Basic Biology of Skeletal Aging: Role of Stress Response Pathways

Abstract

Although a decline in bone formation and loss of bone mass are common features of human aging, the molecular mechanisms mediating these effects have remained unclear. Evidence from pharmacological and genetic studies in mice has provided support for a deleterious effect of oxidative stress in bone and has strengthened the idea that an increase in reactive oxygen species (ROS) with advancing age represents a pathophysiological mechanism underlying age-related bone loss. Mesenchymal stem cells and osteocytes are long-lived cells and, therefore, are more susceptible than other types of bone cells to the molecular changes caused by aging, including increased levels of ROS and decreased autophagy. However, short-lived cells like osteoblast progenitors and mature osteoblasts and osteoclasts are also affected by the altered aged environment characterized by lower levels of sex steroids, increased endogenous glucocorticoids, and higher oxidized lipids. This article reviews current knowledge on the effects of the aging process on bone, with particular emphasis on the role of ROS and autophagy in cells of the osteoblast lineage in mice.

Characteristics of the Aged Skeleton

Advancing age represents a major risk factor for the decline in bone mass and strength and, therefore, the rise in the incidence of bone fractures. Indeed, it is expected that the incidence of osteoporotic-related fractures in the United States alone will reach 3 million annually by 2025 (1). The adult skeleton is continuously remodeled throughout life by the coordinated activities of osteoclasts and osteoblasts. Osteoclasts derive from cells of the hematopoietic lineage and are responsible for bone resorption, whereas osteoblasts originate from the mesenchymal lineage and are responsible for bone formation. With aging, the amount of bone resorbed by osteoclasts is not fully restored with bone deposited by osteoblasts and this imbalance leads to loss of bone mass and strength (Table 1). The decline in whole bone strength is due to reductions in cancellous and cortical bone density, decreased cortical thickness, and a marked increase in cortical porosity (2–5). Strikingly, the age-associated increase in cortical porosity accounts for 76% of the reduction in cortical strength (6).

Table 1.

Age-Related Changes in Bone that are Common to Humans and Rodents.

Age-Related Changes in Bone	Humans	Rodents
Morphologic changes
Decreased BMD	(7–9)	(10–12)
Decreased cancellous bone	(2)	(10,13,14)
Decreased cortical bone	(2)	(15–17)
Increased cortical porosity	(3–6)	(15,16)
Decreased wall width	(18,19)	(10)
Cellular changes
Decreased remodeling in cancellous bone	(20)	(10,21)
Increased remodeling in cortical bone	(3)	(22)
Increased Ob/Ot apoptosis	(23,24)	(10)
Decreased osteocyte density	(23,25,26)	—
Increased marrow adiposity	(27–29)	(30)
Dysfunctional MSCs	(31)	(32,33)
Molecular changes
Increased oxidative stress	—	(34)
Decreased Wnt signaling	—	(17,34)
Increased PPARγ	—	(30,35)
Decreased growth factors	(36–38)	(39)
Decreased/increased RANKL	(40)	(21)
Decreased/increased Sost	(40)	(21)

Notes: BMD = bone mineral density; Ob/Ot = osteoblast and osteocyte; PPAR = peroxisome proliferator-activated receptor; RANKL = receptor activator of nuclear factor kappa-B ligand.

The most common histological finding in aged human bone is reduced wall width—an index of the reduced amount of work (deposition of bone matrix) performed by teams of osteoblasts during bone remodeling (18,19). This deficient bone formation is due primarily to an insufficient number of osteoblasts (Table 1) (41). The defective osteoblast number in the aging skeleton has been attributed to a decrease in the number of mesenchymal stem cells (MSCs), defective proliferation and differentiation of progenitor cells, or diversion of these progenitors toward the adipocyte lineage, as well as to increased apoptosis (Figure 1). Another common histologic feature of aged human bone, which may be related to the reduced osteoblast number, is a decrease in osteocytes density (25,26). Mineralization of osteocyte lacunae, a phenomenon termed micropetrosis, may also contribute to the decrease in osteocytes density with age (23). The conditions that cause micropetrosis are unclear, but this process may be one potential outcome of osteocyte death.

