Introduction

Many clinical and experimental conditions that mechanically unload bone cause disuse osteoporosis and increase fracture risk. For example, disuse due to spinal cord lesion significantly decreases bone mineral density (BMD) in the tibia and femur (Dauty et al., 2000), and significantly reduces the cross-sectional moment of inertia of the femoral diaphysis (Modlesky et al., 2005). Thus, bone bending strength is reduced by spinal cord injury and fracture risk is increased (Wang et al., 2001). These disuse-induced structural changes to bone are due to unbalanced bone remodeling, which typically increases bone resorption and decreases bone formation (Rantakokko et al., 1999; Weinreb et al., 1989).

Inactivity-induced changes in bone remodeling are reflected in serum and urinary markers of bone turnover. Human bedrest increases bone resorption, leading to increased urinary calcium excretion and decreased BMD (Arnaud et al., 1992; Inoue et al., 2000; LeBlanc et al., 1995; Watanabe et al., 2004; Zerwekh et al., 1998). Serum markers of bone resorption consistently increase during bedrest, whereas serum markers of bone formation have been shown to increase, decrease or remain unchanged during bedrest (Inoue et al., 2000; Pedersen et al., 1995; Watanabe et al., 2004; Zerwekh et al., 1998). Bone biopsies have shown that bedrest decreases bone formation rate, decreases osteoblast surface and increases osteoclast surface (Arnaud et al., 1992; Zerwekh et al., 1998). Bone lost during rat hindlimb suspension is primarily due to decreased bone formation (Sakata et al., 2004). Immobilization of canine forelimbs increases both resorption and formation, but substantial bone loss occurs because of increased remodeling space, unbalanced relative increases in resorption and formation, and possibly because of an abnormally long lag time between resorption and formation (Li et al., 2005; Yang Li et al., 2005). Taken together, these studies suggest that bone is lost during disuse as a result of unbalanced remodeling (i.e. more resorption than formation).

Disuse also alters circulating hormone concentrations. Serum parathyroid hormone (PTH) concentration is unchanged or decreased during bedrest (Arnaud et al., 1992; Inoue et al., 2000; Shackelford et al., 2004; Watanabe et al., 2004). Decreased PTH is expected since serum calcium concentration increases during bedrest. A decrease in PTH probably contributes to decreased 1,25 (OH)₂ vitamin D (25-OH D) during bedrest, which reduces intestinal calcium absorption leading to increased calcium excretion (LeBlanc et al., 1995; Shackelford et al., 2004; Zerwekh et al., 1998). Insulin-like growth factor I (IGF-I) has been shown to increase in a human bedrest study that also showed a decrease in the bone formation marker PICP (carboxy-terminal propeptide of type 1 procollagen) (Inoue et al., 2000). The negative correlation between IGF-I and PICP is surprising since IGF-I has anabolic effects on bone. However, rat hindlimb suspension did not affect serum levels of IGF-I and prevented the ability of exogenous IGF-I to stimulate bone formation (Sakata et al., 2004). This suggests that disuse prevents the anabolic effects of IGF-I. Two weeks of rat hindlimb suspension decreased serum leptin, bone formation rate, and trabecular bone volume (Martin et al., 2005). However, low dose intraperitoneal injection of leptin was able to restore serum leptin levels and prevent bone loss by restoring bone formation rate.

Hibernation is a mechanism for surviving prolonged periods of food shortages. Bears do not eat, drink, urinate or defecate during hibernation (Nelson, 1987). During hibernation, bears have a remarkable ability to recycle catabolic waste products (e.g. urea) through neuroendocrine control (Nelson, 1978; Nelson et al., 1983). Bone catabolic products also appear to be recycled in bears during hibernation, probably to keep blood calcium concentration within a homeostatic range. Hibernation causes bone loss in ground squirrels, golden hamsters and little brown bats (Haller and Zimny, 1977; Kwiecinski et al., 1987; Steinberg et al., 1979; Steinberg et al., 1981; Steinberg et al., 1986). However, serum markers of bone formation, trabecular bone volume and BMD do not decrease in black bears during hibernation (Donahue et al., 2003a; Floyd et al., 1990; Pardy et al., 2004). Additionally, cortical bone strength and ash fraction increase with age, and porosity does not change with age in black bears, despite annual periods of disuse (Harvey and Donahue, 2004). Taken together, these findings suggest that bears have evolved biological mechanisms to prevent disuse osteoporosis (Donahue et al., 2005). Systemic hormones and growth factors may be involved in maintaining bone formation in hibernating bears to maintain bone mass in the absence of mechanical loading. We hypothesized that the serum concentrations of PTH, 25-OH D, leptin and IGF-I are correlated with the serum concentration of the bone formation marker osteocalcin. Since histomorphometric indices of bone resorption and formation increase during hibernation (Floyd et al., 1990), we hypothesized that serum markers of bone resorption and formation also increase. We also hypothesized that serum collected in different seasons would cause different levels of prostaglandin E₂ (PGE₂) to be released from MC-3T3 osteoblastic cells, because exogenous PGE₂ prevents disuse-induced bone loss (Akamine et al., 1992; Bakker et al., 2003) possibly by stimulating IGF-I production (McCarthy et al., 1991) to increase collagen production (Canalis et al., 1989). Elucidating the biological mechanisms that regulate bone formation in hibernating bears may help us design anabolic therapies to mitigate bone loss in human osteoporoses. This may be accomplished by targeting genes and circulating hormones (e.g. PTH), which are differentially expressed in bears and humans during disuse.

