Oxidative stress and gamma radiation-induced cancellous bone loss with musculoskeletal disuse

Abstract

Exposure of astronauts in space to radiation during weightlessness may contribute to subsequent bone loss. Gamma irradiation of postpubertal mice rapidly increases the number of bone-resorbing osteoclasts and causes bone loss in cancellous tissue; similar changes occur in skeletal diseases associated with oxidative stress. Therefore, we hypothesized that increased oxidative stress mediates radiation-induced bone loss and that musculoskeletal disuse changes the sensitivity of cancellous tissue to radiation exposure. Musculoskeletal disuse by hindlimb unloading (1 or 2 wk) or total body gamma irradiation (1 or 2 Gy of ¹³⁷Cs) of 4-mo-old, male C57BL/6 mice each decreased cancellous bone volume fraction in the proximal tibiae and lumbar vertebrae. The extent of radiation-induced acute cancellous bone loss in tibiae and lumbar vertebrae was similar in normally loaded and hindlimb-unloaded mice. Similarly, osteoclast surface in the tibiae increased 46% as a result of irradiation, 47% as a result of hindlimb unloading, and 64% as a result of irradiation + hindlimb unloading compared with normally loaded mice. Irradiation, but not hindlimb unloading, reduced viability and increased apoptosis of marrow cells and caused oxidative damage to lipids within mineralized tissue. Irradiation also stimulated generation of reactive oxygen species in marrow cells. Furthermore, injection of α-lipoic acid, an antioxidant, mitigated the acute bone loss caused by irradiation. Together, these results showed that disuse and gamma irradiation, alone or in combination, caused a similar degree of acute cancellous bone loss and shared a common cellular mechanism of increased bone resorption. Furthermore, irradiation, but not disuse, may increase the number of osteoclasts and the extent of acute bone loss via increased reactive oxygen species production and ensuing oxidative damage, implying different molecular mechanisms. The finding that α-lipoic acid protected cancellous tissue from the detrimental effects of irradiation has potential relevance to astronauts and radiotherapy patients.

space is a unique environment that challenges the skeletal health of astronauts. Long-duration spaceflight causes a negative calcium balance and a decline in bone density selectively in tissues that are normally loaded on Earth, posing an increased risk for fracture (45). On entry into space, musculoskeletal disuse and a cephalad fluid shift in microgravity trigger rapid physiological adaptations (26). In addition to the risks imposed by reduced gravity, astronauts are exposed to radiation at low doses and various dose rates and energies as a result of background galactic cosmic radiation and occasional solar particle events (17). Musculoskeletal disuse and radiation exposure each can cause degeneration of connective tissues, including bone (20, 52). Whether the weightless environment of space affects tissue responses to space radiation is unknown. In an early spaceflight experiment, rats were exposed to a relatively high dose (800 rad) of ¹³⁷Cs during flight (Cosmos 690; 20.5 days of spaceflight), and results were compared with ground-based simulation controls and a previous flight without an onboard radiation source (Cosmos 605) (41). Spaceflight reportedly worsened some of the adverse effects of radiation; however, because this conclusion relied heavily on qualitative and histological analyses, rather than functional tests, further research is needed. Better insight into the cellular and molecular mechanisms for bone loss in space may yield new strategies for early intervention to prevent subsequent bone loss and osteoporosis.

Limited access to the spaceflight environment and the importance of understanding physiological mechanisms underlying disuse motivated the development of a ground-based animal model to simulate certain aspects of weightlessness (36, 53). Hindlimb unloading removes weight bearing from the hindquarters of rats or mice and causes a cephalad fluid shift. Hindlimb unloading reduces bone mass (osteopenia), bone strength, and bone formation by osteoblasts and, in some cases, can increase the numbers of osteoclasts in unloaded bones (8, 9, 11, 15, 25, 42, 46). Hindlimb unloading of mice recapitulates many of the cardiovascular, hormonal, immunosuppressive, and musculoskeletal changes observed in astronauts and humans subjected to bed rest, which is a spaceflight analog for humans and, thus, provides a well-characterized model for studying physiological adaptations to the weightless environment of space (2).