Figure 1.

Mechanisms of age-related changes in bone. Both cell intrinsic and cell extrinsic mechanisms contribute to the reduced osteoblast generation and increased osteoblast and osteocytes apoptosis that lead to the loss of bone mass with aging. For example, an increase in ROS and endogenous hyperglucocorticoidism with age, as well as sex steroid deficiency, promote osteoblast and osteocyte apoptosis. Osteocyte apoptosis leads to a reduction in osteocytes density.

Similar to humans, loss of bone mass and strength with advancing age in rodents is associated with an increase in the prevalence of apoptotic osteoblasts and osteocytes and a corresponding decrease in osteoblast number, and bone formation rate (10–12,42,43). Increased intracortical porosity, associated with intense remodeling activity, is also a feature of murine bone aging (15,16,22). In contrast, the number of osteoclasts in cancellous bone is diminished with age in the mouse, in line with a concomitant decrease in receptor activator of nuclear factor kappa-B ligand (RANKL) levels in the bone marrow plasma (10,21). Thus, aging seems to exert opposite effects on bone remodeling in the cancellous versus the cortical bone compartments.

The imbalance between bone resorption and formation with age is due to multiple factors, which include extrinsic and intrinsic mechanisms of cell dysfunction. This article reviews current knowledge on the mechanisms of age-related bone loss with emphasis on the role of reactive oxygen species (ROS) and autophagy.

Which Bone Cells Age?

Although there is no universally accepted definition, cellular aging is frequently described as a decline in function due to the accumulation of damage to lipids, proteins, and DNA (88). This definition then leads to the question of which bone cells experience declines in function that mediate the age-associated changes to the skeleton. Mature osteoblasts and osteoclasts responsible for the synthesis and resorption of bone matrix, respectively, are terminally differentiated postmitotic cells with a short life span (19,44). For bone formation to continue uninterrupted throughout life, osteoblasts need to be constantly replaced with new ones (45) originating from MSCs and produced through the replication and differentiation of a lineage-committed progenitor cell (Figure 1). MSCs and osteocytes—former osteoblasts buried in the bone matrix—are long-lived cells and, consequently, are likely more prone to suffer the intrinsic damaging effects of aging.

Mesenchymal Stem Cells

The loss of regenerative capacity of different tissues with aging, which in turn leads to an impaired response to stress, has led to the idea that aging is due, at least in part, to the loss of functional adult stem cells needed for tissue repair (46,47). In support of this hypothesis, old mice have a significant reduction in the number, proliferative capacity, or differentiation potential of distinct stem cells including neuronal (48,49), germline (50,51), hematopoietic (52), and muscle stem cells (53,54).

MSCs that give rise to the osteoblast lineage are commonly defined by their in vitro ability to differentiate into osteoblasts, chondrocytes, and adipocytes (55,56). These cells are found in the bone marrow and in the perivascular niche in multiple human organs (57). However, the identity and location of adult MSCs in vivo have remained elusive due to lack of specific cellular markers. Lineage tracing studies in mice have identified subsets of mesenchymal cells in the bone marrow that express smooth muscle α-actin and/or myxovirus resistance-1 that can give rise to osteoblasts present on the bone surfaces (45,58). Because these cells also persist in the marrow, these markers may identify bona fide stem cells.

Changes in the behavior of bone marrow–derived MSCs with aging have been reported in humans and rodents. These include loss of potential to proliferate and differentiate, loss of capacity to form bone in vivo, and increased senescence (31–33,59). Similar to the MSCs from bone marrow, multipotent cells from adipose tissue show an age-dependent loss of self-renewal capacity as well as an increased propensity for adipogenesis (60). In addition, MSCs from patients with Hutchinson–Gilford progeria syndrome, a disease characterized by accelerated aging, exhibit defective ability to differentiate (61).

