Mechanical unloading, such as in a microgravity environment in space or during bed rest (for patients who require prolonged bed rest), leads to a decrease in bone mass because of the suppression of bone formation and the stimulation of bone resorption. To address the challenges presented by a prolonged stay in space and the forthcoming era of a super-aged society, it will be important to prevent the bone loss caused by prolonged mechanical unloading. Nuclear factor κB (NF-κB) transcription factors are activated by mechanical loading and inflammatory cytokines. Our objective was to elucidate the role of NF-κB pathways in bone loss that are caused by mechanical unloading. Eight-week-old wild-type (WT) and NF-κB1-deficient mice were randomly assigned to a control or mechanically unloaded with tail suspension group. After 2 weeks, a radiographic analysis indicated a decrease in bone mass in the tibias and femurs of the unloaded WT mice but not in the NF-κB1–deficient mice. An NF-κB1 deficiency suppressed the unloading-induced reduction in bone formation by maintaining the proportion and/or potential of osteoprogenitors or immature osteoblasts, and by suppression of bone resorption through the inhibition of intracellular signaling through the receptor activator of NF-κB ligand (RANKL) in osteoclast precursors. Thus, NF-κB1 is involved in two aspects of rapid reduction in bone mass that are induced by disuse osteoporosis in space or bed rest.
Mechanical stimulation is critical for the maintenance of skeletal integrity and bone mass.[1, 2] Especially in a super-aged society, disuse osteoporosis is a critical issue in patients with reduced locomotor function in bedridden patients due to cardiovascular, brain, and skeletal diseases. In such an unloading-induced osteoporosis, the loss of mechanical stress is responsible for impairments in the maintenance of bone mass. Even in a general physical point of view, reduced physical activity per se in aged patients also significantly reduces bone mass. This disuse-induced bone loss and osteoporosis is not limited to the aged population but also observed in young adults who are suffering from joint immobilization due to fractures and dislocations. Disuse-induced bone loss clearly occurs even in healthy astronauts. This phenomenon is acutely observed during spaceflights and continues to persist long after the astronauts return to the Earth. Thus, the link between disuse and bone loss is well-established, and unloading suppresses bone formation and activates bone resorption. These phenomena are exerted by many types of bone cells, including osteoblasts, osteocytes, and osteoclasts.[1-4] However, the molecular mechanism underlying the alterations of cellular activity and disuse-induced bone loss is not fully understood.
The transcription factor nuclear factor κB (NF-κB) plays a key role in immune and inflammatory responses, proliferation, tumorigenesis, and survival.[5, 6] Currently, five proteins with a conserved homology in the Rel domain (p65 [RelA], RelB, cRel, NF-κB1 [p50/p105], and NF-κB2 [p52/p100]) have been identified. These members can homodimerize and heterodimerize in numerous combinations, with the predominant cellular species being p50:p65, p50:cRel, and NF-κB2:RelB. All five members share an N-terminal domain of 300 amino acids, designated the Rel homology domain (RHD), which is responsible for DNA binding, dimerization, and interactions with the inhibitory IκB proteins. NF-κB1 and NF-κB2 are synthesized as the large precursors of p105 and p100, respectively, and contain long COOH-terminal domains with multiple ankyrin repeats, rendering these precursors functionally similar to the inhibitor IκB. The inhibitory effect of p105 and p100 is relieved when these proteins are processed into p50 and p52, respectively. Three members, p65, c-Rel, and RelB, contain C-terminal transcriptional activation domains (TADs) that are crucial for their ability to induce target gene expression, whereas homodimers of p50 and p52 lack TADs and therefore have no intrinsic ability to drive transcription. In unstimulated cells, NF-κB is predominantly localized to the cytoplasm as part of a complex with inhibitory IκB proteins, including IκBα, IκBβ, IκBϵ, and IκBγ. In response to a variety of stimuli, such as tumor necrosis factor α (TNFα) and interleukin 1β (IL-1β), IκBs are phosphorylated (Ser-32 and Ser-36 for IκBα and Ser-19 and Ser-21 for IκBβ) by the activated IκB kinase (IKK) complex. Phosphorylated IκBs are ubiquitinated and subsequently degraded by the 26S proteasome. The IKK complex consists of two catalytic kinase subunits, specifically, IKKα (IKK1) and IKKβ (IKK2), as well as a regulatory subunit, NF-κB essential modulator (NEMO), also known as IKKγ. IKKβ is critical for the classical (canonical) NF-κB pathway that depends on IκB degradation. In this pathway, the p50/p65 heterodimer enters the nucleus and binds to NF-κB-responsive elements to regulate the expression of genes.[5, 6]
