or aging is caused in part by osteoblast apoptosis, and parathyroid hormone (PTH), bisphosphonate, and calcitonin exert anabolic effects on bone by inhibiting osteoblast apoptosis [1][2][3][4][5][6][7][8].The intrinsic pathway, also called the BCL-2-regulated or mitochondrial pathway, is activated by various developmental cues or cytotoxic insults, such as viral infection, DNA damage, and growth-factor deprivation.This pathway is strictly controlled by the BCL-2 family of proteins and predominantly leads to the activation of caspase-9 [9].Bcl2 subfamily proteins, including Bcl2 and Bcl-XL, inhibit apoptosis by preventing the release of caspase activators from mitochondria through the inhibition of Bax subfamily proteins [10].
Five isoforms of Bcl2l1 (Bcl-XL, Bcl-XS, Bcl-Xγ, Bcl-Xβ, and Bcl-X TM) have been identified to date, and Bcl-XL is the major isoform of Bcl2l1 [15][16][17][18][19]. Bcl2l1 −/− mice die on embryonic day (E) 13.5 and exhibit severe defects in erythropoiesis and neuronal development [20].The osteoclast-specific deletion of Bcl2l1 using Cre knock-in mice in the cathepsin K locus did not affect osteoclast numbers, but increased bone resorption and resulted in osteopenia.Bcl2l1 in osteoclasts is suggested to attenuate osteoclastic boneresorbing activity through a decrease in the production of extracellular matrix proteins, such as vitronectin and fibronectin [21].Therefore, Bcl2 and Bcl2l1 have distinct functions in osteoclasts.
We recently showed that Bcl2 deficiency up-regulated p53 in osteoblasts, which induced Pten and Igfbp3, inactivated Akt, leading to the activation of FoxO1 and FoxO3a, and accelerated osteoblast differentiation [13].The overexpression of Bcl2 in osteoblasts impaired osteocyte process formation, leading to a reduction in the number of canaliculi and osteocyte apoptosis.In contrast, the overexpression of Bcl-XL in osteoblasts increased the volumes of both the trabecular and cortical bone with normal structures and these increases were maintained during aging.Therefore, the functions of Bcl2 and Bcl-XL in osteoblasts may also differ.
Apoptotic cells have been shown to release adenosine triphosphate (ATP) through pannexin 1 (Panx1) as a 'find me' signal in the earliest stage of death to recruit phagocytes [22].ATP binds to P2Y1 receptors on osteoblasts and enhances Tnfsf11 expression [23].ATP also activates NF-κB signaling in osteoclasts through P2X7 receptors on osteoclasts, and ATP drives osteoclast fusion through P2X7 receptors [24,25].Therefore, the activation of P2Y and P2X receptors by ATP leads to enhanced bone resorption [26].
We herein demonstrated that osteoblast numbers and bone formation were maintained in osteoblast-specific Bcl2l1-deficient mice, whereas osteoclastogenesis was accelerated through the induction of Tnfsf11 expression in osteoblasts likely through the release of ATP from apoptotic osteoblasts, resulting in a reduction in trabecular bone.

Generation of Osteoblast-Specific Bcl2l1 Conditional Knock-Out Mice
To examine the expression levels of Bcl2l1 in various tissues, Bcl2l1 expression was examined using RNA from bone marrow, femur, brain, lung, heart, liver, stomach, kidney, and spleen of 10-week-old mice.Bcl2l1 expression in the femur was moderate among these tissues (Figure 1A).To specifically delete Bcl2l1 in osteoblasts, we confirmed Cre expression in 2.3-kb Col1a1 promoter Cre transgenic mice.The 2.3-kb Col1a1 promoter Cre transgenic mice were mated with β-galactosidase reporter mice under the control of the CAG promoter.In double transgenic mice, β-galactosidase activity was specifically detected in osteoblasts (Figure 1B,C).Bcl2l1 fl/+ mice were crossed with Cre transgenic mice and Bcl2l1 fl/flCre mice were generated.Lower Bcl2l1 protein levels in Bcl2l1 fl/flCre femurs than in Bcl2l1 fl/fl femurs were confirmed by Western blotting (Figure 1D,E).Body weights in Bcl2l1 fl/flCre mice were similar to those in Bcl2l1 fl/fl mice of both sexes (Figure 1F).and Bcl2l1 fl/flCre mice were generated.Lower Bcl2l1 protein levels in Bcl2l1 fl/flCre femurs than in Bcl2l1 fl/fl femurs were confirmed by Western blotting (Figure 1D,E).Body weights in Bcl2l1 fl/flCre mice were similar to those in Bcl2l1 fl/fl mice of both sexes (Figure 1F).

Reduction in Bone Mass in Bcl2l1 fl/flCre Mice
In the µCT analysis, the trabecular bone volume and the trabecular number in the femur were lower in Bcl2l1 fl/flCre mice than in Bcl2l1 fl/fl mice of both sexes (Figure 2A,B,E).Trabecular thickness and trabecular bone mineral density (BMD) were lower in male Bcl2l1 fl/flCre mice than in male Bcl2l1 fl/fl mice (Figure 2E).At the mid-diaphysis of the femoral cortical bone, cortical thickness, endosteal perimeters, and cortical BMD in Bcl2l1 fl/flCre mice were similar to those in Bcl2l1 fl/fl mice in both sexes, while the periosteal perimeter was shorter in Bcl2l1 fl/flCre mice than in Bcl2l1 fl/fl mice among males, but not females (Figure 2C,D,F).