The loss of mesenchymal progenitor functionality with age might be due to the accumulation of oxidative damage. Indeed, progenitor cells from old rats exhibit higher levels of oxidized proteins and lipids and showed decreased levels of antioxidant enzyme activity compared with cells from young rats (62). The increased oxidative stress was associated with decreased colony-forming unit fibroblast numbers, increased levels of apoptosis, and reduced proliferation and potential for differentiation. Oxidative stress is also a critical determinant of hematopoietic and neuronal stem cell dysfunction (63,64). These studies suggest a mechanistic link between intracellular oxidants and the decline in regenerative function that occurs as a normal consequence of aging.

Importantly, recent evidence has established that adult stem/progenitor cell dysfunction contributes to aging-related degeneration. Indeed, administration of muscle-derived stem/progenitor cells, isolated from young wild-type mice, to a mouse model of a human progeria conferred significant life-span and health-span extension (65). It remains unknown, however, whether the changes in adult MSCs with aging are functionally related to skeletal involution.

Osteocytes

Within cancellous bone, a subset of osteoblasts becomes embedded in bone matrix with each remodeling cycle thereby generating a fresh population of osteocytes within the new packet of bone. This process also occurs within remodeling cortical bone, but osteocytes near the periosteum are derived from the osteoblasts that continuously expand the periosteum via modeling. Regardless of their origin, all osteocytes are only as old as the bone matrix in which they are embedded, which depends on the rate of remodeling at that site. For example, osteocytes within auditory ossicles, which remodel very little after birth, are essentially as old as the individual (66). Thus, depending on their anatomical location, osteocyte life span can range from a few months to several decades.

Osteocytes have many important functions, including the control of bone resorption and formation, as well as the regulation of phosphate homeostasis. Many of these functions were only recently identified as a result of the development of genetic tools to manipulate gene expression specifically in osteocytes (67,68). For example, osteocyte-specific gene deletion studies have revealed that osteocytes control bone resorption via production of the osteoclastogenic cytokine RANKL (69,70), whereas they control phosphate homeostasis via production of Dmp1 and FGF23 (71). In addition, a combination of mouse and human genetic studies has revealed that osteocytes control bone formation via production of the Wnt-antagonist sclerostin, which is a product of the Sost gene (72–74).

Because osteocytes are long-lived cells, they are likely to be relatively more susceptible than osteoblasts or osteoclasts to the molecular changes caused by aging. Consistent with this idea, accumulation of the age-associated pigment lipofuscin has been demonstrated in osteocytes in rodent bone (75). Nonetheless, there has been little quantitative measurement of molecular damage in aged osteocytes, likely due to the difficulty in isolating pure populations of this cell type. Despite the paucity of quantitative evidence of molecular damage, extensive evidence has shown that osteocyte survival and number decline with age in rodents and humans (10,23,24). Based on the observation that osteocytes control bone resorption via production of RANKL and bone formation via production of sclerostin, one might anticipate that reduced osteocyte number might lead to altered production of these factors and consequent changes in bone remodeling. Consistent with this idea, recent studies by Halloran and colleagues have revealed that the levels of soluble RANKL and sclerostin decline with age in bone marrow supernatants from C57BL/6 mice (21). Although the change in soluble RANKL levels are consistent with reduced osteoclast number on cancellous bone, the low sclerostin levels would be expected to promote Wnt signaling and thereby osteoblast formation; yet osteoblast number is reduced with age. Thus, altered production of known osteocyte products does not, by itself, provide a clear explanation for the reduced cancellous bone turnover associated with aging. It is possible that oxidative stress together with other age-intrinsic mechanisms may alter bone remodeling by decreasing not only the number of osteocytes but also the synthetic capacity of osteocytes. Why these mechanisms have an overall different effect on the rate of remodeling in cancellous versus cortical bone remains unclear.

