Maintenance of skeletal integrity requires a balance between bone formation and bone resorption.1 This coordinated process, known as bone turnover, is regulated by multiple factors, including some that are secreted from antigen-stimulated immune cells.2 This interaction between the immune and skeletal systems is believed to play an important role in maintaining skeletal homeostasis through the lifespan.2, 3
Among the cytokines that regulate bone turnover, interferon γ (IFN-γ) has been shown to play an important role in the regulation of osteoclastogenesis.4 In vitro, IFN-γ has a marked inhibitory effect on osteoclast formation in receptor activator of nuclear factor-κB ligand (RANKL)–stimulated bone marrow monocyte precursors.5 However, the role of IFN-γ in vivo is more complex because it was shown to either decrease osteoclastic bone resorption, leading to an improvement of bone mass,6 or to increase osteoclastic bone resorption, leading to a decrease in bone mass,7 depending on the experimental model and conditions used.8
In contrast to the many studies on the role of IFN-γ in osteoclasts, studies exploring the role of IFN-γ in bone formation both in vitro and in vivo still are scarce. We recently reported that IFN-γ is required for the osteogenic differentiation of mesenchymal stem cells (MSCs) in vitro and for the maintenance of bone mineral density (BMD) in vivo,9 suggesting a potential role of IFN-γ in differentiation and bone formation in vivo. Another report by Rifa10 demonstrated that T-cell cytokines, including IFN-γ, induce bone morphogenic protein 2 (BMP-2) in MSCs and therefore may regulate their differentiation and mineralization.
Based on this evidence, we explored the physiologic role of IFN-γ on bone turnover following ablation of its receptor (IFNγR1−/− mice) as well as the effect of its exogenous administration in sham-operated (SHAM) and ovariectomized (OVX) animals. Our results demonstrate that disruption of IFN-γ signaling leads to a decrease in bone mass associated with a concomitant reduction in markers of osteoblast differentiation and bone formation. In contrast, administration of IFN-γ to SHAM mice increased bone turnover in favor of bone formation, leading to improved bone mass, architecture, and mechanical properties. In addition, IFN-γ administration rescued OVX mice from osteoporosis. Our data therefore strongly support a physiologic role of IFN-γ for maintenance of skeletal integrity and suggest that modulation of its signaling pathway may be used advantageously to improve bone strength.
Materials and Methods
We purchased IFNγR1−/− and IFNγR1+/– mice (strain name 129-Ifngrtm1 on a C57BL/6 background; n = 8/group) from Jackson Laboratories (Bar Harbor, ME, USA). As a control strain, we purchased strain name 129S1/SvImJ mice on a C57BL/6 background from the same source (n = 8). In addition, we used 8-week-old virgin female C57BL/6 mice (Charles River Laboratories, Quebec, Canada) to study the effect of exogenous administration of IFN-γ on bone strength. Bilateral ovariectomies (OVX) were performed under general anesthesia. Another group of animals was sham-operated (SHAM), in which ovaries were exteriorized but replaced intact. Two weeks after surgery, mice (OVX and SHAM) received intraperitoneal injections of either 2000 or 10,000 IU of IFN-γ (R&D Systems, Inc., Minneapolis, MN, USA) or vehicle (PBS) three times weekly for a total of 6 weeks. Mice were housed in cages in a limited-access room. Animal husbandry adhered to Canadian Council on Animal Care Standards, and all protocols were approved by the McGill University Health Center Animal Care Utilization Committee.