Reduction in Bone Mass in Bcl2l1 fl/flCre Mice
In the µCT analysis, the trabecular bone volume and the trabecular number in the femur were lower in Bcl2l1 fl/flCre mice than in Bcl2l1 fl/fl mice of both sexes (Figure 2A,B,E).Trabecular thickness and trabecular bone mineral density (BMD) were lower in male Bcl2l1 fl/flCre mice than in male Bcl2l1 fl/fl mice (Figure 2E).At the mid-diaphysis of the femoral cortical bone, cortical thickness, endosteal perimeters, and cortical BMD in Bcl2l1 fl/flCre mice were similar to those in Bcl2l1 fl/fl mice in both sexes, while the periosteal perimeter was shorter in Bcl2l1 fl/flCre mice than in Bcl2l1 fl/fl mice among males, but not females (Figure 2C,D,F).The bone histomorphometric analysis was performed using male femurs at 10 weeks of age.In the trabecular bone, osteoblast parameters, including the osteoid surface, osteoid thickness, osteoblast surface, and osteoblast number, and the parameters for bone formation, including the mineral apposition rate, mineralizing surface, and bone formation rate, were similar between Bcl2l1 fl/flCre and Bcl2l1 fl/fl mice, whereas the osteoclast parameters, including the osteoclast surface and eroded surface, were increased in Bcl2l1 fl/flCre mice (Figure 3A).In the cortical bone, the mineralizing surface in the periosteum, but not in the endosteum, was slightly larger in Bcl2l1 fl/flCre mice than in Bcl2l1 fl/fl mice, while the mineral apposition rate and bone formation rate in both the periosteum and endosteum were similar in Bcl2l1 fl/flCre mice and Bcl2l1 fl/fl mice (Figure 3B,C).The bone histomorphometric analysis was performed using male femurs at 10 weeks of age.In the trabecular bone, osteoblast parameters, including the osteoid surface, osteoid thickness, osteoblast surface, and osteoblast number, and the parameters for bone for- rate, were similar between Bcl2l1 fl/flCre and Bcl2l1 fl/fl mice, whereas the osteoclast parameters, including the osteoclast surface and eroded surface, were increased in Bcl2l1 fl/flCre mice (Figure 3A).In the cortical bone, the mineralizing surface in the periosteum, but not in the endosteum, was slightly larger in Bcl2l1 fl/flCre mice than in Bcl2l1 fl/fl mice, while the mineral apposition rate and bone formation rate in both the periosteum and endosteum were similar in Bcl2l1 fl/flCre mice and Bcl2l1 fl/fl mice (Figure 3B,C).

Increased Osteoblast Apoptosis and Serum TRAP5b Levels in Bcl2l1 fl/flCre Mice
Since the number of osteoblasts was not decreased in Bcl2l1 fl/flCre mice, the proliferation of osteoblasts was examined by BrdU labeling.The frequencies of BrdU-positive cells were similar between Bcl2l1 fl/flCre and Bcl2l1 fl/fl mice (Figure 4A,B).We then examined osteoblast apoptosis by TUNEL staining.The frequency of TUNEL-positive osteoblastic cells was higher in Bcl2l1 fl/flCre mice than in Bcl2l1 fl/fl mice (Figure 4C,D).The level of the serum marker for bone formation, osteocalcin, was similar in Bcl2l1 fl/flCre mice and Bcl2l1 fl/fl mice, whereas that of the serum marker for bone resorption, tartrate-resistant acid phosphatase 5b (TRAP5b), was higher in Bcl2l1 fl/flCre mice than in Bcl2l1 fl/fl mice (Figure 4E,F).

Increased Osteoblast Apoptosis and Serum TRAP5b
Levels in Bcl2l1 fl/flCre Mice Since the number of osteoblasts was not decreased in Bcl2l1 fl/flCre mice, the proliferation of osteoblasts was examined by BrdU labeling.The frequencies of BrdU-positive cells were similar between Bcl2l1 fl/flCre and Bcl2l1 fl/fl mice (Figure 4A,B).We then examined osteoblast apoptosis by TUNEL staining.The frequency of TUNEL-positive osteoblastic cells was higher in Bcl2l1 fl/flCre mice than in Bcl2l1 fl/fl mice (Figure 4C,D).The level of the serum marker for bone formation, osteocalcin, was similar in Bcl2l1 fl/flCre mice and Bcl2l1 fl/fl mice, whereas that of the serum marker for bone resorption, tartrate-resistant acid phosphatase 5b (TRAP5b), was higher in Bcl2l1 fl/flCre mice than in Bcl2l1 fl/fl mice (Figure 4E,F).

Osteoblastogenesis and Osteoclastogenesis In Vitro
Primary osteoblasts from Bcl2l1 fl/fl mice were infected with the green fluorescent protein (GFP)-or Cre-expressing adenovirus.Osteoblast differentiation was evaluated by alkaline phosphatase (ALP) and von Kossa staining.Staining was similar between cells infected with the GFP-and Cre-expressing adenoviruses, indicating that the deletion of Bcl2l1 did not affect osteoblast differentiation in the early or late stage (Figure 5A-D).Since osteoclast parameters and serum TRAP5b levels were increased in Bcl2l1 fl/flCre mice, Tnfsf11 and Tnfrsf11b expression levels were examined by real-time RT-PCR.The expression levels of Tnfsf11, but not Tnfrsf11b, were increased and the Tnfsf11/Tnfrsf11b ratio was higher with the Cre-expressing adenovirus than with the GFP-expressing adenovirus