Cell Intrinsic Mechanisms of Skeletal Aging

Oxidative Stress

Free radical damage has been considered a key component in the tissue degeneration associated with aging (76,77). The majority of cellular ROS is generated by the mitochondria during oxidative phosphorylation. In addition, ROS can be produced in other cellular compartments by nicotinamide adenine dinucleotide phosphate oxidases, lipoxygenases, and other enzymes in response to growth factors and cytokines (78). To prevent excessive ROS production, cells scavenge ROS by multiple mechanisms including production of enzymes such as superoxide dismutases (SODs) and catalase as well as thiol-containing oligopeptides with redox-active sulfhydryl moieties, the most abundant of which are glutathione and thioredoxin (79). Several transcription factors, including FoxOs and p53, have been identified as important defense mechanism against oxidative stress. Indeed, FoxOs promote the maintenance of hematopoietic and neuronal stem cells by attenuating ROS (63,64). Growth factors inhibit FoxO activity by promoting Akt-mediated phosphorylation of FoxO1, 3, and 4, which results in their nuclear export (80–84). In contrast, growth factor depletion or increased ROS levels activate FoxOs and thereby the transcription of genes involved in free radical scavenging, cell cycle, DNA repair, and life span (85,86).

The accumulation of dysfunctional mitochondrial and a decline of antioxidant defense mechanisms with age result in an increase in the levels of free radicals. A sustained excess of ROS damages proteins, lipids, and DNA (87,88). Indeed, oxidative stress is responsible for age-related tissue damage and various disease states such as diabetes, cardiovascular diseases, cancer, and neurodegeneration (89).

Oxidative stress also increases in bone with age (10,17). Evidence from pharmacological and genetic studies in mice has provided support for a deleterious effect of oxidative stress in bone. Specifically, the loss of bone caused by gonadectomy in males or females is attenuated by antioxidants (10,90,91). In contrast, mice treated with the prooxidant buthionine sulfoximine and murine models of premature aging associated with oxidative damage exhibit low bone mass (91–93). Importantly, mice with global deletion of the antioxidant gene Sod1 exhibit low bone mass, which worseness with age (94). Similar to the aged skeleton of wild-type mice, Sod1^{− /−} mice exhibit decreased trabecular and cortical bone mass, which is associated with lower osteoblast and osteoclast numbers. Administration of an antioxidant to Sod1^{− /−} mice normalizes these defects implicating ROS as the main cause of the low bone mass in this model.

Further support for a role of oxidative stress in skeletal homeostasis was provided by murine models of FoxO loss or gain of function. Specifically, conditional deletion of FoxO1, 3, and 4 in young mice or targeted deletion of FoxO1 in osteoblasts resulted in an increase in oxidative stress in bone, a decrease in the number of osteoblasts, and low bone mass (95,96). Conversely, overexpression of a FoxO3 transgene in mature osteoblasts decreased oxidative stress and increased osteoblast number and bone mass.

At the cellular level, oxidative stress decreases the life span of osteoblast in bone, as highlighted by the observation that administration of antioxidants abrogates osteoblast apoptosis in ovariectomized or aged mice (10,17). In addition, osteoblasts from Sod1- or FoxO-null mice exhibit decreased life span (94,95).