The importance of NF-κB in both bone formation and bone resorption is well-known. In osteoclast development, two groups have reported that mice lacking both the NF-κB1 (p50) and NF-κB2 (p52) subunits develop typical osteopetrosis, which is accompanied by a dramatic reduction in the number of osteoclasts due to the defective tracking of the osteoclast lineage.[7, 8] Certain inhibitors of the NF-κB pathway suppress the receptor activator of the NF-κB ligand (RANKL), inducing osteoclastogenesis and animal models of inflammatory bone destruction, such as collagen-induced arthritis or ovariectomized mice.[9, 10] A recent study indicates that the inhibition of NF-κB in mature mice osteoblasts expressing a dominant-negative form of IKKβ increases bone mineral density and bone volume, due to the increased activity of the osteoblasts. Furthermore, the selective inhibition of NF-κB restores the inhibitory effect of TNF on bone morphogenic protein 2 (BMP2)-induced osteoblast differentiation and prevents bone loss in ovariectomized mice.[12, 13] We have previously reported that the activation of the classical NF-κB pathway induced by TNFα inhibits BMP2-induced osteoblast differentiation by inhibiting the DNA-binding activity of the Smad protein to its target genes. These results strongly suggest that the activation of NF-κB induces osteoclastogenesis and suppresses bone formation.
Although NF-κB transcription factors are activated by mechanical loading and inflammatory cytokines,[14, 15] the role that NF-κB exerts on the maintenance of bone mass with disuse-induced osteoporosis still remains unclear. Shear stress stimulates osteoblastic differentiation via the prostaglandin synthesis induced by NF-κB.[14, 16] Conversely, the activity of the NF-κB-dependent reporter gene is markedly increased in unloaded muscles.[17, 18] Thus, there are many unknown mechanisms with respect to whether the NF-κB activation induced by mechanical unloading produces a positive or negative regulation of bone formation. In this study, our objectives were to elucidate (1) the NF-κB activation mechanism regulated by mechanical unloading, and (2) the role of NF-κB activation on bone formation using gene targeting for molecules that are important for NF-κB signaling.
Materials and Methods
Eight-week-old WT and NF-κB1–deficient mice were purchased from Nihon Crea (Tokyo, Japan) and JAX (Bar Harbor, ME, USA), respectively. Mice homozygous for NF-κB1 were backcrossed with C57/BL6 more than 10 times. The genotypes of the offspring were screened using PCR. All mice were maintained at the Animal Resource Center. These experimental procedures were approved by the Animal Care and Use Committee of Kyushu Dental College (approval no. 12–003).
Tail suspension model
Tail suspensions were performed as described,[19, 20] with slight modifications. WT and NF-κB1–deficient mice were placed in special cages for 2 weeks to perform the tail suspensions (Takatsuka Life Science Co., Okayama, Japan).
Glutathione S-transferase (GST)-RANKL was generously provided by the Oriental Yeast Company, Ltd. (Shiga, Japan). Recombinant human macrophage colony-stimulating factor (M-CSF) was purchased from PeproTech, Inc. (Rocky Hill, NJ, USA). Anti-p50 (sc-8414), RelB (sc-226), and IκBα (sc-371) antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-p65 (SA-171) and nuclear factor of activated T cells, cytoplasmic 1 (NFATc1) (ab25916) were purchased from Biomol (Plymouth Meeting, PA, USA) and abcam (Cambridge, MA, USA), respectively. The anti-β-actin antibody was purchased from Sigma-Aldrich (St. Louis, MO, USA).
After euthanasia, the tibias and femurs were dissected and the soft tissue was removed. All bones were subsequently fixed in a phosphate-buffered saline (PBS)-buffered glutaraldehyde (0.25%)–formalin (4%) fixative (pH 7.4) for 2 days (4°C) and washed with PBS for further studies. The bone mineral densities (BMD) of the tibias and femurs were measured using peripheral quantitative computed tomography (pQCT; XCT Research SA + , Stratec Medizintechnik GmbH, Pforzheim, Germany) as described.[21, 22] Three-dimensional (3D) reconstruction images of proximal tibias were obtained using focal micro–computed tomography (µCT) (ScanXmate-E090; Comscan, Kanagawa, Japan) as described.[21, 22]
Histological preparation and bone histomorphometry
Tibias and femurs from the WT or NF-κB1–deficient mice, either control or unloaded, were embedded in mixtures of methyl methacrylate (MMA) and 2-hydroxyethyl methacrylate (GMA) resins as described.[21, 22] Sagittal sections (4 µm) of the long bones were prepared. These sections were then stained according to von Kossa with modified Van Giesson method to clarify the mineralized tissue. Some sections were stained for tartrate-resistant acid phosphatase (TRAP) and counterstained with methyl green. Osteoclasts were designated as the TRAP-positive (TRAP+) multinucleated cells (MNCs) containing more than three nuclei and located on the bone surface. Standard bone histomorphometric analyses were performed in the secondary spongiosa of the tibias, starting at 0.3 mm distal from the proximal growth plate to exclude the primary spongiosa, using an image analyzing system (KS400, Carl Zeiss, Jena, Germany).