3D micro–computed tomography (3D µCT) of bone samples
Tibias and spines obtained from IFNγR1−/−, IFNγR1+/–, and IFNγR1+/+ mice were scanned by 3D µCT at ×40 magnification with a SkyScan 1072 (Skyscan, Aartselaar, Belgium) and analyzed using bone-analysis software (Version 2.2f, Skyscan). Analyses of the trabecular bone were carried out in a 2.3-mm-thick region for tibias and a 1.55-mm-thick region for femurs distal to the growth plate Femoral cortical thickness was measured on the three-hundredth cross section (3.375 mm) from the distal growth plate. For spine, we used L2 from both groups; the entire vertebra was analyzed. Parameters were acquired with a rotation of 0.9 degree between each picture with the X-ray source set at 100 kV and 98 µA. The segmentation of the image was made by a global threshold and a voxel size of 21.90 × 21.90 × 21.90 µm; the same threshold setting was used for all the samples. Architectural measurements were made as described previously.11 Similar 3D µCT scan analysis was performed in tibias obtained from SHAM and OVX mice treated with either IFN-γ or vehicle alone. Nomenclature and abbreviations of 3D µCT parameters follow the recommendations of the American Society of Bone and Mineral Research.12
Histologic and histomorphometric analysis of bone
The details of these methods were described previously.11 For dynamic histomorphometry, tetracycline labeling was achieved through the intraperitoneal injection of demeclocycline (20 mg/kg; Sigma Chemical Co., St Louis, MO, USA) to IFNγR1−/− and IFNγR1+/+ mice at 5 and 2 days before euthanization. One side femur from each animal in each group was removed at the time of euthanization, fixed in 70% ethanol, dehydrated, and embedded undecalcified in methyl methacrylate (J-T Baker, Phillipsburg, NJ, USA). At 50-µm intervals, longitudinal sections 5- and 8-µm thick were cut using a Polycut-E microtome (Reichert-Jung Leica, Heerbrugg, Switzerland), placed on gelatin-coated glass slides, deplasticized, and stained with Goldner's trichrome. Histomorphometry was done with a semiautomatic image-analyzing system combining a microscope equipped with a camera lucida and digitizing tablet linked to a computer using the OsteoMeasure Software (Osteometrics, Inc., Decatur, GA, USA). Nomenclature and abbreviations of histomorphometric parameters follow the recommendations of the American Society of Bone and Mineral Research.13 Similar histomorphometry analysis was performed in distal femurs obtained from SHAM and OVX mice treated with either IFN-γ or vehicle alone.
Histochemistry, immunohistochemistry, and immunofluorescence
Detection of alkaline phosphatase (ALP) and tartrate-resistant acid phosphatase (TRACP) activity was carried out as described previously.11 Naphtol-AS-TR (Sigma-Aldrich Canada, Ltd., Oakville, Ontario, Canada) was used as substrate for both enzymes; Fast Blue BB salt (Sigma-Aldrich Canada, Ltd.) was used as a coupler for ALP. For immunohistochemistry, sections were incubated overnight at 4°C with a goat polyclonal antibody IgG against Runx2 (anti-mouse Runx2 C-19; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), osteopontin (OPN; anti-mouse osteopontin sc-10593; Santa Cruz Biotechnology, Inc.), or RANKL (anti-mouse RANK-ligand C-20; Santa Cruz Biotechnology, Inc.). Primary antibody was detected by incubation with an anti-goat IgG secondary antibody conjugated with horseradish peroxidase (1:300 in bovine serum albumin 1%; Sigma-Aldrich). Antibody complexes were visualized with DAB, a 3,3-diaminobenzine solution containing hydrogen peroxide (Zymed Laboratories, Inc., San Francisco, CA, USA) and then counterstained in 1% hematoxylin.
For immunofluorescence, sections were incubated for 1 hour at room temperature with a rat monoclonal antibody IgG against CD3 (anti-mouse CD3, sc-70614; Santa Cruz Biotechnology, Inc.). Primary antibody was detected by incubation with a fluorescein isothiocyanate (FITC)–conjugated goat anti-rat IgG secondary antibody (1:500 in bovine serum albumin 1%). Following washing with PBS, the slides were covered with a coverslip and mounted with 90% glycerol. Photographs were taken under an Olympus fluorescence microscope (North Ryde, NSW, Australia) controlled by an IPLab system (BioVision Technologies, Exton, PA, USA). Brightness, overlap, and contrast adjustments were performed in Photoshop (Adobe Systems Inc., San Jose, CA, USA).
Bone mass measurements by dual-energy X-ray absorptiometry (DXA)
Hip and spine bone mineral density (BMD) were measured in SHAM and OVX mice treated with either IFN-γ or vehicle alone using a PIXIMUS bone densitometer (GE Medical Systems, Schenectady, NY, USA). A quality-control phantom was used to calibrate the densitometer prior to each experiment.