Osteoblastogenesis and Osteoclastogenesis In Vitro
Primary osteoblasts from Bcl2l1 fl/fl mice were infected with the green fluorescent protein (GFP)-or Cre-expressing adenovirus.Osteoblast differentiation was evaluated by alkaline phosphatase (ALP) and von Kossa staining.Staining was similar between cells infected with the GFP-and Cre-expressing adenoviruses, indicating that the deletion of Bcl2l1 did not affect osteoblast differentiation in the early or late stage (Figure 5A-D).Since osteoclast parameters and serum TRAP5b levels were increased in Bcl2l1 fl/flCre mice, Tnfsf11 and Tnfrsf11b expression levels were examined by real-time RT-PCR.The expression levels of Tnfsf11, but not Tnfrsf11b, were increased and the Tnfsf11/Tnfrsf11b ratio was higher with the Cre-expressing adenovirus than with the GFP-expressing adenovirus (Figure 5E).The expression of Tnfsf11 and Tnfrsf11b was also examined in Bcl-XL transgenic (tg) mice under the control of the 2.3-kb Col1a1 promoter [27] using the newborn calvaria and osteoblast fraction from femurs and tibiae at 6 weeks of age.Although Tnfrsf11b expression levels in the calvaria were slightly higher in Bcl-XL tg mice than in wild-type mice, Tnfsf11 expression levels and the Tnfsf11/Tnfrsf11b ratio were similar between wild-type and Bcl-XL tg mice in both the newborn calvaria and osteoblast fraction from femurs and tibiae (Figure 5F,G).Similar results were also obtained for primary osteoblasts from Bcl-XL tg mice (Figure 5H).
(Figure 5E).The expression of Tnfsf11 and Tnfrsf11b was also examined in Bcl-XL t genic (tg) mice under the control of the 2.3-kb Col1a1 promoter [27] using the new calvaria and osteoblast fraction from femurs and tibiae at 6 weeks of age.Alth Tnfrsf11b expression levels in the calvaria were slightly higher in Bcl-XL tg mice th wild-type mice, Tnfsf11 expression levels and the Tnfsf11/Tnfrsf11b ratio were simila tween wild-type and Bcl-XL tg mice in both the newborn calvaria and osteoblast fra from femurs and tibiae (Figure 5F,G).Similar results were also obtained for primary oblasts from Bcl-XL tg mice (Figure 5H).Bcl2l1 fl/fl primary osteoblasts, which were infected with the GFP-or Cre-expressing adenovirus, were co-cultured with bone-marrow-derived monocyte/macrophage lineage cells (BMMs) from Bcl2l1 fl/fl mice.Multinucleated TRAP-positive cells were higher in the co-culture using primary osteoblasts infected with the Cre-expressing adenovirus than in that with the GFP-expressing adenovirus (Figure 6A,B).The pit assay was performed on the co-culture of primary osteoblasts from Bcl2l1 fl/fl or Bcl2l1 fl/flCre mice with BMMs from Bcl2l1 fl/fl mice.The area of resorption pits in the co-culture using Bcl2l1 fl/flCre primary osteoblasts was larger than in that using Bcl2l1 fl/fl primary osteoblasts (Figure 6C,D).
Int. J. Mol.Sci.2023, 24, x FOR PEER REVIEW 8 of 15 Bcl2l1 fl/fl primary osteoblasts, which were infected with the GFP-or Cre-expressing adenovirus, were co-cultured with bone-marrow-derived monocyte/macrophage lineage cells (BMMs) from Bcl2l1 fl/fl mice.Multinucleated TRAP-positive cells were higher in the co-culture using primary osteoblasts infected with the Cre-expressing adenovirus than in that with the GFP-expressing adenovirus (Figure 6A,B).The pit assay was performed on the co-culture of primary osteoblasts from Bcl2l1 fl/fl or Bcl2l1 fl/flCre mice with BMMs from Bcl2l1 fl/fl mice.The area of resorption pits in the co-culture using Bcl2l1 fl/flCre primary osteoblasts was larger than in that using Bcl2l1 fl/fl primary osteoblasts (Figure 6C,D).

Increases in Apoptosis and ATP Release by the Bcl2l1 Deletion
To confirm that ATP was released in the medium by osteoblast apoptosis, primary osteoblasts from wild-type mice were cultured in the presence or absence of serum for 24 h.In the absence of serum, TUNEL-positive cells were increased and ATP release in the medium, which was measured by luciferase, was also enhanced (Figure 7A-C).Elevations were observed in Tnfsf11 expression levels and the Tnfsf11/Tnfrsf11b ratio in the absence of serum (Figure 7D).Moreover, Tnfsf11 expression levels and the Tnfsf11/Tnfrsf11b ratio were increased after the culture for 24 h in serum-free conditioned medium supplemented with 10% FBS, indicating that released factors were responsible for the induction of Tnfsf11 expression (Figure 7E).
The number of TUNEL-positive cells in Bcl2l1 fl/fl primary osteoblasts infected with the Cre-expressing adenovirus was higher than in those infected with the GFP-expressing adenovirus (Figure 7F,G).ATP levels in the medium of Bcl2l1 fl/fl primary osteoblasts infected with the Cre-expressing adenovirus were higher than in that of those infected with the GFP-expressing adenovirus, and this increase was eliminated by the addition of apyrase, which catalyzes the hydrolysis of ATP (Figure 7H).