Osteoblasts and osteocytes isolation
Osteoblastic as well as osteocytic cells were enriched by sequential collagenase digestion as reported. After removing bone marrow cells, pieces of femur and tibia bones from loaded or unloaded were digested with 250 mg/mL collagenase (Wako Pure Chemical Industries, Ltd., Osaka, Japan) for 30 minutes. Osteoblast fractions were released after the first 30 minutes and three subsequent digestions at 37°C for 20 minutes each. Osteocyte-rich populations were obtained from the residual bone pieces by an initial incubation in PBS containing 4 mM EDTA for 15 minutes to yield Fraction 5-1, followed by a 15-minutes incubation in a collagenase solution for 15 minutes to give Fraction 5-2. This procedure was repeated to yield Fractions 7-1 and 7-2. Following the final digestion, cell suspensions were sequentially passed through 100-µm and 40-µm filters.
Real-time PCR analysis
After the bone marrow cells were collected, loaded, and unloaded, the femurs were homogenized in TRIzol reagent (Invitrogen, Carlsbad, CA, USA). Two μg of total RNA was synthesized from first-strand cDNA using SuperScript II transcriptase and random primers (Invitrogen). Real-time PCR was performed using SYBR Green PCR master mix and the 7300 Real-time PCR system (Applied Biosystems, Foster City, CA, USA) according to the manufacturer's instructions. Samples were matched to a standard curve generated by amplifying serially diluted products under the same PCR conditions. GAPDH expression served as an internal control. The primer sequences were as follows: p50, 5′-GCAAACCTGGGAATACTTCATGTGACTAAG-3′ (forward) and 5′-ATAGGCAAGGTCAGAATGCACCAGAAGTCC-3′ (reverse); RANKL, 5′-CAGCATCGCTCTGTTCCTGTA-3′ (forward) and 5′-CTGCGTTTTCATGGAGTCTCA-3′ (reverse); osteoprotegerin (OPG), 5′-ACCCAGAAACTGGTCATCAGC-3′ (forward) and 5′-CTGCAATACACACACTCATCACT-3′ (reverse); BMP2, 5′-GCTCCACAAACGAGAAAAGC-3′ (forward) and 5′-AGCAAGGGGAAAAGGACACT-3′ (reverse); BMP4, 5′-CCTGGTAACCGAATGCTGAT-3′ (forward) and 5′-AGCCGGTAAAGATCCCTCAT-3′ (reverse); Wnt3a, 5′-ATGGCTCCTCTCGGATACCT-3′ (forward) and 5′-GGGCATGATCTCCACGTAGT-3′ (reverse); Wnt5a, 5′-ATTGTCCCCCAAGGCTTAAC-3′ (forward) and 5′-AACTTGGAAGACATGGCACC-3′ (reverse); Sost, 5′-AGCCTTCAGGAATGATGCCAC-3′ (forward) and 5′-AAGGCAGTAGGGAAACTGGGA-3′ (reverse); Keratocan, 5′-TGGGATGTCCACGACGACTT-3′ (forward) and 5′-AAGGCAGTAGGGAAACTGGGA-3′ (reverse); alkaline phosphatase (ALP), 5′-CCAACTCTTTTGTGCCAGAGA-3′ (forward) and 5′-GGCTACATTGGTGTTGAGCTTTT-3′ (reverse) and GAPDH, 5′-AACTTTGGCATTGTGGAAGG-3′ (forward) and 5′-ACACATTGGGGTAGGAACA-3′ (reverse).