ROS also inhibit the Wnt/β-catenin signaling pathway. β-Catenin is indispensable for osteoblastogenesis during development and in adulthood, and loss or gain of function of this pathway is associated with a pronounced decrease or increase of bone mass, respectively, in humans and mice (97). Wnt proteins bind to the Frizzled/LRP5 or LRP6 receptor complex, thereby preventing the proteasomal degradation of the transcriptional coactivator β-catenin (98). β-Catenin translocates into the nucleus where it associates with the T-cell factor (TCF) lymphoid-enhancer binding factor family of transcription factors and regulates the expression of Wnt target genes. In the setting of oxidative stress or nutrient depletion, FoxOs divert the limited pool of active β-catenin from TCF- to FoxO-mediated transcription in diverse cell types, including osteoblasts, colon cancer cells, and hepatocytes (34,99,100). In line with this, targeted deletion of FoxOs in osteoblast progenitors has elucidated that FoxOs restrain the pro-proliferative effects of Wnt signaling on these cells and thereby attenuate bone formation throughout life (101). Similar to the osteoblast progenitors, deletion of FoxOs in enteroendocrine progenitors or in neuronal progenitors increases β-catenin/TCF-mediated transcription and proliferation (64,102). Thus, diversion of β-catenin from TCF- to FoxO-mediated transcription in response to stressful conditions, such as the increased ROS and growth factor deficiency that occur with aging (Table 1), may represent a pathogenetic mechanism for osteoporosis and perhaps several other degenerative diseases associated with old age (103–107).

Autophagy

Autophagy is a process in which cellular components, such as long-lived proteins and organelles, are degraded by the lysosome to maintain the health and viability of the cell. Under normal physiological conditions, autophagy acts as a quality control system that removes defective organelles or protein aggregates. However, during stressful conditions, such as nutrient deprivation or hypoxia, autophagy is increased to break down cellular components for use as an energy source. There are several forms of autophagy including chaperone-mediated autophagy, microautophagy, and macroautophagy (108). Chaperone-mediated autophagy involves the targeting of proteins containing a specific five amino acid motif directly to the lysosome for degradation, whereas microautophagy involves invagination of the lysosomal membrane to engulf small portions of the cytoplasm. Macroautophagy refers to a process whereby large components of the cytoplasm, including mitochondria and protein aggregates, are surrounded by a double-membrane structure to form a vacuole known as an autophagosome, which eventually fuses with the lysosome to allow degradation of the components (109) (Figure 2). The most prevalent form of autophagy in many cell types is macroautophagy and hereafter we will use the general term autophagy to refer to this process.

Figure 2.

Autophagy degrades and recycles cellular components. Autophagy is an intracellular recycling pathway in which cellular components, including protein aggregates and organelles such as mitochondria, are targeted to the lysosome for degradation. Cellular components targeted for degradation are engulfed by a double-membrane structure known an autophagosome. Autophagosome formation depends on series of ubiquitin-like conjugation reactions. In this process, Atg7, which is an E1-like enzyme, activates a ubiquitin-like protein known as LC3. LC3 then becomes conjugated to phosphatidyl ethanolamine (PE) and thereby promotes autophagosome production. Importantly, Atg7 is essential for autophagy and conditional deletion of this gene can be used to examine the importance of autophagy in specific cell types.

After initiation of the autophagy process, autophagosome formation is controlled in part by activation of a ubiquitin-like protein known as LC3 (110). LC3 is activated by the E1-like enzyme Atg7 and is eventually conjugated to phosphatidyl ethanolamine in the growing double-membrane structure. It is important to note that this conjugation process is absolutely dependent on Atg7 and that genetic inactivation of Atg7 effectively suppresses autophagy (111). Thus, conditional deletion of Atg7 is frequently used to determine the role of autophagy in a particular cell type or tissue.

Because aging is associated with the accumulation of damaged cellular components, it has been proposed that autophagy may decrease or become less efficient with age (112–117). Consistent with this idea, autophagy in the kidney declines in aged mice and can be restored by caloric restriction (118). Similarly, expression of some autophagy-related genes declines in brain tissue from aged humans and in chondrocytes in osteoarthritic bones from aged mice (119,120).