Increases in Apoptosis and ATP Release by the Bcl2l1 Deletion
To confirm that ATP was released in the medium by osteoblast apoptosis, primary osteoblasts from wild-type mice were cultured in the presence or absence of serum for 24 h.In the absence of serum, TUNEL-positive cells were increased and ATP release in the medium, which was measured by luciferase, was also enhanced (Figure 7A-C).Elevations were observed in Tnfsf11 expression levels and the Tnfsf11/Tnfrsf11b ratio in the absence of serum (Figure 7D).Moreover, Tnfsf11 expression levels and the Tnfsf11/Tnfrsf11b ratio were increased after the culture for 24 h in serum-free conditioned medium supplemented with 10% FBS, indicating that released factors were responsible for the induction of Tnfsf11 expression (Figure 7E).
The number of TUNEL-positive cells in Bcl2l1 fl/fl primary osteoblasts infected with the Cre-expressing adenovirus was higher than in those infected with the GFP-expressing adenovirus (Figure 7F,G).ATP levels in the medium of Bcl2l1 fl/fl primary osteoblasts infected with the Cre-expressing adenovirus were higher than in that of those infected with the GFP-expressing adenovirus, and this increase was eliminated by the addition of apyrase, which catalyzes the hydrolysis of ATP (Figure 7H).(A-C) Primary osteoblasts from wildtype mice were cultured in the presence or absence of 10% FBS for 24 h and then stained for TUNEL (A).The frequencies of TUNEL-positive cells were calculated (B).The amount of ATP in the culture medium was quantitated by luciferase (C).n = 8. (D,E) Real-time RT-PCR analysis of Tnfsf11 and Tnfrsf11b.Primary osteoblasts were cultured in the presence or absence of 10% FBS for 24 h (D).The supernatant with 10% FBS was used for the culture of primary osteoblasts (control), and the supernatant without 10% FBS was used as conditioned medium in the culture of primary osteoblasts after the addition of FBS to 10% (conditioned medium + 10% FBS) (E).n = 5. (F-H) Primary osteoblasts from Bcl2l1 fl/fl mice were infected with the GFP-or Cre-expressing adenovirus.After 48 h, cells were stained for TUNEL (F), and the frequencies of TUNEL-positive cells were calculated (G).The amount of ATP in the culture medium with or without apyrase was quantitated (H).Trabecular bone volume was reduced in Bcl2l1 fl/flCre mice due to increased bone resorption.The deletion of Bcl2l1 in osteoblasts did not affect osteoblast differentiation, but enhanced osteoclastogenesis through the induction of Tnfsf11 expression in osteoblasts.However, Bcl2l1 did not appear to exert intrinsic effects on the expression of Tnfsf11 because the overexpression of Bcl-XL in osteoblasts did not affect Tnfsf11 expression.Osteoblast apoptosis was increased in Bcl2l1 fl/flCre mice, and the deletion of Bcl2l1 in vitro increased osteoblast apoptosis and ATP, which is released from apoptotic osteoblasts and induces Tnfsf11 expression and osteoclastogenesis [28], in the medium.These results suggest that the Bcl2l1 deletion in osteoblasts increased apoptosis, apoptotic osteoblasts released ATP, and ATP enhanced osteoclastogenesis, which promoted bone resorption.
Although osteoblast apoptosis was increased in Bcl2l1 fl/flCre mice, osteoblast parameters and the bone formation rate in the bone histomorphometric analysis and serum osteocalcin levels were not reduced.In contrast, osteoclast parameters in the bone histomorphometric analysis and serum TRAP5b levels were increased in Bcl2l1 fl/flCre mice compared with Bcl2l1 fl/fl mice.Therefore, the reduction in trabecular bone volume in Bcl2l1 fl/flCre mice was due to enhanced bone resorption.The deletion of Bcl2l1 in primary osteoblasts increased the expression of Tnfsf11, whereas the overexpression of Bcl-XL had no effect.Moreover, bone resorption in osteoblast-specific Bcl-XL tg mice was similar to that in wild-type mice [27].Therefore, increased bone resorption in Bcl2l1 fl/flCre mice and enhanced osteoclastogenesis and Tnfsf11 induction by the Bcl2l1 deletion in osteoblasts were not considered to have been caused by the intrinsic effects of Bcl2l1.Since the Bcl2l1 deletion in osteoblasts increased apoptosis and ATP release into the medium, the deletion of Bcl2l1 was considered to have enhanced osteoclastogenesis through the release of ATP from apoptotic osteoblasts.Apoptotic osteoblasts are quickly removed by phagocytes; therefore, it is difficult to observe the induction of Tnfsf11 in neighboring osteoblasts in vivo.In the case of apoptotic osteocytes, which are not phagocytosed, the induction of Tnfsf11 was observed in neighboring osteocytes and osteoblasts [29,30].Therefore, ATP released from apoptotic osteoblasts may have induced Tnfsf11 expression by binding to P2Y receptors on neighboring osteoblasts in vivo [23].ATP will also bind to P2X receptors on neighboring osteoclast progenitors and osteoclasts and enhance their differentiation and activity [24,25,31].
Since the increase in osteoblast apoptosis did not affect osteoblast parameters or the bone formation rate in the bone morphometric analysis or the serum level of osteocalcin in Bcl2l1 fl/flCre mice, some mechanisms may have compensated for osteoblast death.P2Y receptors has been shown to positively or negatively function in the proliferation, differentiation, and functions of osteoblasts [32].A previous study demonstrated that ATP potentiated parathyroid hormone-mediated signaling though P2Y1 receptors, ATP released by low-intensity ultrasound induced osteoblastogenesis through P2Y1 receptors, and an antagonist of P2Y12 receptors, which was selective for ADP and ATP, inhibited osteoblast proliferation and differentiation and reduced trabecular bone volume [33][34][35].Therefore, ATP released from apoptotic osteoblasts may have enhanced osteoblastogenesis and compensated for the loss of osteoblasts.
In conclusion, Bcl2l1 was involved in the survival of osteoblasts, the Bcl2l1 deletion in osteoblasts appeared to enhance bone resorption by inducing osteoblast apoptosis, and ATP released from apoptotic osteoblasts may be responsible for enhanced bone resorption.This is the first study to suggest that osteoblast apoptosis reduces bone mass by enhancing bone resorption, although the mechanism shown here by in vitro studies needs to be confirmed in vivo.Since the overexpression of Bcl-XL in osteoblasts has been shown to enhance bone formation and increase the trabecular and cortical bone by preventing osteoblast apoptosis [27], the inhibition of osteoblast apoptosis is important not only for enhancing bone formation, but also for suppressing bone resorption.