Markers of bone turnover
IFNγR1−/− and IFNγR1+/+ mice were euthanized at 4 and 8 weeks of age, and blood was removed by cardiac puncture. Osteocalcin (OCN) was measured in 20 µL of serum using the Mouse OCN Immunoradiometric Assay Kit (Immutopics, San Clemente, CA, USA). N-Telopeptide was measured in 20 µL of serum to assess osteoclastic activity using the RAT Telopeptide Kit (Osteometer Biotech, Hovedgade, Denmark).
Western blot analysis
Bone marrow cells were obtained from the left femurs of 4-week-old IFNγR1−/− and IFNγR1+/+ mice (n = 8/group) by flushing with Dubelcco's α − modified medium (DMEM). Red blood cells in marrow cells were hemolyzed in 0.017 M tris-HCl (pH 7.5) buffer containing 0.8% ammonium chloride. Hemolyzed bone marrow suspensions were rinsed twice with PBS. Protein extracts were obtained after suspending the cells in 2 volumes of buffer containing 10 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, and one Complete protease inhibitor mixture tablet (Boehringer Mannheim, Laval, Quebec, Canada). Solutions then were centrifuged at 25,000g for 20 minutes at 4°C and resuspended in 20 mM HEPES (pH 7.9), 25% glycerol, 0.42 M NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, and 0.5 mM dithiothreitol. Protein content was determined with a protein assay kit (Bio-Rad, Mississauga, Ontario, Canada), and samples then were aliquoted and stored at −80°C. For Western blot analysis, lysates were dissolved in SDS electrophoresis buffer (Bio-Rad, Hercules, CA, USA), and proteins were separated on SDS-polyacrylamide gels and subsequently electrotransferred to polyvinylidene difluoride membranes. After blocking with PBS containing 0.1% Tween 20 and 10% nonfat dry milk, membranes were incubated overnight at 4°C using mouse monoclonal antibodies directed against Runx2, OPN, and RANKL (1:1000; Santa Cruz Biotechnology, Inc.). Secondary antibodies conjugated to horseradish peroxidase were from Sigma Chemical Co. (1:5000). Antigen-antibody complexes were detected by chemiluminescence and blots were exposed using an Amersham-enhanced chemiluminescence (ECL) Plus Western Blotting Detection kit (GE Healthcare, Rydalmere, NSW, Australia). Films were scanned, and the optical density (OD) of each specific band was analyzed using the ImageMaster program (GE Healthcare) and expressed as OD/mm2/100 µg of total protein. Relative intensity of the samples was determined comparing the protein of interest in the mutant mice using the values of wild-type mice as controls (100%). Values are reported as the average of samples obtained from 8 mice.
Three-point bending mechanical test of bone
A three-point breaking test14 was performed on the midshaft of the tibias obtained from both SHAM and OVX mice treated with either IFN-γ or vehicle alone. A commercial bench-mounted vertical tensile/compression tester, the Instron 5569 (Instron Corp., Canton, MA, USA), was used. The extrinsic parameters [ultimate force (Fult), stiffness, work to failure, and Young modulus] were extracted from a force-displacement curve. Stiffness (K) is a measure of resistance of bone to displacement; Young's modulus (E) is a measure of stiffness related to the shape of the object; Fult (newtons) is the maximum force that bone can resist; work to failure (U) is the energy required to break bone. The span of two support points was 10 mm, and the deformation rate was 1.0 mm/min.
All results were expressed as the mean ± SE. Differences in the structural and static parameters of bone histomorphometry between groups of mice were determined using Levene's test for homogeneity of variances and the unpaired t test for equality of means. In all experiments, a value of p < .05 was considered significant.
Consequences of IFNγR1 disruption on bone mass, bone microarchitecture, and bone turnover in vivo
We examined microarchitectural and morphometric changes in IFNγR1−/− and IFNγR1+/– mice compared with wild-type controls (IFNγR1+/+ mice). Measurement of structural parameters by 3D µCT analyses demonstrated a 45% reduction in bone volume (BV/TV) as well as a significant reduction in other structural parameters (eg, trabecular thickness and number and cortical thickness) in IFNγR1−/− mice compared with IFNγR1+/+ mice (p < .01; Fig. 1A and Table 1). No differences between heterozygous and wild-type mice were found (data not shown). We then performed histomorphometric analysis to further examine the changes in bone architecture. Similar to the changes observed by 3D µCT, significant deterioration in the structural parameters of bone was observed in IFNγR1−/− animals (Fig. 1B and Table 1). In addition, dynamic histomorphometry showed a significant reduction of mineralized surfaces, bone-formation rate (BFR), and mineral-apposition rate (MAR) in 4-week-old IFNγR1−/− mice compared with IFNγR1+/+ mice at 4 weeks of age (Fig. 1C and Table 1). Finally, the mean number of osteocytes per cortical bone at the metaphyseal level was significantly lower in the IFNγR1−/− mice compared with IFNγR1+/+ mice (Table 1).