Additional work suggests that autophagy contributes to the long-term health of cells and organisms. Specifically, an intact autophagy system is required for the life-span-extending action of caloric restriction or daf-2 pathway mutations in Caenorhabditis elegans (114) and promotion of autophagy extends life span in Drosophila (121). Moreover, deletion of autophagy genes in myocytes, pancreatic β cells, or T cells leads to accumulation of damaged mitochondria and increased ROS production (122–127). These latter studies suggest the possibility that age-associated cellular damage is due in part to a decline in autophagy, leading to damaged mitochondria, which in turn produce more ROS. Elevated ROS would then be expected to perpetuate the cycle of damage. Another possibility is that ROS is elevated with age, independent of changes in autophagy, and this leads to damaged mitochondria, which are not replaced effectively due to reduced levels of autophagy in aged cells. The accumulation of damaged mitochondria would then be expected to accentuate ROS production (128).

Because osteocytes are long-lived postmitotic cells that can only be replaced by bone turnover, it is reasonable to hypothesize that autophagy plays an important role in their survival and function. Specifically, autophagy may help osteocytes defend against stresses such as elevated ROS. To address the role of autophagy in osteocytes, we have deleted Atg7 in osteocytes using the dentin matrix protein 1 (Dmp1)-Cre transgenic mouse developed by Feng and colleagues (68). Preliminary analyses of these mice revealed that they have low bone mass at 6 months of age that is associated with low bone remodeling (129). The magnitude of these changes was similar to that observed when comparing old (>18 months old) and young (<8 months old) mice (10). Importantly, oxidative stress was elevated in the bones of mice lacking autophagy in osteocytes, as measured by ROS production in bone marrow cells and by p66^shc phosphorylation in bone (130). Thus, suppression of autophagy in osteocytes was sufficient to cause skeletal changes in young adult mice that are similar to those observed in aged wild-type mice. Additional studies will be required to identify the molecular mechanisms by which autophagy in osteocytes controls bone remodeling and bone mass. Perhaps more importantly, it will be important to determine whether autophagy does indeed decline with age in osteocytes or cells at any stage of osteoblast differentiation.

Recent evidence suggests that autophagy may also be important for the long-term health of progenitors as well as fully differentiated long-lived cell types. Specifically, deletion of Atg7 from hematopoietic stem cells (HSCs) in mice led to accumulation of mitochondria and increased oxidative stress in the HSCs, which was associated with increased proliferation and DNA damage (131). These results led to the conclusion that autophagy is required for maintenance of the HSC compartment in adult mice. Another study has shown that the basal level of autophagy is elevated in HSCs of old mice and is required for the survival of these cells under stressful conditions, such as nutrient deprivation (132). Importantly, FoxO3 is critical for expression of autophagy genes and the induction of autophagy in response to stress in HSCs. Thus, genes encoding the machinery for the autophagic response are important targets for FoxOs in HSCs and other cell types. It is possible that autophagy may play a similar role in the maintenance of MSCs in bone and that changes in autophagy in this compartment may underlie changes in MSC behavior with age.

Cell Extrinsic Mechanisms of Skeletal Aging

Loss of Sex Steroids

In women, the loss of bone occurs at a faster rate after the menopause, attesting to the adverse role of estrogen deficiency on bone mass and its contribution to the acceleration of skeletal involution with age. Estrogen deficiency causes an increase in bone remodeling, increased osteoclastogenesis and osteoblastogenesis, increased osteoclast and osteoblast numbers, and increased resorption and formation—albeit unbalanced. Conversely, estrogens slow the rate of bone remodeling and promote a positive balance between bone formation and resorption by attenuating the generation of osteoclast and osteoblast progenitors in the bone marrow and exerting a proapoptotic effect on osteoclasts and an antiapoptotic effect on osteoblasts and osteocytes (41,43,133). Estrogens decrease osteoclast generation and life span via direct effects on osteoclasts mediated by the estrogen receptor α (134,135). These actions are responsible for the protective effects of estrogens on the cancellous bone compartment. On the other hand, the estrogen receptor α present in osteoblast progenitors mediates a protective effect of estrogens against endocortical bone resorption (136). Attenuation of the production of cytokines like IL6, IL1, TNFα, and RANKL by other cells present in the bone marrow environment, including T- and B lymphocytes and stromal/osteoblastic cells, also contribute to the antiosteoclastogenic properties of estrogens (137,138).