Although radiation at high doses can cause substantial degeneration of various tissues, including bone marrow, little is known about how the complex spectrum of space radiation may affect bone. Space radiation cannot be precisely replicated on Earth, although recent insight has been gained by exposure of mice to a single ≤2-Gy dose of total body irradiation (3, 16, 23, 56, 57); a total ∼2-Gy dose corresponds to the dose to which an astronaut would be exposed on a lengthy mission outside the Earth's magnetosphere and to a typical single dose of fractionated therapeutic radiation. The dose rate used in these published reports exceeds that of space, although a test of different radiation sources that contribute to galactic cosmic radiation (protons and heavy ions) reveals that total body irradiation causes a cancellous bone loss similar to that observed following exposure to X-rays or gamma sources (16; Yumoto et al., unpublished observations). In efforts to simulate the low-linear energy transfer (LET), high-energy (≥100-meV) proton component of the space radiation environment, we chose to use gamma rays emitted from the decay of ¹³⁷Cs. These low-LET photons elicit radiobiological effects similar to those elicited by higher-energy protons and provide the means to compare the effects of higher-LET heavy ions in the galactic cosmic radiation. Thus gamma irradiation provides a suitable experimental model to investigate the influence of space-relevant doses of radiation on bone.

Reactive oxygen species (ROS) and reactive nitrogen species (RNS) may mediate tissue degeneration in age-related diseases, such as osteoporosis, and after insults, such as radiation exposure (30). In addition, spaceflight followed by return to Earth can lead to oxidative damage in blood and various soft tissues (18, 48, 49). ROS can directly stimulate bone resorption by osteoclasts (7, 28), and the bone loss due to aging and estrogen deficiency may be attributed, at least in part, to oxidative stress (1, 6, 34). The release of ROS/RNS can cause oxidative damage, which is associated with excess bone resorption by osteoclasts over bone formation by osteoblasts, reduced viability of the putative mechanosensory cells (i.e., osteocytes), and osteoporosis (1). Furthermore, treatment with the potent antioxidant α-lipoic acid (α-LA) can prevent inflammation-induced bone loss (13). The generation of excess ROS/RNS and resulting oxidative damage within skeletal tissues during spaceflight may contribute to later bone loss and weakening during recovery.

We showed previously that gamma irradiation leads to a rapid decrease in cancellous fractional bone volume [bone volume (BV) as a fraction of total volume (TV)] in postpubertal mice (23) and an increase in bone-resorbing osteoclasts (23, 57). We also showed that in vitro irradiation of osteogenic cells from the bone marrow leads to increased generation of ROS (22). These findings support the following hypotheses: 1) irradiation causes oxidative stress and bone loss, and 2) acute radiation damage may be worsened by concomitant musculoskeletal disuse. To begin to test these possibilities, we subjected mice to hindlimb unloading followed by irradiation with ¹³⁷Cs and evaluated parameters relevant to cancellous microarchitecture, cell dynamics, and oxidative stress. We treated animals with α-LA to determine its effectiveness in ameliorating acute radiation-induced bone loss. Our results show that hindlimb unloading and irradiation each acutely stimulated osteoclastic bone resorption and bone loss and that α-LA inhibited the acute osteopenic effects of gamma irradiation.

METHODS

Animals.

Male C57BL/6J mice (Jackson West, Bar Harbor, ME) at 17 wk of age were used for all experiments. Mice were randomized by weight, assigned to groups (n = 6/group), and acclimatized to cages for 2 days before initiation of the experiments. Mice were housed in an animal room under controlled conditions (24 ± 2°C, 55 ± 5% humidity, 12:12-h light-dark cycle). Food and water were available ad libitum. Mice were maintained in standard cages with the same footprint as custom-designed hindlimb-unloading cages. Mice were hindlimb unloaded by tail traction according to previously described methods (35). We measured body weights throughout the experiments to monitor the animals' health. The Ames Research Center Institutional Animal Care and Use Committee approved all procedures.

Experimental protocols.

The experimental conditions for two separate hindlimb-unloading experiments (Fig. 1) were selected on the basis of our previous results showing dose and time dependence of radiation-induced cancellous bone loss (23) and separate hindlimb-unloading studies (40). Two different durations of hindlimb unloading and doses of radiation were used, with the goal of capturing interaction effects of hindlimb unloading and irradiation, should this occur. Our previous time-dependent, radiation dose-response experiments revealed that a longer time was required to elicit significant changes with 1 Gy than with 2 Gy (23). In planning the hindlimb-unloading experiments, we reasoned that hindlimb unloading may accelerate and/or worsen the effects of irradiation, not knowing a priori the effects hindlimb unloading might have on the radiation response. Therefore, we tested the more effective (i.e., higher) dose in the short-term experiment (2 Gy, 3 days), as well as the lower dose in the longer-term experiment (1 Gy, 10 days). Our rationale for irradiating mice during ongoing disuse was that this regimen reflects a likely spaceflight scenario.