Table 1. Microarchitectural Parameters and Measurement of Cellular Activities in Proximal Tibias Assessed by Histomorphometry and 3D µCT at 8 Weeks of Age
Histology (n = 8)
µCT (n = 8)
Mean ± SEM
Mean ± SEM
Mean ± SEM
Mean ± SEM
BV/TV = bone volume; Tb.Th = trabecular thickness; Tb.Sp = trabecular separation; Tb.N = trabecular number; Ct.Wi = cortical width; BS/TV = bone surface per tissue area; MAR = mineral apposition rate; BFR/BS = bone-formation rate per bone surface; N.Ob/T.Am = osteoblast number; ALP/BS = alkaline phosphatase per bone surface; N.Oc/E.Pm = osteoclast number; N/A = nonapplicable; N.Ot/Ct.Ar = osteocyte number (lacuna with or without cell) per cortical area; NS = nonsignificant.
9 ± 0.7
5.8 ± 0.4
8.6 ± 0.4
4.8 ± 0.2
0.25 ± 0.11
0.19 ± 0.01
0.12 ± 0.01
0.9 ± 0.02
0.25 ± 0.01
0.32 ± 0.02
0.25 ± 0.01
0.43 ± 0.02
3.6 ± 0.1
3 ± 0.2
2.4 ± 0.3
1.4 ± 0.2
0.22 ± 0.02
0.18 ± 0.03
0.26 ± 0.01
0.15 ± 0.02
6 ± 4
7.2 ± 4
13 ± 1.9
7.2 ± 1.2
4.5 ± 0.5
2.2 ± 0.2
BFR/BS (mm2 × 103/mm/d)
4.3 ± 0.3
2.3 ± 0.4
22 ± 3
9 ± 3
32.8 ± 3.5
19 ± 3
762 ± 20
644 ± 27
We then identified Runx2 expression, a marker of early osteoblast differentiation, in the distal femur by immunohistochemistry (Fig. 2A) and Western blot (Fig. 2C). We observed a significant reduction of Runx2 expression in bone marrow cells of IFNγR1−/− mice compared with IFNγR1+/+ controls (p < .05; Fig. 2C). We next assessed OPN expression, a marker of early mineralization, and demonstrated a marked reduction of its expression within the trabeculae and in cortical bone (Fig. 2B) of IFNγR1−/− mice compared with IFNγR1+/+ mice. The reduction in OPN protein expression in bone observed by immunohistochemistry in IFNγR1−/− mice was confirmed by Western blot (p < .05; Fig. 2C). Finally, number of ALP-expressing cells (osteoblasts), a marker of late differentiation, was significantly reduced in IFNγR1−/− mice compared with IFNγR1+/+ mice (Table 1). These immunohistochemical changes were associated with lower circulating concentrations of the bone-formation marker OCN compared with IFNγR1+/+ controls (Fig. 2D). Subsequently, we analyzed parameters of bone resorption and osteoclastic activity by bone histomorphometry, immunohistochemistry, and circulating concentration of bone markers. The expression of RANKL by immunostaining, a regulator of osteoclastic differentiation and activity at the cell membrane of osteoblast precursors, was significantly reduced in bone marrow cells of IFNγR1−/− compared with IFNγR1+/+ mice (p < .05; Fig. 3A, B). Histomorphometry showed a significant reduction in osteoclast number (N.Oc/E.Pm; Table 1) and eroded surfaces (ES/BS; Fig. 3C) associated with a significant reduction in circulating concentrations of the bone resorption marker N-telopeptide in IFNγR1−/− mice compared with the IFNγR1+/+ controls (Fig. 3D).
