INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

MECHANICAL LOADING OF THE skeleton is a requirement for normal tissue function. At a primary level, when loading is diminished, as occurs during disuse, bone mass is rapidly and substantially reduced (16). At a secondary level, disuse diminishes the tissue deformations that also serve to greatly enhance nutrient exchange to cellular populations within bone (21). For the osteocyte, which resides within small lacunae entrapped in the mineralized matrix far from direct contact with the vascular supply, diminished nutrient exchange precipitated by a loss of loading may stimulate a number of cellular pathways.

Due to its phenotypic morphology reminiscent of neuronal cells (20) and generation of a communication network via gap junctions (6, 36), the osteocyte has been proposed as a likely mechanotransducer within bone. Recent studies have demonstrated, both in vivo (23, 27, 30) and in vitro (14, 15), that the osteocyte responds to alterations in its physical environment by rapid regulation of a variety of factors. However, the process by which loss of loading, or disuse, is translated into biochemical signals that eventually result in locally mediated bone resorption is poorly understood.

Recently, we used an in vivo model of bone adaptation to demonstrate that osteocytes rapidly become hypoxic in response to acute disuse and that this condition can be countered by a brief period of mechanical loading (5). Disuse-induced osteocyte hypoxia arises, presumably, because nutrient exchange within the tissue is greatly diminished when loading is inhibited (21). Given known cellular responses to hypoxia (29, 34), we hypothesized that disuse-induced osteocyte hypoxia may serve to initiate a signal transduction pathway that ultimately results in bone resorption. In the present study, we demonstrate that osteocytes upregulate the hypoxia-dependent transcription factor HIF-1 in response to in vivo disuse and that this response is mimicked when osteocyte-like cells are deprived of oxygen in vitro. Because several HIF-1 target genes have the potential to induce osteoclastic activation [e.g., vascular endothelial growth factor (VEGF)], we speculate that HIF-1 may serve to initiate a cascade of events that eventually result in disuse-induced bone resorption.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In vivo model. Ten adult male turkeys (1-1.5 yr) underwent functional isolation of the left ulna (24). In this well-developed model, the diaphysis of the ulna is deprived of mechanical loading by two parallel metaphyseal osteotomies performed at both ends of the bone. A 3-mm-thick cross section of bone is removed at each end, and the exposed diaphyseal bone is covered with methacrylate-filled Delrin caps. The vascular supply of the diaphysis is maintained through the nutrient artery, and the bones remain viable and capable of responding to mechanical stimuli, consistent with other models of bone adaptation. After surgery, turkeys were assigned to 1-day (n = 5), 3-day (n = 3), or 5-day ( n = 2 ) disuse groups. The animals were killed at appropriate time points in accordance with approved Institutional Animal Care and Use Committee policy at the University of Cincinnati.

Immunohistochemistry. At death, 3-mm-thick sections were extracted from the middiaphysis of the left (experimental) and right (intact) ulna of each turkey. The sections were fixed in 10% buffered formalin and decalcified in EDTA (~10 days at 40°C). The decalcified sections were paraffin embedded, sectioned at 5 µm, and mounted on charged slides. In preparation for staining, the sections were blinded and deparaffinized in xylene and graded ethanol washes. As previously described, brief Pronase digestion (10 min at 38°C; Biomeda) was used to aid in antigen retrieval (5). The sections were then rinsed in PBS-2% Brij 35 and blocked with 10% horse serum (10 min at room temperature; Vector Laboratories). The sections were then incubated with a HIF-1 monoclonal antibody (Neomarkers, clone OZ12), followed by an anti-mouse, FITC-conjugated secondary antibody. In addition, five sections were selected at random and subjected to identical staining procedures, with the exception of the primary antibody, to serve as a negative control.