Fig. 1.

Experimental protocols. Postpubertal, male C57BL/6 mice for each experiment were randomized by weight and then assigned to groups (n = 6/group). In experiment 1, mice were hindlimb unloaded (HU) by tail traction or normally loaded (NL) for 4 days and then irradiated (IR) with 2 Gy of ¹³⁷Cs or sham-irradiated. Mice were returned to their respective cages for an additional 3 days before tissue harvest. In experiment 2, mice were hindlimb unloaded by tail traction or normally loaded for 4 days and then irradiated with 1 Gy of ¹³⁷Cs or sham-irradiated. Mice were returned to their respective cages for an additional 10 days before tissue harvest. In experiment 3, mice were normally loaded throughout the experimental period and irradiated with 1 or 2 Gy of ¹³⁷Cs or sham-irradiated 10 days or 3 days before tissue harvest, respectively. In experiment 4, mice were normally loaded throughout the experimental period, irradiated with 2 Gy or sham-irradiated, and treated twice daily with α-lipoic acid (α-LA, 25 mg/kg body wt) or vehicle, and tissues were harvested 3 days later.

In experiment 1, mice were hindlimb unloaded or normally loaded, 4 days later they were exposed to 2 Gy of ¹³⁷Cs or sham-irradiated, and after an additional 3 days they were euthanized and tissues were harvested (Fig. 1). Mice were injected with calcein (4 mg/kg sc) 5 days and 1 day before tissue harvest for measurements of bone formation. In experiment 2, mice were hindlimb unloaded or normally loaded, 4 days later they were exposed to 1 Gy of ¹³⁷Cs or sham-irradiated, and after an additional 10 days they were euthanized and tissues were harvested. In experiment 3, normally loaded mice were irradiated with 1 or 2 Gy of ¹³⁷Cs 10 days or 3 days before tissue harvest, respectively. In experiment 4, normally loaded mice were irradiated with 2 Gy of ¹³⁷Cs or sham-irradiated and injected subcutaneously twice daily with α-LA (25 mg/kg body wt) or vehicle (5% ethanol), and tissues were harvested 3 days later.

Gamma irradiation.

Mice were subjected to uniform whole body irradiation using a ¹³⁷Cs irradiator (Mark I, JL Shepherd) equipped with a turntable at a dose rate of 0.92 Gy/min. The animals were positioned in a ventilated animal holding chamber within the irradiator, which provided exposures within 95% of the selected dose.

Sample collection and handling.

After the mice were euthanized, bones were harvested from the carcass. Left tibiae were fixed in 70% ethanol and stored at room temperature until analysis. Lumbar vertebrae were removed as a unit and wrapped with saline-soaked gauze and then stored at −20°C. Muscle was removed from the tibia, and the vertebrae were thawed at 4°C overnight before micro-CT scanning at room temperature.

Micro-CT.

Tibiae were subjected to three-dimensional (3-D) micro-CT analysis using a Viva CT 40 (Scanco Medical, Bassersdorf, Switzerland), as previously described in detail (23). Briefly, we evaluated several images from control mice to establish nominal segmentation values, which were selected as a best average match to gray-scale images to capture bone structure without excessive porosity. The segmentation values were held fixed for all 3-D trabecular evaluations throughout the study. The proximal metaphyses were scanned at maximum resolution (voxel size = 10.5 μm) into a single stack consisting of 210 slices. Cancellous tissue was analyzed in a region 0.2–1.2 mm distal to the growth plate. The first lumbar vertebral body was scanned at maximum resolution into a single stack consisting of 360 slices (3.8 mm) transverse to the craniocaudal axis. A 2.1-mm region was analyzed, beginning 0.5 mm proximal to the caudal end cap and extending proximally toward the cranial end cap. Data are based on calculations for total area, bone volume, and BV/TV. BV/TV describes the fraction of total volume within the selected region of interest that is occupied by bone itself, not marrow space, and values are generally proportional to mechanical properties of the tissue.

Cancellous histomorphometry.