Effect of exogenous administration IFN-γ on resident T-cell population, bone strength, and bone cellularity in SHAM mice and OVX-induced osteoporosis
To assess the effect of OVX and IFN-γ treatment on resident T-cell populations, we investigated the changes in CD3+ cells within the bone marrow of SHAM and OVX mice treated with either vehicle or IFN-γ (2000 and 10,000 IU). OVX mice showed a reduction in bone marrow expression of CD3+ cells. This effect of OVX on CD3+ cells was reverted by treatment with IFN-γ (2000 and 10,000 IU; Fig. 4). In addition, treatment of SHAM mice with IFN-γ (2000 and 10,000 IU) induced higher levels of CD3+-expressing cells corresponding to a higher level of resident T-cell population (Fig. 4). Finally, the level of CD3+ cells was similar in OVX and SHAM mice treated with 10,000 IU of IFN-γ.
Furthermore, we examined BMD, bone microarchitecture, histomorphometric changes, and mechanical properties of bone after 6 weeks of treatment with either IFN-γ (2000 and 10,000 IU subcutaneously three times per week) or vehicle to 10-week old SHAM and OVX C57BL/6 mice. Treatment with 10,000 IU of IFN-γ to the SHAM group induced a significantly higher bone mass than in SHAM animals treated with vehicle at the lumbar spine (0.070 ± 0.004 g/cm2 versus 0.058 ± 0.002 g/cm2; Fig. 5A). OVX induced a significant reduction in BMD at both the lumbar spine (approximately –17%) and femur (approximately –25%) compared with untreated SHAM controls (p < .01). Treatment of OVX mice with IFN-γ (2000 and 10,000 IU) corrected this bone loss at the lumbar spine (Fig. 5A). At the femoral level (Fig. 5B), IFN-γ treatment (2000 and 10,000 IU) induced a significant increase in BMD compared with vehicle-treated OVX animals but did not reach the levels of BMD found in the SHAM groups (p < .01; Fig. 5B). Furthermore, 3D µCT analysis (Fig. 5C, D, and Table 2) demonstrated a significant increase in bone volume (BV/TV), trabecular thickness, trabecular number, and cortical width in SHAM and OVX animals treated with IFN-γ compared with SHAM and OVX animals treated with vehicle (Fig. 5D and Table 2). We next performed mechanical testing in the same groups of animals (Fig. 5E). A significant increase in all mechanical indices was observed in both SHAM and OVX animals treated with IFN-γ (2000 and 10,000 IU) compared with OVX vehicle-treated animals (Fig. 5E).
Table 2. Microarchitectural Parameters and Measurements of Cellular Activity in Femurs Assessed by Histomorphometry and 3D µCT at 28 Weeks of Age in OVX C57BL/6 Mice Treated With IFN-γ (n = 8 per Group)
SHAM + IFN-γ (2000 IU)
SHAM + IFN-γ (10,000 IU)
OVX + PBS
OVX + IFN-γ (2000 IU)
OVX + IFN-γ (10,000 IU)
Mean ± SEM
Mean ± SEM
Mean ± SEM
Mean ± SEM
Mean ± SEM
Mean ± SEM
BV/TV = bone volume; Tb.Th = trabecular thickness; Tb.Sp = trabecular separation; Tb.N = trabecular number; Ct.Wi = cortical width; BS/TV = bone surface per tissue area; N.Ob/T.Am = osteoblast number; ALP/BS = alkaline phosphatase per bone surface; N.Oc/E.Pm = osteoclast number; Max. cortical MAR = maximal cortical apposition rate; Max trabecular MAR = maximal trabecular mineral apposition rate.