Detection of ALP activity
Bone marrow cells were seeded at a density of 1.0 × 104 cells/well in 96-well plates and cultured in α-minimal essential medium (α-MEM) containing 10% fetal bovine serum (FBS) in the presence or absence of β-glycerophosphate (10 mM) and ascorbic acid (50 μg/mL) for 7 days. The cells were fixed with an acetone/ethanol mixture (50:50, vol/vol) and incubated with a substrate solution (0.1 M diethanolamine, 1 mM MgCl2, and 10 mg/mL p-nitrophenyl phosphate). The reaction was terminated by the addition of 5 M NaOH, and the absorbance was measured at 405 nm using a microplate reader (Bio-Rad, Hercules, CA, USA). All experiments were performed in triplicate.
Colony formation assay
Bone marrow cells from the tibias of the loaded and unloaded groups of WT or NF-κB1–deficient mice were inoculated at a cell density of 2.5 × 105 cells/well in 12-well plates and cultured with or without β-glycerophosphate and ascorbic acid for 7 days. The cells were stained for ALP, and percentages of ALP-positive colony-forming unit–fibroblasts (CFU-Fs) (%) were counted separately in each well.
In vitro osteoclastogenesis
Bone marrow–derived macrophages (BMMs) were prepared as osteoclast precursors from WT or NF-κB1–deficient mice. Bone marrow cells obtained from mouse tibias were suspended in 96-well plates for 16 hours in the presence of M-CSF (100 ng/mL) in an α-MEM containing 10% FBS, 100 units/mL penicillin, and 100 µg/mL streptomycin. Nonadherent cells were harvested and further cultured for 2 days with M-CSF (100 ng/mL). The adherent cells, most of which expressed macrophage-specific antigens, such as Mac-1, Moma-2, and F4/80, were used as BMMs.[21, 22] BMMs were cultured for 3 days with RANKL (100 ng/mL). Cultures were fixed with 3.7% formaldehyde, and osteoclasts were detected by staining for TRAP. TRAP+ MNCs were observed under a microscope and counted as osteoclasts.[21, 22]
Single-cell suspension of bone marrow cells were centrifuged and resuspended in PBS containing 1% FBS. Cells were stained with anti-c-kit–fluorescein isothiocyanate (FITC), anti-RANK-phycoerythrin (PE), and ant-c-fms-Cy5 antibodies for 30 minutes on ice, washed with PBS 1% FBS and analyzed using a FACSCalibur flow cytometer (Becton Dickinson).
Coculture and pit formation
Osteoblasts were obtained from the calvariae of newborn C57BL6 mice by digestion with 0.1% collagenase (Wako) and 0.2% dispase (Goso Shusei, Tokyo, Japan), and bone marrow cells obtained from each mice were cocultured in α-MEM containing 10% FBS, 1α,25-dihydroxyvitamin D3 (10−8 M) (Wako), and prostaglandin E2 (10−6 M) (Sigma) in 100-mm-diameter dishes coated with collagen gel (Nitta Gelatin, Osaka, Japan). Osteoclasts were formed within 6 days in culture and were removed from the dishes by treating them with 0.2% collagenase (Wako). Mature osteoclasts were seeded onto dentin slices (φ 4 mm; Kyushu-do, Kitakyushu, Japan). The slices were placed in 48-well plates containing α-MEM supplemented with 10% FBS and then cultured for 48 hours. The cells were scraped off the dentine slices, and the slices were stained with Mayer's hematoxylin. The stained pit areas were photographed under a microscope at ×40 magnification. The area of each pit was measured from the photographs with Scion image software. In some experiments, mature osteoclasts were seeded onto dentin slices for 18 hours and then were fixed and stained for TRAP.
Western blot analysis
Bone tissue extracts and BMMs treated with RANKL for the indicated periods of time were lysed in Tris-buffered saline (20 mM Tris-HCl and 200 mM NaCl) containing 1% Triton X-100 and protease inhibitors (eg, aprotinin, pepstatin, dithiothreitol, and leupeptin). The lysates were resolved by 10% SDS-PAGE, transferred to Immobilon-P membranes (Millipore, Billerica, MA, USA), and immunoblotted with individual antibodies. The membranes were later washed and incubated with horseradish peroxidase–conjugated secondary antibodies (Santa Cruz Biotechnology). The immunoreactive proteins were visualized using enhanced chemiluminescence (ECL; Amersham, Piscataway, NJ, USA) and analyzed using a Luminescent image analyzer (Fujifilm, Tokyo, Japan).
Comparisons were made using factorial ANOVA. When significant F values were detected, Fisher's projected least significant difference (PLSD) post hoc test was performed to compare each of the groups. The data were expressed as the mean ± SD; values of p < 0.05 were considered significant.