The left tibiae were fixed in 4% paraformaldehyde overnight and then decalcified in 20% EDTA for 2 wk. For quantification of osteoclasts, decalcified bones were embedded in paraffin, and 5-mm-thick sagittal sections were cut and stained for tartrate-resistant acid phosphatase (TRAP), as previously described (37, 50). Measurements were performed within a 0.24-mm² sample area 0.3 mm distal to the growth plate-metaphyseal junction and extending distally 0.7 mm. TRAP-positive cells that also contained more than two nuclei were scored as osteoclasts, and cell numbers were normalized to the total length of bone surface (BS) within the region of interest. For quantification of the cancellous surfaces occupied by osteoclasts, the lengths of bone surfaces occupied by TRAP-positive multinucleated osteoclasts were measured, summed, and normalized to BS within the region of interest. For measurements of bone formation rate (BFR) in mineralized sections, left femora were embedded in polymethylmethacrylate, and 0.5-mm sections were processed for fluorescent detection of calcein. Sections were analyzed by fluorescent confocal microscopy using a Zeiss 510 confocal microscope with postimaging software to optimize signal detection. Surface labeling of bone sections using calcein is visualized as distinct lines, providing a measure of newly formed bone in the period between injection of labels. Calcein labeling was quantified within a 0.34-mm² area (0.5 × 0.67 mm) that was 0.3–0.8 mm from the growth plate of the distal femur. The percentage of cancellous perimeter (BS) covered with a calcein label [mineralizing surface (MS)] and the average distance between double labels normalized per day [mineral apposition rate (MAR), μm/day] were measured using the Zeiss image examiner software. BFR (μm³ · μm⁻² · day⁻¹) was calculated as follows: MS/BS × MAR. Nomenclature, symbols, and units are those recommended by the Nomenclature Committee of the American Society of Bone and Mineral Research (39).

Fluorescent flow cytometry of bone marrow cells.

Cells were flushed from the right femora and then dispersed using an 25-gauge needle. Total and apoptotic cell numbers were quantified by flow cytometry using the Guava Via Count reagents, which are DNA-binding dyes with differential membrane permeability (Guava Technologies, Hayward, CA), in conjunction with the Guava EasyCyte Mini cell cytometer system. For detection of intracellular ROS, freshly isolated bone marrow cells were treated for 30 min at 37°C with the ROS-sensitive fluorogenic dye 5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate (CM-H₂DCFDA, 5 μM; Molecular Probes), as previously described (31). Analysis by flow cytometry was done immediately after tissue harvest and treatment with the dye, and measurements of cells from each animal were performed in duplicate.

Assessment of oxidative damage to lipids.

Marrow cells were flushed from the bones, and the remaining mineralized bones were stored at −80°C in 5 μM butylated hydroxytoluene until quantification of lipid peroxidation. A commercially available kit that measures malondialdehyde (MDA) and 4-hydroxy-2-nonenal (4-HNE) levels in tissue homogenates was used (Lipid Peroxidation Assay Kit, Oxford Biomedical Research, Rochester Hills, MI). Tissue homogenates and experimental procedures were performed according to the manufacturer's instructions. MDA levels were measured in duplicate and calibrated against a standard curve. For localization of peroxidated lipids within bone, decalcified, paraffin-embedded sections of the proximal tibia were blocked with 2% bovine serum albumin and then immunolabeled with a goat anti-MDA antibody and a goat anti-hydroxyalkenal antibody (Alpha Diagnostics, San Antonio, TX). Primary antibody was detected with FITC-conjugated anti-goat IgG secondary antibody.

Data analysis and statistics.

Values are means ± SD (n = 6 per group). For experiments 1 and 2 (hindlimb unloading) and 4 (α-LA treatment), results were analyzed statistically by two-factor ANOVA. Post hoc analysis, using the Tukey-Kramer test, was performed on values exhibiting an interaction effect. For experiment 3 (oxidative stress), results were analyzed by one-factor ANOVA with the Tukey-Kramer post hoc test (Stat View Software, SAS Institute). P ≤ 0.05 was considered to be statistically significant.

RESULTS

Influence of hindlimb unloading and irradiation on body weight.

The health of the mice was monitored by twice-daily weighing and observation of the animals. Body weights at the time of tissue harvest did not differ significantly between groups (Table 1), although the mice that were both hindlimb unloaded (1 wk) and irradiated (2 Gy) weighed 8% less than normally loaded controls at the time of sample recovery. Irradiation (2 Gy) and hindlimb unloading (1 or 2 wk) each caused splenic atrophy, as expected (54, 55), with no greater effect in combination (data not shown). Thus the combination of hindlimb unloading and gamma irradiation was well tolerated by the mice.