Materials and methods

Results

Both bone resorption and formation markers increase during hibernation and appear to remain coupled and balanced

The bone resorption marker (ICTP) began to increase immediately after the onset of hibernation (Fig. 1). After 10–20 days, the bone formation markers (osteocalcin and PICP) also increased. This is consistent with the 1–2 week histological `reversal' period between resorption and formation (Jee, 2001). These remodeling markers showed trends of increased resorption and formation throughout the hibernation period, and formation appeared to remain coupled and balanced with resorption. Mean osteocalcin levels were higher (P<0.0001) during and after hibernation compared to pre-hibernation (Table 2).

View this table:

• View inline
• View popup

Table 2.

Mean serum metabolite concentrations for the three study periods

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Normalized serum resorption (ICTP) and formation (PICP and osteocalcin) marker concentrations during the 3-month hibernation period. Each data point is the mean value from five bears. Serum was collected every 10 days; the first sample was taken December 31, the last was March 31. The resorption marker increased immediately after the onset of hibernation. After 10–20 days, the formation markers increased and appeared to remain coupled to the increased resorption for the duration of hibernation. The PICP and ICTP concentration values for the entire study period (i.e. pre-hibernation, hibernation and post-hibernation seasons) have previously been published (Donahue et al., 2003a).

Discussion

For many animals, mechanical unloading of bone decreases bone formation and/or increases bone resorption, resulting in net bone loss. However, both resorption and formation increase during canine forelimb immobilization, but significant bone loss occurs because of unbalanced increases in resorption over formation (Li et al., 2005; Yang Li et al., 2005). Likewise, high bone turnover occurs in patients with spinal cord injury, which leads to bone loss and increased fracture incidence (Modlesky et al., 2005; Pitsillides et al., 1995). Our previous analyses of serum remodeling markers suggested that bears may lose some bone during hibernation because of increased resorption and unchanged formation, but that they may be able to gain it back more quickly than other animals and thus prevent bone loss with aging despite long annual periods of inactivity (Donahue et al., 2003a; Donahue et al., 2003b). However, these conclusions were based on the mean seasonal values of the remodeling markers, not the relative, temporal changes in formation and resorption markers during hibernation. Our current analyses on serum markers of bone metabolism suggest that both resorption and formation increase during disuse in bears, with a normal lag time (i.e. reversal period) between resorption and formation. The increase in formation appears to remain coupled and balanced with the increase in resorption, which probably prevents bone loss during hibernation. This is corroborated by histological data from black bear iliac crest biopsies, which also showed increased resorption and formation activity during hibernation, and by the fact that bears do not lose bone volume or BMD during hibernation (Floyd et al., 1990; Pardy et al., 2004).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Serum osteocalcin levels were significantly correlated with serum PTH levels for pre-hibernation, hibernation and post-hibernation. Samples were pooled (N=27).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

The amount of PGE₂ released by osteoblastic cells was greatest when the cells were treated with serum collected in the post-hibernation period (N=9). P-values are for comparisons with the post-hibernation value. Pre-hibernation (N=8) and hibernation (N=7) values were not different from each other (P=0.48). Values are means ± s.e.m.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Proposed differences in bone remodeling and calcium metabolism between human bedrest (Shackelford et al., 2004), spinal cord injury (SCI) (Maimoun et al., 2005; Modlesky et al., 2005), hyperparathyroidism (HPT) (Brockstedt et al., 1995; Monchik and Gorgun, 2004) and hibernating bears. There is increased remodeling activity in all four conditions; however, bone resorption (Resorb) and formation (Form) appear to be balanced in hibernating bears whereas they are unbalanced in the other three conditions. This imbalance leads to increased cortical porosity and decreased bone mineral density (BMD) and cortical thickness in humans. However, cortical porosity decreases and cortical cross-sectional area and mineral content are unchanged in hibernating bears (McGee et al., 2006). As in human inactivity, resorption increases and formation decreases in hibernating bats and hamsters, leading to increased cortical porosity and decreased cortical thickness (Kwiecinski et al., 1987; Steinberg et al., 1979; Steinberg et al., 1981; Steinberg et al., 1986). In the human conditions, calcium liberated by bone resorption is excreted in urine and increases total (tCa) and ionized (iCa) serum calcium concentration. Bears, however, do not urinate during hibernation and thus it is likely that the calcium liberated by bone resorption is recycled back into bone, which is possibly facilitated by PTH increasing renal reabsorption of calcium.