Subsequently, we determined the changes in bone volume to total volume (BV/TV) ratio, rate of bone turnover, and bone cell numbers in the different groups of treated mice. Figure 6 shows sections of undecalcified bone stained with Von Kossa for histomorphometry (Fig. 6A), ALP to identify osteoblasts (purple; Fig. 6B), TRACP to identify osteoclasts (red; Fig. 6C) and tetracycline labeling for dynamic histomorphometry (Fig. 6D). Von Kossa staining shows a significant and dose-dependent increase in BV/TV in the SHAM and OVX animals treated with IFN-γ (2000 and 10,000 IU) compared with vehicle-treated animals (p < .01; Table 2). Quantification of cell numbers after normalization with bone surface (Table 2) revealed a significant and dose-dependent increase in osteoblast and osteoclast numbers in both SHAM and OVX animals treated with IFN-γ compared with animals treated with vehicle alone (p < .01; Table 2). In the case of SHAM animals, treatment with 2000 IU of IFN-γ induced an increase of approximately 0.5-fold in osteoblast number and a 3-fold increase in osteoclast number (p < .001). The increase in bone cellularity in SHAM mice was significantly higher in mice treated with 10,000 IU compared with 2000 IU, with an approximately 15-fold increase in osteoblast number and approximately 30-fold increase in osteoclast number (p < .001). OVX induced a significant change in osteoclast number (approximately 2-fold) compared with SHAM animals treated with vehicle. Treatment with 2000 IU of IFN-γ to OVX mice induced an increase of approximately 2.5-fold in osteoblast number and approximately 2-fold in osteoclast number compared with vehicle-treated OVX mice. Furthermore, treatment of OVX mice with 10,000 IU of IFN-γ induced a significant increase in osteoclast (approximately 3-fold) and osteoblast (approximately 13-fold) compared with vehicle-treated OVX mice. Finally, dynamic histomorphometry shows a significantly higher mineral-apposition rate, both trabecular and cortical, in all the IFN-γ-treated animals compared with vehicle-treated mice (Fig. 6D and Table 2).
Recent advances in osteoimmunology have suggested that there is a close interplay between the immune and skeletal systems.2, 4 However, most of these advances have focused on the interaction between the immune system and the bone-resorpting osteoclasts.3 In contrast, there is limited information looking at the relationship between the immune system and the osteoblasts.9, 10, 15 Considering that aging is associated with a significant reduction in osteoblastogenesis and bone turnover,16 identification of the potential role of the immune system in this decline is essential. Among the cytokines that have been found to regulate osteoclastogenesis, IFN-γ seems to be a critical regulator of bone resorption.7, 17, 18 Although the role of IFN-γ in osteoclast differentiation and activity has been widely studied,7, 17 its role in osteoblast differentiation and function remains poorly understood.8
In this study we have assessed the role of IFN-γ in bone turnover. We used two different approaches. First, we assessed the changes in bone quality and turnover in a mouse model in which the IFN-γ R1 had been disrupted by homologous recombination, the IFNγR1−/− mouse. These animals are normal at birth, are fertile, have a normal growth rate and body weight, and cannot be differentiated from their heterozygous (IFNγR1+/–) and wild-type (IFNγR1+/+) controls, except for subtle changes in the immune system.19, 20. We have reported previously that disruption of IFN-γ signaling results in lower osteoblast differentiation of MSCs and a significant and sustained decrease in bone mass over time, suggesting that bone mass acquisition soon after birth cannot proceed normally in the absence of IFN-γ.9 In this study we have found that this defect in bone acquisition is due to low levels of bone turnover owing to a predominant reduction in bone formation. In addition, we examined the effect of IFN-γ in a model of osteoporosis used previously to test the efficacy of various therapeutic agents in osteoporosis, including parathyroid hormone (PTH) and bisphosphonates,21, 22 and their SHAM controls. Treatment with IFN-γ in both SHAM and OVX mice led to improved microarchitecture and strength compared with vehicle-treated animals.
Although one previous study has reported an effect of IFN-γ on bone formation through the induction of bone morpgogenetic protein 2 (BMP-2),10 there is indirect evidence linking IFN-γ with osteoblastogenesis. Disruption of nitric oxide (NO) signaling in mice results in defective bone formation.23, 24. Because IFN-γ, together with interleukin 1 (IL-1), stimulates NO production to high levels in bone,25, 26 these earlier studies support a role for IFN-γ in bone formation. Furthermore, IFN-γ induces expression of Best5, a gene expressed during bone formation in rats.27
In contrast, another study supports a negative effect of IFN-γ on bone formation. Stat1 disruption, a checkpoint of the IFN-γ activated pathway, has been associated with an increased bone mass owing to a marked increase in bone formation.28 However, it has been suggested that this effect may be related to the role of Stat1 as a cytoplasmic attenuator of Runx2,29, 30 a process that is independent of the presence of IFN-γ.