Table 1.

Body weight at euthanization

	Body Wt, g
	Sham-irradiated (0 Gy)	Irradiated
	Experiment 1
Normally loaded	29.3±1.9	28.9±1.7
Hindlimb unloaded	28.9±2.3	26.7±1.1
	Experiment 2
Normally loaded	25.8±1.9	25.8±2.1
Hindlimb unloaded	26.3±2.0	25.7±2.4

Values are means ± SD. In experiment 1, mice were hindlimb unloaded for 1 wk and irradiated with 2 Gy 4 days after initiation of unloading. In experiment 2, mice were hindlimb unloaded for 2 wk and irradiated with 1 Gy of ¹³⁷Cs 4 days after initiation of unloading.

Influence of hindlimb unloading and irradiation on cancellous microarchitecture.

Mice were exposed to two different experimental regimens on the basis of our previous findings that irradiation caused time- and dose-dependent cancellous bone loss (23). Micro-CT revealed similar structural changes after irradiation and hindlimb unloading, with less bone and more void space (marrow) evident in 3-D reconstructions of the tissue (Fig. 2A). At 3 days after 2-Gy irradiation, a significant decrement in cancellous BV/TV of tibiae (Fig. 2, A and B) and lumbar vertebrae (Fig. 2D) was observed; in normally loaded mice, irradiation (2 Gy) caused a 16% decrement in tibiae compared with sham-irradiated controls and a 5% decrement in lumbar vertebrae. The lower dose (1 Gy) after 10 days (Fig. 2C) caused a similar decline in the tibia, but not vertebrae (Fig. 2E), which was consistent with our previous results (23). Similarly, hindlimb unloading for 7 or 14 days caused cancellous bone loss in tibiae and lumbar vertebrae. There were no interaction effects in BV/TV by two-factor ANOVA, indicating that the response to irradiation was not significantly different due to loading condition. Thus acute effects of irradiation in normally loaded or hindlimb-unloaded mice were similar in both experiments, and no additional effect was observed when the two factors were combined, suggesting that cellular pathways mediating acute bone loss may be shared.

Fig. 2.

Hindlimb unloading and irradiation effects on cancellous bone volume in axial and appendicular bones. Mice were hindlimb unloaded or normally loaded for 7 days and irradiated with 2 Gy of ¹³⁷Cs 3 days before tissue harvest (experiment 1; A, B, and D) or hindlimb unloaded for 14 days and irradiated with 1 Gy of ¹³⁷Cs 10 days before tissue harvest (experiment 2; C and E). Cancellous tissue within the proximal tibiae (A, B and C) or the centra of the 1st lumbar vertebrae (D and E) were analyzed by micro-CT. Hindlimb unloading and irradiation each reduced cancellous fractional bone volume [bone volume/total volume (BV/TV)], and there was no additional effect when hindlimb unloading was combined with irradiation. Values are means ± SD. Data were analyzed statistically by 2-factor ANOVA, with irradiation and loading condition (IR × load) as factors. P ≤ 0.05 was accepted as significant.

Influence of hindlimb unloading and irradiation on bone cells, marrow cells, and oxidative damage to mineralized tissue.

Hindlimb unloading for 1 wk caused a 47% increase in cancellous bone surface covered with TRAP-positive osteoclasts (OcS/BS) and a 33% increase in the numbers of osteoclasts per bone surface (N.Oc/BS; Fig. 3, A and B). Irradiation at 2 Gy also increased OcS/BS and N.Oc/BS to an extent similar to that observed after hindlimb unloading. No further effects were found when mice were subjected to hindlimb unloading and irradiation. In contrast, irradiation had no effect on BFR, whereas hindlimb unloading of sham-irradiated mice reduced BFR 17% relative to normally loaded controls, as expected (Fig. 3E). The decline in BFR due to unloading in sham-irradiated mice was due to a reduction in MAR (reflecting osteoblast activity; Fig. 3C), but not MS/BS (reflecting osteoblast number; Fig. 3D). The lack of change in BFR due to irradiation was due to a combination of reduced MS/BS and a trend (P < 0.0586) toward increased MAR; thus irradiation did not significantly affect BFR (product of MAR and MS/BS). These results indicate that the rapid cancellous bone loss caused by irradiation was due primarily to increased bone resorption by osteoclasts, rather than reduced BFR.

Fig. 3.