Expression levels of NF-κB family members in unloaded bone
We first examined the changes in the expression levels of NF-κB family members in bone after unloading for 2 weeks. Among these proteins, the expression levels of NF-κB1 and p50 were increased by 1.7-fold; however, p105 could not be detected (Fig. 1A). In contrast, the expression levels of other NF-κB family members, such as p65, RelB, or IκBα, were unaltered, even after unloading (Fig. 1A). To determine whether unloading enhances p50 expression through changes at the transcription level or through protein stability, we performed real-time PCR analyses on either loaded or unloaded tibias using mouse p50 primers. Unloading induced p50 expression, suggesting that unloading increased p50 protein through upregulation at the transcriptional level (Fig. 1B). We next examined the expression levels of p50 in osteoblasts and osteocytes during unloading for a period of 1 week. We confirmed the purity of osteocyte and osteoblast fractions by real-time PCR using the expression levels of Sost, a marker of osteocytes, and Keratocan, a marker of osteoblasts (Fig. 1C). We observed that p50 expression was increased in osteocytes but not in osteoblasts after unloading, suggesting that the increased expression of the p50 protein in total bone tissue is due to increased p50 expression in osteocytes not osteoblasts (Fig. 1C).
NF-κB1 deficiency suppresses unloading-induced secondary trabecular bone loss
To test the hypothesis that NF-κB1 is involved in the mechanical regulation of bone mass, NF-κB1–deficient mice and WT mice were subjected to hindlimb unloading. The baseline body weight levels were similar between the WT and NF-κB1–deficient mice. The soft X-ray pictures indicated that unloading led to trabecular bone loss in the tibias and femurs of the WT mice but not in the NF-κB1–deficient mice (Fig. 2A). A three-dimensional μCT analysis indicated that unloading led to sparsity in the architecture of the secondary trabecular bone of the WT mice, which has been previously demonstrated (Fig. 2B).[19, 20] In the control NF-κB1–deficient mice, the μCT pictures revealed that a pattern of trabecular bone architecture that was slightly more dense than that of the control WT mice. In contrast to the WT mice, the NF-κB1–deficient mice did not demonstrate any pattern changes induced by hindlimb unloading (Fig. 2B). In the WT mice, the quantification of BMD indicated that unloading induced a reduction of approximately 30% after 2 weeks (Fig. 2C). The NF-κB1–deficient mice exhibited slightly increased base line levels of BMD compared to the WT mice, and the level of this parameter was not reduced in the NF-κB1–deficient mice even after unloading (Fig. 2C). After unloading, von Kossa staining also clearly showed decreased trabecular bone volume in WT mice but not in NF-κB1–deficient mice (Fig. 2D). Other microarchitectural parameters, such as the bone volume/tissue volume (BV/TV) and the number of trabecular bones (Tb.N), but not the trabecular thickness (Tb.Th), were also reduced by approximately 50% due to unloading in the WT mice, whereas these parameters were not reduced in the NF-κB1–deficient mice even after unloading (Fig. 2E). In the WT, unloading reduced the cortical bone area and cortical thickness of the midshaft region of the bone, whereas a lack of NF-κB1 suppressed such unloading-induced cortical bone loss (Fig. 2F). Thus, NF-κB1 deficiency preserved bone volume, even after unloading.
NF-κB1 deficiency suppresses the unloading-induced reduction of bone formation
To understand further the mechanisms for such an NF-κB1 deficiency-induced suppression of bone loss, we examined dynamic bone formation parameters using the analysis of calcein double-labeling. This analysis provided an in vivo estimation of osteoblastic activity with respect to the accumulation of bone mass (Fig. 3A). In the WT mice, the mineral apposition rate (MAR) was reduced by approximately 40% as a result of the unloading (Fig. 3B). NF-κB1 deficiency did not alter the baseline levels of MAR; however, the deficiency suppressed the reduction in MAR levels induced by unloading (Fig. 3B). The bone formation rate (BFR) was also reduced by approximately 45%, which was due to unloading in the WT but not in the NF-κB1–deficient mice (Fig. 3B). After unloading, the expression of ALP mRNA was reduced by approximately 40% in WT but not NF-κB1–deficient mice (Fig. 3C). These data indicate that even though NF-κB1 is not involved in the maintenance of baseline bone mass, at least one of the targets of the NF-κB pathway during the changes caused by mechanical stimuli is bone formation. To analyze the cellular effects of a lack of osteoblastic differentiation, bone marrow cells from either the WT or NF-κB1–deficient mice after being unloaded for 2 weeks were cultured in the presence or absence of β-glycerophosphate and ascorbic acid to induce osteoblastic differentiation. Cells from the WT but not the NF-κB1–deficient mice displayed an unloading-induced reduction in ALP activity (Fig. 3D). Because it has been suggested that the committed progenitor cells for osteogenesis are CFU-Fs expressing a high level of ALP activity, we performed a CFU-F assay using bone marrow cells isolated from the tibias of either loaded or unloaded WT or NF-κB1–deficient mice. Unloading significantly reduced the proportion of ALP-positive CFU-Fs from the WT mice in the presence of β-glycerophosphate and ascorbic acid, whereas an NF-κB1 deficiency did not alter the proportion of ALP-positive CFU-Fs during unloading (Fig. 3E). Overall, these data revealed that a lack of NF-κB1 prevents an unloading-induced reduction in bone-forming activity both in vivo and in vitro.