Imaging. A Zeiss LSM510 laser-scanning microscope was used to image the sections (25-mW argon laser, 488-nm blue filter, ×63 water objective). On the basis of preliminary studies, optimal laser transmission excitation, amplifier gain and offset, and pinhole diameter were established. Identical settings were used for all imaging. HIF-1 expression was assessed systematically around the cortex of each section. Four adjacent images were obtained on three surfaces (periosteal, intracortical, endocortical) at each of six anatomically located sites (i.e., 72 images per section, Fig. 1). In each field, the number of osteocytes staining positive for HIF-1 was expressed as a percentage of the total osteocytes in the field (i.e., 25-30). t-Tests were used to determine whether disuse elevated expression of HIF-1 and whether elevation of HIF-1 expression was uniform around the cortex or across bone surfaces.

View larger version (28K): [in this window] [in a new window]   Fig. 1.   Schematic of the ulna middiaphysis cross section illustrating confocal sampling sites. At each of 6 sites (geometrically determined from the corners of the ulna and the midpoints), 4 adjacent confocal images were taken from the endocortical, intracortical, and periosteal surfaces (E, I, and P, respectively, on enlarged site 2). Dorsal (D) and cranial (CR) surfaces are noted.

Cell culture. MLO-Y4 osteocyte-like cells were provided by Dr. Lynda Bonewald (University of Texas Health Science Center, San Antonio, TX). These cells were derived from murine long bones and were cultured as described previously (13). Maintenance of dendritic phenotype was observed as well as expression of a 43.8-kDa protein phenotypic of these cells. In preparation for hypoxia studies, the cells were plated on 100-mm dishes coated with rat tail type I collagen and grown to 50% confluence in -MEM supplemented with 2.5% fetal bovine serum and 2.5% calf serum.

Western blot analysis. Medium was changed 12 h before exposure to either normal oxygen (19% O₂) or hypoxia (2% O₂) for 4, 8, or 12 h. Cells were trypsinized, and nuclear extracts were prepared for each time point as described previously (4). An equivalent amount of nuclear protein (45 µg) from each experimental condition was loaded onto a 7-12% gradient SDS-polyacrylamide gel, electrophoresed, transferred to nitrocellulose, and immunoblotted with anti-HIF-1 monoclonal antibody (Biomol, clone H167). Antibody was detected using a commercial chemiluminescence enhancement kit, following manufacturer's instructions (Super Signal West Pico, Pierce Chemical). As a positive control, MLO-Y4 cells were incubated in the presence of N-CBZ-Leu-Leu-norvalinal, a proteosome inhibitor that blocks the normal ubiquitin-mediated pathway of HIF-1 degradation (35). The experiment was repeated four times with independent extracts. Mean increases in elevation of HIF-1 band density under hypoxic conditions were quantified from scanned autoradiograms using National Institutes of Health Image software.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Depriving bone of mechanical loading significantly elevated the mean (±SE) percentage of osteocytes staining positive for HIF-1 compared with intact normal bones (30.9 ± 6.1 vs. 14.1 ± 3.8%, P < 0.001; Fig. 2). The percentage of HIF-1-positive osteocytes did not vary between 1 and 5 days of disuse (1 day: 33.3 ± 7.5%, 3 days: 26.0 ± 5.7%, 5 days: 32.2 ± 4.5%; Fig. 3); therefore, data were grouped together to assess the distribution of HIF-1 expression around the cortex. In intact bones, HIF-1 expression at the six cortical sites was consistent (range from 10.7 ± 1.9 to 16.5 ± 5.9%). Disuse uniformly elevated HIF-1 expression (range from 28.3 ± 5.9 to 34.4 ± 5.7%; Fig. 3). Examination of HIF-1 expression across bone surfaces in intact bone revealed that expression levels were lower intracortically (8.5 ± 2.8%) compared with both endocortical (16.8 ± 3.7%) and periosteal (16.2 ± 3.0%) surfaces (Fig. 3). In response to disuse, HIF-1 expression was elevated to a consistent level at each of the surfaces (range from 26.8 ± 4.1 to 33.7 ± 5.6%), although increases in elevation were higher intracortically (3.2-fold) than on endocortical (1.8-fold) or periosteal (2.1-fold) surfaces.