Hindlimb unloading and irradiation effects on osteoclasts and bone formation in situ. Mice were hindlimb unloaded or normally loaded for 7 days and irradiated with ¹³⁷Cs (2 Gy) 3 days before tissue harvest (experiment 1). Proximal tibiae were analyzed for number of osteoclasts normalized to bone surface (OcS/BS, A), number of osteoclasts normalized to total length of bone surfaces within region of interest (N.Oc/BS, B), mineral apposition rate (MAR, C), mineralizing surface as a fraction of bone surface (MS/BS, D), and bone formation rate (BFR, E). Hindlimb unloading and irradiation each increased numbers and bone surfaces covered by osteoclasts, and there was no additional effect when hindlimb unloading was combined with irradiation. Hindlimb unloading alone inhibited BFR as a result of inhibition of MAR (reflecting osteoblast activity). Although irradiation reduced MS/BS (reflecting osteoblast number), it had no effect on BFR. Values are means ± SD. Data were analyzed statistically by 2-factor ANOVA, with irradiation and loading condition as factors. P ≤ 0.05 was accepted as significant.

Given the marked radiosensitivity of bone marrow that contains progenitors for both mesenchymal-derived osteoblasts and hematopoietic-derived osteoclasts, we investigated viability and apoptosis in marrow cells freshly extracted from femora. Irradiation (2 Gy) markedly reduced marrow cell viability (−58%) and increased the fraction of cells that were undergoing apoptosis within 3 days, whereas hindlimb unloading for 7 days had no such effects (Fig. 4, A and B).

Fig. 4.

Hindlimb unloading and irradiation effects on marrow cells and lipid peroxidation of mineralized tissue. Mice were hindlimb unloaded or normally loaded for 7 days and irradiated with 2 Gy of ¹³⁷Cs 3 days before tissue harvest (experiment 1). Bone marrow cells were flushed from femora and analyzed for cell number (A) and apoptosis (B) by nuclear dye exclusion using flow cytometry, and residual mineralized tissue was analyzed for levels of lipid peroxidation by measurement of malondialdehyde (MDA) and 4-hydroxy-2-nonenal (4-HNE) levels (C). Irradiation, but not hindlimb unloading, reduced marrow cell number, increased the fraction of cells undergoing apoptosis, and increased oxidative damage to lipids in mineralized tissue. Values are means ± SD. Data were analyzed statistically by 2-factor ANOVA, with irradiation and loading condition as factors. P ≤ 0.05 was accepted as significant.

To determine whether irradiation and/or hindlimb unloading caused oxidative damage to the mineralized compartment of bone, which may be related to the observed loss of marrow cell viability, we assayed mineralized tissue of femora for lipid peroxidation by measuring the levels of MDA and 4-HNE. Irradiation with 2 Gy of normally loaded mice caused a modest increase in lipid peroxidation after 3 days of exposure, whereas hindlimb unloading for 1 wk had no effect (Fig. 4C).

Gamma irradiation and oxidative stress.

To explore further the cellular and molecular changes triggered by gamma radiation, we irradiated mice with 1 or 2 Gy and recovered samples 3 or 10 days later (experiment 3 in Fig. 1). Irradiation markedly reduced marrow cell viability within 3 days, as previously observed (Fig. 4), but only with the higher 2-Gy dose (Fig. 5A). This effect was transient, inasmuch as viability returned to near-control levels by day 10 (Fig. 5A). The early drop in viability coincided with a transient increase in the fraction of apoptotic cells detected at day 3 (Fig. 5B). To test for the generation of ROS after irradiation, we loaded marrow cells with the fluorogenic dye precursor CM-H₂DCFDA and conducted flow cytometric analysis. Relatively low doses of gamma rays were sufficient to increase ROS levels 3 and 10 days after irradiation (Fig. 5C). Lipid peroxidation levels in the mineralized tissue increased significantly at 1 and 2 Gy 10 days after irradiation (Fig. 5D), indicative of persistent and cumulative oxidative damage. Immunoreactivity for MDA and 4-HNE was evident throughout marrow and mineralized tissue and appeared qualitatively more intense in samples from irradiated mice than controls (Fig. 5, E and F).

Fig. 5.