NF-κB1 deficiency suppresses unloading-induced increases in osteoclasts in vivo
Because bone mass levels are determined by bone formation and bone resorption activities, we examined the effects of NF-κB1 deficiency on bone resorption in these mice. A quantification of TRAP-positive MNCs at the primary trabecular regions indicated that unloading-induced increases led to a more than twofold increase in the number of TRAP-positive MNCs in the WT mice (Fig. 4A, B). In the control NF-κB1–deficient mice, the number of TRAP-positive cells was slightly decreased compared with the WT mice (Fig. 4B). In contrast to the WT, NF-κB1 deficiency suppressed unloading-induced increases in the number of TRAP-positive MNCs in the primary trabecular regions (Fig. 4A, B). Unloading induced increases by approximately 2.5-fold at the osteoclast surfaces of the primary trabecular regions of the WT mice, and NF-κB1 deficiency also suppressed such alterations due to unloading (Fig. 4B). Consistent with these results, trabecular bone separation (Tb.Sp) was increased by approximately twofold in the WT mice but not in the NF-κB1–deficient mice (Fig. 4C). However, in the WT and NF-κB1–deficient mice, unloading stimulated the expression level of RANKL mRNA approximately 10-fold, whereas the expression level of OPG mRNA was not changed after unloading (Fig. 4D). Consistent with previous reports, RANKL expression is higher in osteocytes than in osteoblasts during both loading and unloading; however, mRNA expression levels did not differ between WT and NF-κB–deficient cells (Fig. 4E).
NF-κB1 deficiency suppresses unloading-induced increases in osteoclasts by inhibiting RANK downstream signals
To examine further the cellular mechanisms underlying a lack of NF-κB1 suppression during bone resorption upon unloading in vivo, osteoclast differentiation in culture was investigated using bone marrow cells derived from mice subjected to hindlimb unloading. Bone marrow cells were cultured in the presence of M-CSF and RANKL to induce osteoclastogenesis. Unloading in the WT mice increased osteoclastogenesis in vitro; however, unloading in NF-κB1–deficient mice did not alter osteoclastogenesis in vitro (Fig. 5A). We next asked whether the failure to increase the osteoclast number following unloading is intrinsic to osteoclastogenesis in the bone marrow cells of the NF-κB1–deficient mice. The efficiency of osteoclast formation in bone marrow cell cultures obtained from the tibias of the unloaded and loaded mice in the presence of RANKL and M-CSF was enhanced in the WT mice but not in the NF-κB1–deficient mice. This finding suggests that one of the possible mechanisms underlying the failure of unloading to increase the number of osteoclasts in the NF-κB1–deficient mice is intrinsic to the NF-κB1 pathway. Unloading resulted in a slight increase in the population of osteoclast precursors (c-kit+, c-fms+, RANK+) as a percentage of the total cell population in the WT mice. Although the basal population of osteoclast precursors increased in the NF-κB1–deficient mice, unloading did not alter the total population of osteoclasts in these mice (Fig. 5B). We further examined the effect of NF-κB1 deficiency on the downstream signaling of RANK. When osteoclast precursors were treated with RANKL for the indicated periods of time, RANKL-induced expression of c-Fos and NFATc1, which is a master regulator of osteoclastogenesis, were enhanced after unloading in WT cells. Conversely, NF-κB1 deficiency did not alter the RANKL-induced expressions of c-Fos and NFATc1 after unloading (Fig. 5C). Finally, we examined the pit-forming activity of osteoclasts in WT and NF-κB1–deficient mice from the loaded and unloaded groups. Although unloading increased the pit area resorbed by osteoclasts in the WT mice, unloading did not enhance the pit area in NF-κB1–deficient mice (Fig. 5D). Bone resorbing activity corresponded to the difference in the number of osteoclasts between the WT and NF-κB1–deficient mice (Fig. 5D).