Inactivity during hibernation causes bone loss in ground squirrels, golden hamsters and little brown bats by increasing bone resorption and decreasing formation (Haller and Zimny, 1977; Kwiecinski et al., 1987; Steinberg et al., 1979; Steinberg et al., 1981; Steinberg et al., 1986). However, aestivation does not decrease bone strength or cross-sectional area in the long bones of green-striped burrowing frogs (Cyclorna alboguttata) (Hudson et al., 2004). After 9 months of aestivation there was no evidence of osteoclast activity; however, there was also no intracortical remodeling in non-aestivating frogs. Unfortunately, little is known about the endocrinology of aestivating frogs (Hudson et al., 2004), and thus the mechanisms that preserve bone during aestivation are unknown.

Bears have evolved many unique biological mechanisms to survive long periods of immobilization without food (Nelson, 1987). These mechanisms appear to include the recycling of calcium and other products of bone catabolism, since bears increase bone turnover but do not excrete waste during hibernation. In humans, bedrest-induced disuse osteoporosis is caused primarily by increased resorption without a corresponding increase in formation (Shackelford et al., 2004; Zerwekh et al., 1998). This results in hypercalcemia and a negative calcium balance brought about by increased urinary and fecal calcium. Since bears do not urinate or defecate during hibernation, it is probable that most of the calcium released from bone by resorption is recycled back into bone via osteoblastic bone formation. Similar extensive bone turnover and recycling was documented during prolonged fasting in northern elephant seals (Mirounga angustirostris) that remain active, but consume neither food nor water for up to 3 months (Patterson-Buckendahl et al., 1994). Ionized calcium did increase in bears by 23% during hibernation, possibly because of the lag time between resorption and formation. Paradoxically, PTH was highest when ionized calcium was highest (Table 2).

Hibernation is a mechanism to survive famine. Bears do not eat, drink, urinate, or defecate during hibernation (Nelson, 1973). Metabolic energy and water is derived from fat catabolism, and water is conserved by not urinating (Nelson, 1973). Urea, which is produced by protein catabolism and normally excreted in urine, does not increase in the blood of bears during hibernation despite the fact that they do not urinate (Nelson, 1978). During hibernation, bears recycle urea nitrogen by increasing protein synthesis (Nelson et al., 1983). Similar recycling mechanisms may help maintain collagenous bone formation; ICTP and PICP are collagen peptide fragments which were found to be increased, indicating an increased, but balanced bone turnover during hibernation (Fig. 1). Bone calcium is also likely to be recycled during hibernation (Fig. 4). The effect of PTH on the kidneys is to increase calcium reabsorption and decrease phosphate. This mechanism seems to be at work in hibernating bears since we found serum ionized calcium concentration to increase and serum phosphorous concentration has been shown to decrease during hibernation (Hellgren et al., 1993). PTH may also facilitate calcium recycling (via increasing bone formation) by acting directly on osteoblasts to increase differentiation and prevent apoptosis (Hock et al., 2001).

Numerous substances influence bone cell metabolism in concert with mechanical stimulation. Further complicating the regulation of bone cell metabolism is the fact that many molecules (e.g. PTH, PGE₂, leptin) can exert both anabolic and catabolic influences. Cortisol, which inhibits osteoblast activity, increases in bears during hibernation (Donahue et al., 2003a). Here we found that IGF-I (which typically promotes bone formation) decreases during hibernation. Thus, it appears that changes in cortisol and IGF-I that occur during hibernation, which would be expected to reduce bone formation, are dominated by the anabolic influence of other molecules, since bone formation markers increase during hibernation. In the hierarchy of factors that regulate bone metabolism, those that regulate calcium concentration probably dominate, followed by the sex steroids and mechanical stimulation (Harada and Rodan, 2003). Since PTH is the main regulator of blood calcium concentration and can have potent anabolic effects on bone (Neer et al., 2001; Zanchetta et al., 2003), it is possible that PTH helps maintain bone formation in hibernating bears. The anabolic actions of PTH require IGF-I in normally loaded bone (Bikle et al., 2002; Kostenuik et al., 1999). However, even though PTH reduces bone loss during disuse (Halloran et al., 1997; Turner et al., 1998), unloaded bone does not respond to the anabolic actions of IGF-I (Bikle et al., 1994), raising the possibility that the anabolic actions of PTH occur via a mechanism independent of IGF-I during disuse. This is consistent with our findings that serum concentration of the bone formation marker osteocalcin increased during hibernation despite decreased IGF-I (Table 2), and that serum PTH was positively correlated with osteocalcin (Fig. 2). These findings support the idea that PTH may have a dominant role in preserving bone formation in hibernating bears, possibly by stimulating osteoblast differentiation and inhibiting osteoblast apoptosis (Hock et al., 2001).