Mice
Cre transgenic mice driven by the 2.3-kb Col1a1 promoter, CAG-CAT-LacZ transgenic mice [36], and Bcl2l1 flox mice [37] were kindly provided by Dr. H. Akiyama (Gifu University, Gifu, Japan), Dr. J. Miyazaki (Osaka University, Suita, Japan), and Dr. Y.He (Duke University, Durham, NC, USA), respectively.Bcl-XL tg mice under the control of the 2.3-kb Col1a1 promoter were generated as previously described [27].Before the initiation of the present study, all experimental protocols were reviewed and approved by the Animal Care and Use Committee of Nagasaki University Graduate School of Biomedical Sciences (No. 1903131520-9).Animals were housed three per cage in a pathogen-free environment on a 12 h light cycle at 22 ± 2 • C, with standard chow (CLEA Japan, Tokyo, Japan) and free access to tap water.

Real-Time RT-PCR and Western Blot Analyses
Total RNA was extracted from bone marrow, femur, brain, lung, heart, liver, stomach, kidney, and spleen of wild-type mice at 10 weeks of age, newborn calvariae of wild-type and Bcl-XL tg mice, and from osteoblast-enriched samples from the femurs and tibiae of wild-type and Bcl-XL tg mice at 6 weeks of age using ISOGEN (Wako, Osaka, Japan).To obtain the osteoblast-enriched samples, muscle, connective tissue, and periosteum were removed from the femurs and tibiae, and the bones were cut at the metaphyses.After the hematopoietic cells in the diaphyses of femurs and tibiae were flushed out with PBS, the osteoblast-enriched cells were collected using a microintertooth brush (Kobayashi Pharmaceutical Co., Ltd., Osaka, Japan).Total RNA was also extracted from primary osteoblasts using ISOGEN.Real-time PT-PCR was performed using THUNDERBIRD SYBR qPCR Mix (Toyobo, Osaka, Japan) and Light Cycler 480 real-time PCR system (Roche Diagnostics, Tokyo, Japan).Primer sequences were as follows: Actb, 5 -CCACCCGCGAGCACAGCTTC-3 and 5 -TTGTCGACGACCAGCGCAGC-3 , Bcl2l1, 5 -GCAGGTATTGGTGAGTCG-3 and 5 -GGCTGCTGCATTGTTCCC-3 ; Tnfsf11, 5 -CAAGCTCCGAGCTGGTGAAG-3 and 5 -CCTGAACTTTGAAAGCCCCA-3 ; Tnfrsf11b, 5 -AAGAGCAAACCTTCCAGCTGC-3 and 5 -CACGCTGCTTTCACAGAGGTC-3 .We normalized values to those of Actb.Protein was extracted from femurs and tibiae, and Western blot analysis was performed using anti-Cre (Covance, Princeton, NJ, USA) and anti-Bcl-XL (Cell Signaling Technology, Danvers, MA, USA) antibodies.

X-Gal Staining
X-gal staining was performed using mice at 4 weeks of age.Mice were sacrificed and the long bones were dissected.They were fixed for 30 min in 0.1 M phosphate buffer with 5 mM EGTA, 2 mM MgCl 2 , and 4% paraformaldehyde, and rinsed three times for 15 min each with 0.1 M phosphate buffer containing 5 mM EGTA, 2 mM MgCl 2 , 0.02% Nonidet P-40, and 0.01% sodium deoxycholate.They were then stained with the same rinsing solution supplemented with 0.5 mg/mL of X-gal, 10 mM potassium ferrocyanide, and 10 mM potassium ferricyanide.After staining, samples were washed in the same rinsing solution for 24 h.

Micro-CT Analysis
A quantitative micro-CT analysis was performed with a micro-CT system (R_mCT; Rigaku Corporation, Tokyo, Japan).Data from scanned slices were used for a threedimensional analysis to calculate femoral morphometric parameters using the software (Ratoc, TRI-3D BON version R9.02.00.0-H-64,RATOC Systems, Inc., Osaka, Japan).Trabecular bone parameters were measured on a distal femoral metaphysis.Approximately 2.4 mm (0.5 mm from the growth plate) was cranio-caudally scanned, and 200 slices were taken at 12 µm intervals.Cortical thickness was measured at the mid-diaphyses of femurs.We used a threshold value of 275 to binarize the spongiosa and cortex.