Unloading leads to rapid bone loss and results in disuse osteoporosis. However, the molecular basis for the cellular activities underlying this phenomenon is still incompletely understood. In the present study, we determined that the expression level of the NF-κB p50 subunit was increased in bone tissue after unloading for 2 weeks compared with the loaded condition. Unloading significantly reduced trabecular bone volume, whereas NF-κB1 deficiency suppressed such reductions in bone mass. Furthermore, NF-κB1 deficiency suppressed both of the two characteristic features of disuse osteoporosis, namely the unloading-induced reduction in osteoblastic bone formation and the enhancement of osteoclastic bone resorption. Although soft X-ray images indicated that the forelimbs of NF-κB1–deficient mice were denser than those of the control WT mice, unloading only slightly increased bone density in WT and NF-κB1–deficient mice (Supplemental Data 1). However, the expression of ALP or RANKL mRNA and the differentiation into osteoclasts of bone marrow cells from the forelimbs of WT or NF-κB1–deficient mice remained similar after unloading (Supplemental Data 1). These results strongly indicate that disuse-induced osteoporosis in unloaded limbs are associated with an NF-κB1–mediated pathway that is exclusively initiated by unloading.
Members of the NF-κB family are activated by a broad range of signals, including cytokines, mitogens, free radicals, and stress.[5, 6] The activation of NF-κB by inflammatory cytokines, including TNFα or IL-1, suppresses osteoblastic bone formation and enhances osteoclastic bone resorption.[12, 29] The present work demonstrates that disuse osteoporosis in healthy young mice is also associated with the activation of NF-κB. However, the trigger does not appear to be of a cytokine origin, and the NF-κB pathway is distinct from that observed with TNFα because of a lack of NF-κB p65 activation. Previous studies have demonstrated that the activation of a NF-κB–dependent reporter gene was markedly increased in unloaded muscles and that the p50 and Bcl3 proteins are implicated in this induction.[17, 18] Moreover, unloading-induced muscle atrophy is suppressed in both NF-κB1–deficient and Bcl3-deficient mice, suggesting that both the NF-κB1 and Bcl3 genes are necessary for unloading-induced muscle atrophy and the associated phenotype transition. Hindlimb unloading induces significant muscle atrophy, which could contribute to bone loss in this model, due to a reduction in muscle contractile forces. Indeed, the preservation of muscle mass via exercise or with transgenic animal models has been shown to reduce bone loss during unloading.[32, 33] These results strongly suggest that NF-κB1 and Bcl3 are involved in bone loss and the muscle atrophy induced by unloading.[17, 18] However, elucidating the mechanism of activation of NF-κB1 and Bcl3 during bone loss requires further investigation.
The cells that are responsible for the sensation of mechanical stress have been sought for many years. Osteocytes and osteoblasts are candidate cells for sensing mechanical stress and are thought to be involved in the mechanical stress–dependent regulation of bone mass. Recent reports have revealed that the targeted ablation of osteocytes in mice results in a resistance to unloading-induced bone loss. RANKL expression in osteoblasts is increased in WT mice but not in these “osteocyte-less” mice.[19, 20] In contrast, a recent report indicates that purified osteocytes express a significantly higher level of RANKL and have a much greater capacity to support osteoclastogenesis in vitro than osteoblasts. Furthermore, mice specifically lacking RANKL in their osteocytes develop severe osteopetrosis. These results strongly suggest that osteocytes control osteoclastogenesis either directly or indirectly through the modulation of RANKL expression in osteoblasts. We observed that p50 expression was increased in osteocytes but not in osteoblasts after unloading, suggesting that increasing p50 protein in bone tissue reflects the increasing p50 protein level in osteocytes during unloading. However, the RANKL expression level in osteocytes from WT and NF-κB1–deficient mice was comparable during both loading and unloading. We examined the expression level of BMPs and Wnts in osteoblasts and osteocytes during unloading or loading conditions in WT and NF-κB1–deficient mice to further investigate how the response to unloading affects the numbers of osteoprogenitors and the amount of osteoclastogenesis. BMP2 and Wnt3a expression did not differ between WT or NF-κB1–deficient osteoblasts and osteocytes even after unloading (Supplemental Data 2). In both WT and NF-κB1–deficient osteocytes, BMP4 expression was reduced after unloading and the level of reduction was similar between the groups (Supplemental Data 2). The expression of Wnt5a increased by approximately twofold in WT osteoblasts but not in NF-κB1–deficient osteoblasts. The expression of Wnt5a was slightly but not significantly increased in WT osteocytes. In contrast, Wnt5a expression remained unchanged in NF-κB1–deficient osteocytes (Supplemental Data 2). Additional studies are necessary to determine the role of p50 upregulation in osteocytes during unloading and unloading-induced osteoporosis.