Our findings directly demonstrate that disruption of IFN-γ signaling affects bone remodeling through a reduction in bone formation, extending our earlier studies demonstrating reduced osteoblast differentiation of bone marrow precursors in this model.9 Assessment of bone architecture with 3D µCT analysis, histology, and histomorphometry demonstrated that IFNγR1−/− mice had a significant deficit in both trabecular and cortical bones characteristic of osteoporosis.
We then attempted to shed light on the mechanism underlying the bone volume reduction in IFNγR1−/− mice. Our in vivo experiments demonstrated that disruption of IFN-γ signaling in IFNγR1−/− mice reduced the expression of the osteogenic factor Runx2 as well as a reduction in OPN expression in trabecular and cortical bones, the presence of persistently lower concentrations of serum OCN, and a decrease in the quantity of bone surfaces stained for ALP activity in IFNγR1−/− mice. In addition, the reduction in the number of mature osteoblasts and thus osteocytes in IFNγR1−/− mice also was associated with a subsequent decrease in the formation of trabeculae. These findings are further supported by the reduction in mineral-apposition and bone-formation rates observed in IFNγR1−/− mice.
As reported previously,7, 17 we also observed a significant effect of IFN-γ on osteoclastic activity. In our study, disruption of IFN-γ signaling resulted in a strong and sustained inhibition of markers of osteoclastic activity. In addition, IFNγR1−/− mice expressed lower levels of RANKL than their IFNγR1+/+ counterparts. Our results differ from a previous study looking at osteoclast differentiation in a model of inflammatory bone loss in IFNγR1−/− mice. In this model, osteoclastic differentiation was significantly higher than in controls.31 This apparent discrepancy may be related to the presence of surrounding activated T cells, which are known to be central to osteoclast activation and bone loss,8 in contrast with our model, in which RANKL modulation by IFN-γ could be the primary mechanism of action in a normal physiologic environment. However, it remains to be established whether IFN-γ modulates primarily osteoclastic activity indirectly through RANKL expression by osteoblasts and/or has a direct effect on osteoclasts. In our study, the overall effect of IFN-γ disruption was a very significant reduction in bone volume, indicating that deficiency in IFN-γ results in low-turnover osteoporosis with a concomitant reduction in bone formation and osteoclastic activity.
Subsequently, we examined the effect of exogenous administration of IFN-γ in a model of osteoporosis. As reported previously in this model,22 a very significant and rapid decrease in BMD was observed following OVX compared with SHAM animals, particularly at the lumbar spine. In contrast with previous studies in which OVX was performed in transgenic models and inflammatory conditions,7, 33 in this study we used a normal model of OVX in which either inflammation or high levels of activated T cells are not expected to play a role. Nevertheless, we characterized the response of T cells to administration of exogenous IFN-γ in our OVX and SHAM models. Previous studies looking at the effect of OVX on IFN-γ levels and T-lymphocyte activity have shown contradictory results.34 For instance, a previous study has reported an increase in CD4+ T cells expressing IFN-γ and in the IFN-γ concentration in culture supernatants of bone marrow CD90+ T cells after ovariectomy, concluding that this increase is critical in explaining the bone effects in estrogen deficiency.35 However, opposing data indicate that estrogens upregulate and gonadectomy downregulates IFN-γ production,36, 37 a phenomenon that also has been reported in postmenopausal women.38, 39 Considering that these data suggest a difference in the regulation of IFN-γ production by T cells from peripheral tissues (ie, spleen and lymph node) versus bone marrow T cells after gonadectomy, we selected CD3+ cells as a more specific marker of T-cell activation in mouse bone marrow after treatment with IFN-γ.40 Compared with untreated SHAM mice, OVX induced lower levels of CD3+ cells within the bone marrow. In contrast, treatment with exogenous IFN-γ induced a similar stimulatory response in the levels of CD3+-expressing cells in both SHAM and OVX groups in a dose-dependent manner. These data demonstrate that exogenous administration of IFN-γ exerts a similar effect on T cells from both OVX and SHAM mice; thus its effect on bone cells seems to be independent of the number of activated T cells present in the bone marrow.