Histological Analysis
To assess dynamic histomorphometric indices, mice were injected with calcein 13 d and 3 d before sacrifice.For bone histomorphometric analyses, femurs were fixed with 70% ethanol, and the undecalcified bones were embedded in glycol methacrylate.Three µm longitudinal sections from the distal parts of femurs were stained with toluidine blue and analyzed using a semiautomated system (Bone Histomorphometric System, System Supply, Nagano, Japan).The trabecular bone was measured in the secondary spongiosa commencing 480 µm below the growth plate and extending proximally for 560 µm.Calcein labeling in the cortical bone was measured using 50 µm polishing sections from the middiaphyses of femurs by cellSens Standard (OLYMPUS, Tokyo, Japan).
In histological analyses of long bones, mice were sacrificed and fixed in 4% paraformaldehyde/ 0.01 M phosphate-buffered saline, and the long bones were decalcified in 10% EDTA (pH = 7.4) and embedded in paraffin.Sections (thickness of 3 µm) were stained for TUNEL using the ApopTag ® system (Millipore, Darmstadt, Germany).In the BrdU incorporation study, 2-week-old mice were injected intraperitoneally with 100 µg BrdU/g body weight and sacrificed 1 h later.Sections were stained with a BrdU staining kit (Zymed, San Francisco, CA, USA).In the counting of TUNEL-positive or BrdU-positive osteoblastic cells, only cells in the distal primary spongiosa of femurs, which were recognized as osteoblastic cells from their morphology and attachment to trabecular bone, were counted.

Serum Testing
The serum levels of osteocalcin and TRAP5b were measured using a Mouse Osteocalcin Enzyme Immunoassay Kit (BTI, Stoughton, MA, USA) and Mouse TRAP Assay (Immunodiagnostic Systems, Boldon, UK), respectively.

Cell Culture
Primary osteoblasts were isolated from newborn calvariae by sequential digestion with 0.1% collagenase A and 0.2% dispase.Osteoblastic cells from the third to fifth fractions were pooled, plated on 24-well plates at a density of 1 × 10 5 /well in α-MEM supplemented with 10% fetal bovine serum (FBS).At confluency, cells were infected with an adenovirus expressing GFP or Cre at a multiplicity of infection (MOI) of 10 for 3 h.After 24 h, cells were plated on 24-well plates at a density of 1 × 10 5 /well, and after the culture for 24 h, the medium was changed to osteogenic medium containing 50 µg/mL ascorbic acid and 10 mM β-glycerophosphate.ALP and von Kossa staining was performed after the culture for 3 and 12 days, respectively, in the osteogenic medium.Staining for ALP activity and mineralization (von Kossa) was performed as previously described [13].ALP and von Kossa staining was quantified using ImageJ version 1.47.Total RNA was extracted after the culture for 4 days in the osteogenic medium.

In Vitro Osteoclastogenesis
BMMs were isolated by density gradient centrifugation using Ficoll-Paque TM (GE Healthcare, Tokyo, Japan) from the bone marrow of Bcl2l1 fl/fl mice at 12 weeks of age.Primary osteoblasts from the newborn calvariae of Bcl2l1 fl/fl mice were infected with an adenovirus expressing GFP or Cre for 3 h.After 24 h, cells at 1.5 × 10 5 cells/cm 2 were cocultured with BMMs at 2.5 × 10 5 cells/cm 2 in α-MEM containing 10% FBS in the presence of 50 µg/mL of ascorbic acid and 10 −8 M 1α, 25 (OH) 2 D 3 in 48-well plates for 4 days.Osteoclast formation was evaluated by TRAP staining.Cultured cells were fixed with 4% paraformaldehyde for 30 min, washed in 0.01M PBS at room temperature for 5 min, and incubated in acetate buffer (pH 5.0) containing, N, N-dimethylformamide, naphthol AS-MX phosphate, fast red AL salt, and 50 mM sodium tartrate at 37 • C for 15 min.
To evaluate bone resorption, primary osteoblasts from the newborn calvariae of Bcl2l1 fl/fl or Bcl2l1 fl/flCre mice at a density of 5 × 10 5 cells/cm 2 were co-cultured with BMMs at 5 × 10 5 cells/cm 2 on dentin slices (Wako) in 48-well plates for 8 days.The wells were treated by ultrasonication in 1 M ammonia water, dried, and observed by scanning microscopy (HITACHI, Miniscope TM-1000, Tokyo, Japan).

Measurement of ATP
Primary osteoblasts from Bcl2l1 fl/fl mice were plated on 48-well plates at a density of 1 × 10 5 /well.At confluency, cells were infected with an adenovirus expressing GFP or Cre at a MOI of 10 for 3 h.The medium was changed to fresh medium containing 1 U/mL of apyrase (NACALAI TESQUE, Inc., Kyoto, Japan).After a culture for 72 h, ATP levels in the supernatant were measured using an Intracellular ATP detection kit (IC100, TOYO B-Net, Co., Ltd.Tokyo, Japan).

Serum-Free Culture
Primary osteoblasts from Bcl2l1 fl/fl mice were plated on 24-well plates at a density of 1 × 10 5 /well and cultured in α-MEM with or without 10% FBS for 24 h, and TUNEL staining was performed using the ApopTag ® system (Millipore) or RNA was extracted from cells.The supernatant was used to measure ATP.Furthermore, the supernatant with 10% FBS was used for the culture of primary osteoblasts, the supernatant without 10% FBS was used as conditioned medium in the culture of primary osteoblasts after the addition of FBS to 10%, and RNA was extracted after the culture for 24 h.

Statistical Analysis
Statistical analyses were performed by the Student's t-test in the comparison of two groups except for Figure 7H, in which the four groups were compared by the Tukey-Kramer test, using BellCurve version 4.05 for Excel (Social Survey Research Information Co., Ltd., Tokyo, Japan).Data are presented as the mean ± S.D. A p-value < 0.05 was considered to be significant.