The anabolic effects of PTH may be enhanced when physical activity is resumed following arousal from hibernation. Mechanical loading and PTH act synergistically to increase bone formation in vivo and biochemical signaling in vitro (Ma et al., 1999; Ryder and Duncan, 2000). During remobilization in the spring, bone formation, as indicated by serum osteocalcin, remains higher than during pre-hibernation periods. Serum levels of PTH and IGF-I are also highest in the post-hibernation period, and serum from this season induced the greatest PGE₂ release from osteoblasts in vitro. These findings are consistent with the idea of PTH inducing bone formation via mechanisms involving IGF-I and PGE₂ (Bakker et al., 2003; Bikle et al., 2002; McCarthy et al., 1991). PTH stimulates PGE₂ production in osteoblasts and exogenous PGE₂ prevents disuse-induced bone loss by activating more formation than resorption (Akamine et al., 1992; Bakker et al., 2003). PGE₂ stimulates IGF-I production (McCarthy et al., 1991), which can increase collagen production (Canalis et al., 1989). This raises the possibility that the anabolic actions of PTH occur via different mechanisms during periods of normal mechanical loading compared to periods of unloading, since IGF-I was lowest during hibernation when osteocalcin was highest.

Sympathetic tone also has potent effects on bone metabolism and may contribute to the regulation of bone remodeling in bears. During hibernation, resting heart rate drops to approximately one-fourth to one-half of non-hibernating rates, indicating low sympathetic tone during hibernation (Hellgren, 1998; Nelson et al., 2003). Low intracerebroventricular leptin decreases sympathetic tone, which increases bone formation and decreases resorption (Elefteriou et al., 2005; Takeda et al., 2002). We found that serum leptin did not change between pre-hibernation and hibernation; however, it was significantly lower during remobilization following hibernation. Similar seasonal changes have been reported in brown bears (Hissa et al., 1998). This finding suggests that leptin is not involved in the increased bone turnover that occurs during hibernation. However, decreased serum leptin during remobilization following arousal from hibernation may promote bone formation and decrease resorption, assuming there is a corresponding decrease in intracerebroventricular leptin.

Although our findings on seasonal changes in serum metabolites and their effects on in vitro osteoblast metabolism are provocative, they are somewhat limited because a complete complement of measurements was not made on all serum samples in a given year because of volume limitations. However, for the bears studied, hibernation starts and stops within a few days of the same date each year because hibernation is regulated by the availability of food supplies in the captive facility. Furthermore, some molecules (e.g. ICTP and leptin) have been measured over a number of years in different groups of bears and have shown strikingly similar seasonal variations from year to year. Another limitation is that we have focused our discussion on how inactivity may mediate changes in bone metabolism. However, the bears are also anorectic during hibernation and significant bone loss occurs in humans with anorexia nervosa (Legroux-Gerot et al., 2005). In these patients, serum markers of formation are markedly decreased and resorption markers are increased, explaining why up to 40% of women with anorexia nervosa become osteoporotic. Thus, it is even more remarkable that bears can maintain bone mass during hibernation (Pardy et al., 2004) since they are presented with nutritional challenges as well as challenges resulting from mechanical unloading.

Since disuse-induced bone loss is largely caused by increased bone resorption, anti-resorption drugs (i.e. bisphosphonates) are a logical prophylactic treatment. However, bisphosphonates do not prevent disuse induced loss of BMD following spinal cord injuries (Nance et al., 1999). Furthermore, treating dogs with bisphosphonates had little effect on mitigating disuse-induced bone loss (Yang Li et al., 2005). An alternative approach to treating disuse osteoporosis is to increase bone formation to preserve balance between bone resorption and formation. Hibernating bears may do just that during hibernation, possibly in part through the anabolic actions of PTH. Indeed, recombinant PTH reduces fracture risk by increasing bone mineral density in postmenopausal women (Neer et al., 2001). A more thorough understanding of the mechanisms that maintain osteoblastic bone formation in hibernating bears may provide insight for treating disuse osteoporosis.

List of abbreviations

25-OH D

1,25 (OH)₂ vitamin D

BMD

bone mineral density

ELISA

enzyme-linked immunosorbent assay

ICPT

C-terminal cross-linked telopeptide of type 1 procollagen

IGF-I

insulin-like growth factor I prostaglandin E₂ (PGE₂)

PTH

parathyroid hormone

PICP

C-terminal propeptide of type 1 procollagen

TFA

trifluoroacetic acid