In our model, administration of IFN-γ significantly increased BMD in SHAM animals at the lumbar spine in a dose-dependent manner. In addition, IFN-γ rescued OVX mice from osteoporosis in both the lumbar spine and femur. However, since BMD is only one component of bone strength and does not always correlate with the effect of therapy on fracture reduction,41 we carefully assessed other components of bone microarchitecture and strength. Both trabecular and cortical indices assessed by 3D µCT significantly improved with IFN-γ in a dose-dependent manner. This beneficial effect was found in both SHAM and OVX mice. Furthermore, the beneficial effect of IFN-γ on cortical bone architecture was further demonstrated by an increase in cortical mineral apposition and improvement in biomechanical properties of bone using three-point bending in the midshaft of tibias. In addition, this beneficial effect of IFN-γ on the bone mass of osteoporotic mice also was associated with an increase in both osteoblast and osteoclast number.
The changes in cellularity and labeled surfaces observed in IFN-γ-treated mice suggest that the anabolic effect and the potential activation of bone remodeling are not just coupled phenomena. When compared with proposed models of PTH action on bone,42 changes in tetracycline labeling suggest that despite high levels of bone resorption associated with higher osteoclast number, intermittent administration of IFN-γ also could induce bone formation in previously quiescent surfaces, especially in SHAM mice, where the tetracycline labeling shows high levels of bone formation within just 6 weeks of treatment in a context of lower levels of bone resorption compared with OVX animals, in which high levels of bone resorption are a constant. In contrast, in the OVX animals, the differences in osteoclast and osteoblast numbers induced by IFN-γ suggest that, as in PTH-treated OVX animals,22 there could be an initial osteoclast activation that, owing to effective coupling of osteoblasts to osteoclasts in the remodeling unit, intensely recruits osteoblasts that further benefit from the direct stimulatory actions of IFN-γ.
Although our results clearly indicate that IFN-γ plays an important role in bone formation and bone remodeling, its usefulness in osteoporotic patients needs further study. Based on our data and similar to PTH,43 the potential usefulness of IFN-γ in osteoporosis and the changes induced by IFN-γ in bone cells seem to be dose-dependent. Here we demonstrated that intermittent administration of low and moderate doses of IFN-γ (1 and 5 × 105 IU/kg, respectively) are anabolic to bone. In contrast, two previous studies with higher doses of IFN-γ reported a predominant effect on osteoclastic bone resorption, with its systemic administration resulting in bone loss.44, 45 In these studies, short-term administration of recombinant IFN-γ at a dose of 1 × 106 IU/kg (approximately 20,000 to 30,000 units/animal) once a day for 8 days induced a significant decrease in BMD compared with vehicle-treated animals. Indeed, IFN-γ has been used successfully in osteopetrosis to restore osteoclastic activity, and its administration at high dose (1.5 µg/kg of body weight per dose three times per week for at least 6 months) is currently approved to ameliorate this disease.46
Taken together, it would appear that the beneficial effect of lower doses of IFN-γ on bone strength demonstrated here relates to its effect on bone formation, which predominates over its effect on bone resorption. Further studies aimed at determining the optimal anabolic response of IFN-γ on bone while minimizing its potential side effects should be pursued. An attractive strategy would be to modulate IFN-γ signaling at critical checkpoints to investigate the potential therapeutic usefulness of this new anabolic pathway.
In summary, our data clearly indicate that disruption of IFN-γ signaling in mice results in an osteoporotic phenotype with low bone turnover and that administration of IFN-γ increases bone mass and strength in both SHAM and OVX mice through the induction of bone formation in the former and by rescuing the latter from osteoporosis. Further studies are required to examine the therapeutic potential of targeting this novel signaling pathway in human osteoporosis.
The authors state that they have no commercial affiliations, consultancies, stock, or equity interests and patent-licensing arrangements that could be considered conflicts of interest.
GD holds a fellowship from The University of Sydney Medical Research Foundation. RK holds a Chercheur National Award from the Fond de la Recherche en Santé du Québec. This study was supported by the Canadian Institutes for Health Research (CIHR MOP 10839), the Australian National Health and Medical Research Council (NHMRC 632767 to GD), the Dairy Farmers of Canada, and NSERC (to RK). We thank Pat Hales for preparation of the manuscript.