Figure 1 .
Figure 1.Generation of osteoblast-specific Bcl2l1 conditional knock-out mice.(A) Real-time RT-PCR analysis.Bcl2l1 expression was examined using RNA from bone marrow, the femur, brain, lung, heart, liver, stomach, kidney, and spleen of wild-type mice at 10 weeks of age.(B,C) β-Galactosidase staining of sections from the femoral metaphyses (B) and femoral diaphyses (C) of CAG-LacZ transgenic mice (LacZ) and CAG-LacZ: Cre double transgenic mice (LacZ; Cre) at 4 weeks of age.Scale bars: 50 µm (B,C).(D,E) Western blot analysis using anti-Bcl-XL and anti-Cre antibodies.β-actin was used as an internal control.The intensities of the Bcl-XL bands were normalized against β-actin, normalized values in the average of Bcl2l1 fl/fl mice were set as 1, and relative levels are shown in (E).Proteins were extracted from femurs at 5 weeks of age.n = 2. (F) Body weights of male and female Bcl2l1 fl/fl and Bcl2l1 fl/flCre mice at 10 weeks of age.n = 14 (male Bcl2l1 fl/fl ), n = 11 (male Bcl2l1 fl/flCre ), n = 11 (female Bcl2l1 fl/fl ), n = 12 (female Bcl2l1 fl/flCre ).

Figure 1 .
Figure 1.Generation of osteoblast-specific Bcl2l1 conditional knock-out mice.(A) Real-time RT-PCR analysis.Bcl2l1 expression was examined using RNA from bone marrow, the femur, brain, lung, heart, liver, stomach, kidney, and spleen of wild-type mice at 10 weeks of age.(B,C) β-Galactosidase staining of sections from the femoral metaphyses (B) and femoral diaphyses (C) of CAG-LacZ transgenic mice (LacZ) and CAG-LacZ: Cre double transgenic mice (LacZ; Cre) at 4 weeks of age.Scale bars: 50 µm (B,C).(D,E) Western blot analysis using anti-Bcl-XL and anti-Cre antibodies.β-actin was used as an internal control.The intensities of the Bcl-XL bands were normalized against β-actin, normalized values in the average of Bcl2l1 fl/fl mice were set as 1, and relative levels are shown in (E).Proteins were extracted from femurs at 5 weeks of age.n = 2. (F) Body weights of male and female Bcl2l1 fl/fl and Bcl2l1 fl/flCre mice at 10 weeks of age.n = 14 (male Bcl2l1 fl/fl ), n = 11 (male Bcl2l1 fl/flCre ), n = 11 (female Bcl2l1 fl/fl ), n = 12 (female Bcl2l1 fl/flCre ).

2. 3 .
Increases in Osteoclast Parameters in Bcl2l1 fl/flCre Mice in the Bone Histomorphometric Analysis

Figure 5 .
Figure 5.In vitro osteoblastogenesis.ALP (A) and von Kossa (C) staining, their quantification and a real-time RT-PCR analysis (E).Primary osteoblasts from Bcl2l1 fl/fl mice were infected w GFP-or Cre-expressing adenovirus.ALP and von Kossa staining was performed after the c for 3 and 12 days, respectively, in the osteogenic medium, and RNA was extracted after the c for 4 days in the osteogenic medium.Values for the GFP-expressing adenovirus were define and relative levels are shown.n = 9 in (B,D), n = 6 in (E).Similar results were obtained from tw D) and four (E) independent experiments and representative data are shown.(F,G) Real-tim PCR analysis of Tnfsf11 and Tnfrsf11b expression in Bcl-XL transgenic (tg) mice.RNA was pre from the newborn calvariae of wild-type (n = 15) and Bcl-XL tg (n = 9) mice (F), and from osteo enriched samples from the femurs and tibiae of wild-type (n = 5) and Bcl-XL tg (n = 6) mic weeks of age (G).(H) Real-time RT-PCR analysis of Tnfsf11 and Tnfrsf11b expression in pr osteoblasts from wild-type (n = 6) and Bcl-XL tg (n = 6) mice.* p < 0.05, ** p < 0.01, *** p < 0.00 p < 0.0001.

Figure 5 .
Figure 5.In vitro osteoblastogenesis.ALP (A) and von Kossa (C) staining, their quantification (B,D), and a real-time RT-PCR analysis (E).Primary osteoblasts from Bcl2l1 fl/fl mice were infected with a GFP-or Cre-expressing adenovirus.ALP and von Kossa staining was performed after the culture for 3 and 12 days, respectively, in the osteogenic medium, and RNA was extracted after the culture for 4 days in the osteogenic medium.Values for the GFP-expressing adenovirus were defined as 1 and relative levels are shown.n = 9 in (B,D), n = 6 in (E).Similar results were obtained from two (A-D) and four (E) independent experiments and representative data are shown.(F,G) Realtime RT-PCR analysis of Tnfsf11 and Tnfrsf11b expression in Bcl-XL transgenic (tg) mice.RNA was prepared from the newborn calvariae of wild-type (n = 15) and Bcl-XL tg (n = 9) mice (F), and from osteoblast-enriched samples from the femurs and tibiae of wild-type (n = 5) and Bcl-XL tg (n = 6) mice at 6 weeks of age (G).(H) Real-time RT-PCR analysis of Tnfsf11 and Tnfrsf11b expression in primary osteoblasts from wild-type (n = 6) and Bcl-XL tg (n = 6) mice.* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Figure 6 .
Figure 6.In vitro osteoclastogenesis.(A,B) TRAP staining (A) and the number of TRAP-positive cells (B).Primary osteoblasts from Bcl2l1 fl/fl mice were infected with the GFP-or Cre-expressing adenovirus, and infected cells were co-cultured with BMMs from Bcl2l1 fl/fl mice for 4 days.The number of multinucleated TRAP-positive cells in 48-well plates is shown in (B).n = 4. (C,D) Pit assay.Primary osteoblasts from Bcl2l1 fl/fl or Bcl2l1 fl/flCre mice were co-cultured with BMMs from Bcl2l1 fl/fl mice on dentin slices for 8 days (C).The pit area in 1 mm 2 was measured (D).n = 6.Scale bars: 200 µm (C).Similar results were obtained from two independent experiments and representative data are shown.* p < 0.05, *** p < 0.001.