View larger version (34K): [in this window] [in a new window]   Fig. 2.   Gray-scale confocal image of haversian canal (H) and several surrounding osteocytes staining positive for hypoxia-inducible factor HIF-1 in response to acute disuse (A, white arrows). One osteocyte in this field did not stain positive for HIF-1 (open arrow). The percentage of osteocytes staining positive for HIF-1 in response to acute disuse was substantially elevated (B). Compared with control bones (c) in which 14% of osteocytes stained positive for HIF-1, nearly 31% of osteocytes in unloaded bones (x) stained positive for HIF-1 (*P < 0.001).

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Percentage of osteocytes staining positive for HIF-1 assessed by period of acute disuse (A), site around cortex (B), and bone surface (C). Elevation of HIF-1 expression by osteocytes was significant (*P < 0.01) at 1, 3, and 5 days of disuse (x). Similarly, disuse elevated HIF-1 expression at each of the 6 sampling sites (*P < 0.01). HIF-1 expression was lowest intracortically in control bones (c), but all surfaces demonstrated upregulation in response to disuse (*P < 0.01).

Western blot analysis of MLO-Y4 cells indicated that HIF-1 protein expression was upregulated in response to acute oxygen deprivation. Compared with extracts from cells cultured at normal oxygen, 4 h of 2% O₂ induced a 2.2 ± 0.6-fold elevation of HIF-1 protein expression. Exposure to 8 h of 2% O₂ resulted in a similar elevation of HIF-1 (2.1 ± 0.5-fold; data not shown), whereas 12 h of hypoxia induced a 3.7 ± 1.2-fold elevation of HIF-1 protein levels relative to those detected in parallel cultures incubated under normoxic conditions (Fig. 4).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Enhanced expression of HIF-1 by MLO-Y4 osteocytes in response to oxygen deprivation. A and B: representative Western blot analyses of MLO-Y4 cells using a HIF-1 monoclonal antibody. In A, parallel cell cultures were exposed to normoxia (N), the proteosome inhibitor N-CBZ-Leu-Leu-norvalinal [CBZLLN (Ub)], or 4 h of 2% O2 (2%). Elevated HIF-1 protein expression was observed in response to CBZLLN (via blockage of HIF-1 degradation) and the acutely reduced oxygen environment. In B, HIF-1 is elevated by 4 h of hypoxia (N vs. 4 h) and is further elevated after 12 h of 2% O2. The 118-kDa HIF-1 protein is noted. C: across 4 independent experiments, HIF-1 expression was elevated slightly more than 2-fold after 4 and 8 h of hypoxia and nearly 4-fold after 12 h of hypoxia.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study, we used an in vivo model of bone adaptation to demonstrate that the percentage of osteocytes staining positive for HIF-1 increases rapidly in response to acute disuse. In addition, we used Western blot analysis to demonstrate that MLO-Y4 osteocyte-like cells rapidly upregulate HIF-1 protein expression in response to acute in vitro oxygen deprivation.

Although these data represent the first report of HIF-1 regulation by osteocytes, the ability to upregulate HIF-1 in response to oxygen deprivation is nearly universal (26). Under a normal oxygen environment, HIF-1 is expressed at low levels in numerous cell types (11, 12). Our observations of basal expression of HIF-1 by osteocytes in normal bone and in MLO-Y4 cells under normal oxygen culture conditions are consistent with these data. Whether the basal levels of HIF-1 expression observed here are similar to those in other bone cells (e.g., osteoblasts) or are unique to osteocytes remains unanswered until further study. As has been shown previously with other cell types (and suggested by the positive control presented in Fig. 4), it is likely that upregulation of osteocyte HIF-1 expression was achieved by inhibiting the ubiquitin-proteasome degradation pathway as opposed to transcriptional upregulation of the HIF-1 gene (25). One caveat with our data lies with the use of a mouse monoclonal antibody raised against human HIF-1 to detect HIF-1 expression in turkey osteocytes via immunohistochemistry. We have not directly demonstrated that the human HIF-1 antibody used in this study cross-reacts with the turkey HIF-1 protein. However, the HIF-1 protein is highly conserved across species (90% homology between human and mouse overall, 90% homologous in the region used to generate the antibody), suggesting it is also conserved in turkeys. Furthermore, we were able to confirm upregulation of HIF-1 by osteocytes in the subsequent in vitro experiments.