Irradiation dose- and time-dependent effects on marrow cells and lipid peroxidation of mineralized tissue. Mice were normally loaded throughout the experimental period and irradiated at 1 or 2 Gy 3 days or 10 days before tissue harvest (experiment 3). Bone marrow cells were flushed from femora and analyzed for cell number (A) and apoptosis (B) by nuclear dye exclusion and for reactive oxygen species (ROS) generation after incubation with 5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate (CM-H₂DCFDA, C) by flow cytometry. Lysates of residual mineralized tissue were analyzed for levels of lipid peroxidation by measurement of MDA and 4-HNE levels (D). Decalcified sections of cancellous bone within proximal tibiae were immunostained for MDA and 4-HNE and sham-irradiated (E) and irradiated at 2 Gy (F). Scale bar, 100 μm. B, bone; M, marrow. Irradiation at 2 Gy caused a transient decrease in marrow cell number and increase in fraction of cells undergoing apoptosis. Irradiation increased ROS generation by marrow cells and markedly stimulated lipid peroxidation, with a greater effect 10 days than 3 days after irradiation. Values are means ± SD. Data were analyzed by 1-factor ANOVA with Tukey-Kramer post hoc test. *P ≤ 0.05 vs. control (NL/0 Gy, Ctrl) accepted as significant.

In summary, irradiation and hindlimb unloading caused cancellous bone loss and increased osteoclast number. Only hindlimb unloading inhibited BFR, whereas only irradiation decreased marrow cell viability, increased apoptosis of marrow cells, and caused oxidative damage to lipids in mineralized tissue. In no case did the combination of hindlimb unloading and irradiation exert a greater effect than either factor alone.

α-LA mitigated the adverse effects of gamma irradiation on mineralized tissue and marrow cell viability.

To test whether treatment with an antioxidant may prevent radiation-induced skeletal damage, we treated mice with α-LA during and after irradiation. Mice irradiated 3 days before tissue harvest showed the expected decrement in BV/TV (Fig. 6A), as previously observed (see Fig. 2, A, B, and D, and Ref. 23). Injection of α-LA alone did not significantly affect BV/TV in sham-irradiated mice but did reduce bone loss caused by irradiation (Fig. 6A). Analysis of freshly extracted bone marrow cells showed that α-LA partially, but significantly, ameliorated the radiation-induced decline in marrow cell viability (Fig. 6B).

Fig. 6.

α-LA effects on radiation-induced changes in cancellous bone volume and marrow cell viability. Mice were injected twice daily with α-LA and irradiated with 0 or 2 Gy of ¹³⁷Cs (experiment 4). After 3 days, tissues were harvested and analyzed by micro-CT for cancellous architecture of the proximal tibia (A) and viable marrow cells by flow cytometry (B). Treatment with α-LA inhibited radiation-induced bone loss and decline in marrow cell viability. Values are means ± SD. Data were analyzed statistically by 2-factor ANOVA, with α-LA treatment and irradiation (IR × α-LA) as factors. P ≤ 0.05 was accepted as significant. Bars show significant differences by Tukey-Kramer post hoc given the significant interaction effects.

DISCUSSION

We investigated the short-term effects of musculoskeletal disuse and gamma irradiation, reasoning that probing early changes in bone cell function that lead to tissue loss will yield insight into interactions, mechanisms, and potential therapeutic interventions. Some of the skeletal responses to disuse and gamma irradiation, including the degree of cancellous bone loss and the increase in osteoclasts, were similar, even when treatments were combined. These findings imply that musculoskeletal disuse and irradiation share cellular pathways leading to acute bone loss, namely, bone resorption by osteoclasts. There were, however, notable differences in the skeletal responses to the two stimuli. Hindlimb unloading reduced cancellous BFR, although irradiation did not. In contrast, irradiation caused oxidative changes in bone marrow cells and mineralized tissue that were not observed with hindlimb unloading. Together, these results support the proposal that although increased bone resorption by osteoclasts was responsible for the acute cancellous bone loss due to hindlimb unloading or irradiation, only gamma irradiation did so by causing oxidative stress.

Acute cancellous bone loss induced by hindlimb unloading or irradiation was observed in appendicular (tibiae) and axial (vertebrae) bones, as previously reported (23), although the extent of acute bone loss was greater in the tibiae than in the vertebrae. Similarly, cancellous bone loss is evident in axial and appendicular bones of astronauts during long-duration missions (27), presumably due to insufficient mechanical stimulation by weight bearing. The degree of gamma irradiation-induced osteopenia was similar in hindlimb-unloaded and normally loaded mice in this short-duration study, consistent with results obtained using a heavy-particle radiation source (⁵⁶Fe; Yumoto et al., unpublished observations). These findings do not preclude the possibility that disuse influences skeletal responses to radiation in the long term.