Two weeks of functional disuse significantly reduced the metaphyseal bone fraction, predominantly via a decrease in the trabecular number, thereby reducing its connectivity and augmenting the marrow space. Therefore, bone formation is the critical activity to determine the level of bone loss due to unloading and is a major target for elucidating the mechanisms required for unloading-induced losses of bone mass. Dynamic histomorphometric analyses of the osteoblastic cell activity in vivo revealed that the in vivo reduction in bone formation activity due to unloading was suppressed by a total lack of NF-κB1. Furthermore, bone cell culture experiments using the bone marrow cells taken from the animals indicated that NF-κB1 deficiency suppressed the unloading-induced reduction in ALP activity. The interpretation of this reduction in ALP activity after 1 week in culture would be that the progenitor cell populations for the osteoblastic cell lineage could be reduced at the point of preparing the cells from the animals at the end of unloading. This suppression was blocked in NF-κB1–deficient cells. To characterize further the cells from the WT or NF-κB1–deficient mice that responded to β-glycerophosphate and ascorbic acid after unloading, we performed a CFU-F assay. The bone marrow cells formed colonies composed of fibroblastic cells, known as CFU-Fs, when the cells were cultured at a low density. Although the CFU-Fs formed various types of cells, including fibroblasts, osteoblasts, and endothelial cells, it has been suggested that CFU-Fs expressing a high level of ALP activity are the committed progenitors for osteogenesis. In the present study, unloading significantly reduced the populations of ALP-positive colonies from the WT mice. Although NF-κB1 deficiency slightly increased the number of ALP-positive colonies at baseline, the reduction in ALP-positive colonies was smaller than that of the WT cells. These results suggest that NF-κB1 deficiency maintained the proportion and/or potential for osteoprogenitors or immature osteoblasts to differentiate into mature osteoblasts during unloading.
Unloading-induced enhancement in osteoclastic bone resorption by increasing number of osteoclasts was not observed in the NF-κB1–deficient mice. Although unloading increased RANKL expression, and the ratio of RANKL/OPG was also similar for both the WT and NF-κB1–deficient mice. We next examined whether the population of osteoclast precursors increased in the bone marrow of WT mice during unloading. Unloading slightly increased the population of osteoclast precursors (c-kit+, c-fms+, RANK+) of the total cells in the WT mice. Although the basal population of osteoclast precursors increased in the NF-κB1–deficient mice, unloading did not alter the population in the mice, suggesting that an increased population of osteoclast precursors during unloading partially contributes to unloading-induced osteoclastogenesis in WT mice. Another possible mechanism in the unloading-induced stimulation of bone resorption in the WT mice appears to be the enhanced intracellular signaling of RANKL. Unloading enhanced the c-Fos and NFATc1 expression induced by RANKL in the WT cells but not in the NF-κB1–deficient cells. Although the molecular mechanisms by which NF-κB1 deficiency could suppress unloading-induced c-Fos and NFATc1 expression are unknown, these results indicate that NF-κB1–deficient mice prevent unloading-induced osteoclastogenesis in vivo by suppressing the enhancement of RANKL intracellular signaling on osteoclast precursors and by increasing the levels of osteoclast precursors during unloading.
In conclusion, an NF-κB1 deficiency results in the loss of mechanosensitivity in bone and suppresses unloading-induced reductions in bone formation and the enhancement of bone resorption. Our results also indicate that the NF-κB1 p50 subunit might represent a useful target molecule for preventing unloading-induced bone loss.
All authors state that they have no conflicts of interest.
This work was supported by grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan (23390424 to EJ) and the “Ground-based Research Program for Space Utilization” promoted by the Japan Space Forum (to EJ).
Authors' roles: HN, WM, HF, and K Osawa performed the research. KA, NA, KN, and K Ohya performed the radiological assessments, histology preparation, and bone histomorphometry. HY provided the GST-RANKL. YMT provided the technical support for the preparation of the osteoblasts and osteocytes. KM and IN reviewed the intermediate draft. EJ designed the study, performed the literature review, prepared the initial and final versions of the paper, and submitted the document.