In this study, deprivation of loading significantly elevated the percentage of osteocytes staining positive for HIF-1 in a uniform manner around the cortical surface. However, the strain environment of bone is highly nonuniform (1, 8). The enhanced nutrient exchange facilitated by this stimulus is also, presumably, highly nonuniform. Therefore, osteocytes that reside in different portions of the cortex (e.g., near intracortical haversian systems or within lamellar bone near the endocortical or periosteal surface) or are consistently exposed to varying levels of mechanical stimuli (e.g., locations of peak strain or minimal strain at the neutral axis) are likely to have accommodated to widely varying levels of basal nutrient exchange. It is the alteration from the cell's normal environment, then, that appears to precipitate upregulation of the HIF-1 pathway.

Osteocyte hypoxia, and specifically upregulation of the HIF-1 pathway, has substantial potential to mediate disuse-induced bone loss. Although data from this study are not conclusive, they are consistent with this pathway. Upregulation of HIF-1 results in enhanced expression of a wide array of downstream genes that, in general, act to accentuate oxygen delivery or decrease the cell's need to consume oxygen (26). The substantial and growing list of HIF-1 target genes includes at least two factors whose upregulation would suggest how altered osteocyte oxygen homeostasis may lead to the osteoclastogenesis responsible for disuse-induced bone loss. For example, VEGF is rapidly upregulated after elevated HIF-1 expression. VEGF is a potent stimulus for angiogenesis (7), induces osteoclast activation (19), and also enhances osteoclastic activity (18). Recent reports demonstrate that osteoblast-like cells upregulate VEGF mRNA and protein expression in response to acute hypoxia, suggesting an additional role for this process in fracture repair (2, 28). It should be noted that this hypothesis is not solely predicated on VEGF as the mediator and is likely to be more complex and involve other bone-relevant cytokines that are induced by hypoxia such as tumor necrosis factor- and granulocyte-macrophage colony-stimulating factor (10).

Alternatively, hypoxia may influence bone resorption over a broader time course. Recent studies suggest that HIF-1 works in concert with the tumor suppressor protein p53 to mediate programmed cell death (3). Explorations of osteocyte apoptosis have indicated that osteocyte cell death is exacerbated in a variety of conditions, including estrogen depletion (31) and overloading of bone (33). Once osteocyte apoptosis occurs, the tissue must be revitalized via remodeling, and this requires the activation of osteoclast populations. It is interesting to note that cell apoptosis in response to oxygen deprivation is most pronounced when the cell is reoxygenated. On restoration of normal oxygen status after short-term exposure to hypoxia, HIF-1 is rapidly degraded without deleterious effects (22). After prolonged hypoxia, however, it is the return of oxygen metabolism that is often fatal for the cell (32). The VEGF and/or apoptosis pathways would be consistent with the locally mediated bone resorption noted in a variety of in vivo models of disuse osteopenia (9, 17, 37).

In summary, we have determined that acute loss of mechanical loading and direct oxygen deprivation both result in upregulation of HIF-1 by osteocytes. We suggest that this pathway may mediate disuse-induced bone loss. If our premise is supported, it should be possible to develop novel treatment strategies to combat musculoskeletal challenges as diverse as space-induced bone loss and fracture nonunions.