Hindlimb unloading and irradiation each increased the numbers and resorbing surface of osteoclasts in cancellous tissue, and these effects were not additive. Although hindlimb unloading inhibited BFR as expected, irradiation did not. These results, together with earlier findings of a net loss of bone in postpubertal mice within 3 days of exposure to gamma radiation (23), show that acute osteopenia was caused predominantly by increased bone resorption, rather than decreased bone formation. Similarly, astronauts show evidence that bone resorption increases in flight, whereas bone formation does not decrease, on the basis of calcium kinetic studies and biomarker levels (47). The lack of an additive effect of gamma irradiation and hindlimb unloading on osteoclasts may be due to each stimulus saturating a key biochemical pathway, leading to increased osteoclast activity and/or depletion of the pool of osteoclastogenic cells. Further studies are needed to address this issue.

Analysis of bone marrow cell and mineralized tissue responses to irradiation and hindlimb unloading revealed differences in cell viability and oxidative damage. Irradiation, but not hindlimb unloading, decreased viability and increased apoptosis of marrow cells and increased lipid peroxidation within mineralized tissue, as measured by MDA and 4-HNE levels. Furthermore, radiation increased ROS generation by marrow cells. These results are consistent with well-established damaging effects of radiation on other tissues (14). The influence of gamma irradiation on ROS generation was transient and relatively modest (increase of 20–28% relative to controls 3 days after irradiation), whereas radiation-induced oxidative damage to lipids measured by MDA and 4-HNE content accumulated further over time (10 days). Interestingly, immunohistochemical analysis showed that mineralized tissue and bone marrow cells stained positively for MDA adducts. Proteolipids participate in mineralization of the extracellular matrix (5); therefore, the immunopositive material evident in mineralized tissue may be oxidized lipid from extracellular matrix and cell membranes. Lipid peroxidation of chondrocytes in experimental models of osteoarthritis is linked to increased matrix degradation via oxidation of collagen (51), and similarly, the radiation-induced increase in skeletal MDA observed in this study may lead to further degradation of cancellous tissue by osteoclasts.

Our finding that α-LA mitigated radiation-induced cancellous bone loss supports the hypothesis that the generation of ROS in marrow and bone cells leads to cumulative oxidative damage, recruitment and activation of osteoclasts, and resorption of bone. Treatment with α-LA improved the viability of bone marrow cells following irradiation, as observed after other cytotoxic chemotherapeutic treatments, e.g., cyclophosphamide (44). Injection of α-LA in vivo protects connective tissue in experimental models of arthritis (29) and inflammatory bone disease (13, 19). In vitro, α-LA inhibits osteoclast differentiation (19, 21, 29) and protects a bone marrow cell line from DNA damage and apoptosis induced by TNF-α or H₂O₂ (4). In fact, α-LA has pleiotropic functions, including regulation of cellular redox and cytokine and insulin signaling (24, 43). However, our results, together with the findings of others, suggest that the potent antioxidant activity of α-LA mediates its radioprotective effects (32, 33).

Although ground-based models have certain limitations, results from these experiments shed new light on the skeletal effects of unloading and irradiation in animal models subjected to conditions designed to simulate the space environment, (10, 16). In contrast to the space environment, hindlimb unloading removes weight bearing from the hindquarters, whereas in space the entire skeleton is unloaded (36). Furthermore, radiation in space consists of a complex mixture of radiation types (high-energy particles, protons, and secondary particles) with far lower dose rates than the single exposure to gamma rays applied in this study at the higher dose rate (38). Nonetheless, insight into skeletal physiology gained from ground-based studies such as this improves our understanding of the causes and progression of bone loss, which has relevance to human health in space and on Earth.

In conclusion, musculoskeletal disuse and a single dose of gamma irradiation triggered rapid increases in osteoclast numbers and trabecular surfaces covered by osteoclasts and also a loss of cancellous tissue in skeletally mature, male mice; these responses were not additive in the short term. Gamma irradiation caused oxidative damage within bone, and treatment with an antioxidant mitigated the damage. Thus treatment with antioxidants (12) may serve as a useful approach to protect bone from the adverse effects of radiation.

GRANTS

This work was supported by National Aeronautics and Space Administration Grants NNH04ZUU005N/RAD2004-0000-0110 (R. K. Globus) and NNA05CV48A (C. L. Limoli).

DISCLOSURES

No conflicts of interest are declared by the authors.