Figure 6 .
Figure 6.In vitro osteoclastogenesis.(A,B) TRAP staining (A) and the number of TRAP-positive cells (B).Primary osteoblasts from Bcl2l1 fl/fl mice were infected with the GFP-or Cre-expressing adenovirus, and infected cells were co-cultured with BMMs from Bcl2l1 fl/fl mice for 4 days.The number of multinucleated TRAP-positive cells in 48-well plates is shown in (B).n = 4. (C,D) Pit assay.Primary osteoblasts from Bcl2l1 fl/fl or Bcl2l1 fl/flCre mice were co-cultured with BMMs from Bcl2l1 fl/fl mice on dentin slices for 8 days (C).The pit area in 1 mm 2 was measured (D).n = 6.Scale bars: 200 µm (C).Similar results were obtained from two independent experiments and representative data are shown.* p < 0.05, *** p < 0.001.

Figure 7 .
Figure 7. TUNEL-staining and measurement of ATP in vitro.(A-C) Primary osteoblasts from wildtype mice were cultured in the presence or absence of 10% FBS for 24 h and then stained for TUNEL (A).The frequencies of TUNEL-positive cells were calculated (B).The amount of ATP in the culture medium was quantitated by luciferase (C).n = 8. (D,E) Real-time RT-PCR analysis of Tnfsf11 and Tnfrsf11b.Primary osteoblasts were cultured in the presence or absence of 10% FBS for 24 h (D).The supernatant with 10% FBS was used for the culture of primary osteoblasts (control), and the supernatant without 10% FBS was used as conditioned medium in the culture of primary osteoblasts after the addition of FBS to 10% (conditioned medium + 10% FBS) (E).n = 5. (F-H) Primary osteoblasts from Bcl2l1 fl/fl mice were infected with the GFP-or Cre-expressing adenovirus.After 48 h, cells were stained for TUNEL (F), and the frequencies of TUNEL-positive cells were calculated (G).The amount of ATP in the culture medium with or without apyrase was quantitated (H).n = 13 (GFP), n = 6 (GFP + apyrase), n = 16 (Cre), n = 6 (Cre + apyrase).*, # p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.Scale bars: 50 µm.Similar results were obtained from two independent experiments and representative data are shown.3.Discussion

Figure 7 .
Figure 7. TUNEL-staining and measurement of ATP in vitro.(A-C) Primary osteoblasts from wildtype mice were cultured in the presence or absence of 10% FBS for 24 h and then stained for TUNEL (A).The frequencies of TUNEL-positive cells were calculated (B).The amount of ATP in the culture medium was quantitated by luciferase (C).n = 8. (D,E) Real-time RT-PCR analysis of Tnfsf11 and Tnfrsf11b.Primary osteoblasts were cultured in the presence or absence of 10% FBS for 24 h (D).The supernatant with 10% FBS was used for the culture of primary osteoblasts (control), and the supernatant without 10% FBS was used as conditioned medium in the culture of primary osteoblasts after the addition of FBS to 10% (conditioned medium + 10% FBS) (E).n = 5. (F-H) Primary osteoblasts from Bcl2l1 fl/fl mice were infected with the GFP-or Cre-expressing adenovirus.After 48 h, cells were stained for TUNEL (F), and the frequencies of TUNEL-positive cells were calculated (G).The amount of ATP in the culture medium with or without apyrase was quantitated (H).n = 13 (GFP), n = 6 (GFP + apyrase), n = 16 (Cre), n = 6 (Cre + apyrase).*, # p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.Scale bars: 50 µm.Similar results were obtained from two independent experiments and representative data are shown.3.Discussion.
Figure 7. TUNEL-staining and measurement of ATP in vitro.(A-C) Primary osteoblasts from wildtype mice were cultured in the presence or absence of 10% FBS for 24 h and then stained for TUNEL (A).The frequencies of TUNEL-positive cells were calculated (B).The amount of ATP in the culture medium was quantitated by luciferase (C).n = 8. (D,E) Real-time RT-PCR analysis of Tnfsf11 and Tnfrsf11b.Primary osteoblasts were cultured in the presence or absence of 10% FBS for 24 h (D).The supernatant with 10% FBS was used for the culture of primary osteoblasts (control), and the supernatant without 10% FBS was used as conditioned medium in the culture of primary osteoblasts after the addition of FBS to 10% (conditioned medium + 10% FBS) (E).n = 5. (F-H) Primary osteoblasts from Bcl2l1 fl/fl mice were infected with the GFP-or Cre-expressing adenovirus.After 48 h, cells were stained for TUNEL (F), and the frequencies of TUNEL-positive cells were calculated (G).The amount of ATP in the culture medium with or without apyrase was quantitated (H).n = 13 (GFP), n = 6 (GFP + apyrase), n = 16 (Cre), n = 6 (Cre + apyrase).*, # p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.Scale bars: 50 µm.Similar results were obtained from two independent experiments and representative data are shown.3.Discussion.