Like aging, gonadectomy in female or male mice promotes an increase in oxidative stress in bone (43). Administration of antioxidants, similar to estrogens or androgens, abrogates the stimulatory effects of gonadectomy on oxidative stress, osteoclastogenesis, and osteoblast and osteocytes apoptosis, and it prevents the loss of bone mass (10,90). Hence, sex steroid deficiency may contribute to age-related bone loss, at least in part, by increasing oxidative stress.

Relative Excess of Endogenous Glucocorticoids

Administration of glucocorticoids is a common cause of osteoporosis (139). Glucocorticoids are strong inhibitors of bone formation, at least in part, by stimulating osteoblast apoptosis. ROS generation and activation of a PKCβ/p66^shc/JNK signaling cascade are responsible for the proapoptotic effects of glucocorticoids on osteoblastic cells (140). Moreover, glucocorticoids promote osteocyte apoptosis via activation of Pyk2 and JNK, followed by inside-out signaling that leads to cell detachment–induced apoptosis or anoikis (141).

Glucocorticoids also suppress the generation of new osteoblasts by attenuating Wnt signaling. This occurs via at least three separate mechanisms: first, glucocorticoids stimulate expression of the Wnt inhibitor DKK1 in bone and osteoblastic cell cultures (142); second, glucocorticoids suppress the PI3K/Akt/GSK3β signaling pathway (143); and third, suppression of Akt leads to activation of FoxO transcription factors, which antagonize Wnt/β-catenin signaling (140).

Glucocorticoids also modulate bone resorption and increase cortical porosity. The rapid loss of bone caused by glucocorticoid excess results from direct prosurvival actions on osteoclasts (144). The short-term transient increase in bone resorption in combination with long-term impairment of bone formation might be responsible for the increased cortical porosity in patients treated long term with glucocorticoids (145).

Importantly, the production of endogenous glucocorticoids as well as the sensitivity of bone cells to glucocorticoids increases with age (146,147). In line with this evidence, mice with osteoblast/osteocyte-specific transgenic expression of 11β-HSD2, the enzyme that inactivates glucocorticoids, are partially protected from the adverse effects of aging on osteoblast and osteocyte apoptosis, bone formation rate, and bone strength (147). Thus, the rise in endogenous glucocorticoids represents another age-associated pathogenetic mechanism of involutional osteoporosis.

Marrow Adipogenesis and Lipid Oxidation

With advancing age, the number of adipocytes in the bone marrow increases dramatically, particularly in long bones, in both humans and rodents. In line with the changes in adiposity, the expression of peroxisome proliferator-activated receptor (PPAR) γ2, a transcription factor that is essential for adipogenesis, is increased with age in the bone of mice (30,35). Osteoblasts and bone marrow adipocytes arise from a common progenitor cell and lineage allocation into adipocytes or osteoblasts is considered to be reciprocally exclusive (Figure 1). In view of these findings, it has been proposed that enhanced PPARγ2 expression in progenitor cells plays an important role in the pathogenesis of age-related bone loss by favoring the commitment of mesenchymal progenitors to the adipocyte instead of the osteoblast lineage. In support of this hypothesis is the evidence that mice haploinsufficient for PPARγ2 exhibit increased osteoblast number and high bone mass (148).

Entry of mesenchymal progenitor cells into the osteoblastic lineage is dependent on Wnt/β-catenin signaling. In addition to its role in osteoblastogenesis, Wnt signaling is a strong inhibitor of adipogenesis. Wnt/β-catenin signaling inhibits adipogenesis, at least in part, by suppressing PPARγ expression (149,150). In agreement with this, deletion of β-catenin in osteoprogenitor cells in adult mice increases the expression of PPARγ2 and bone marrow adiposity and decreases bone mass (151). On the other hand, activation of PPARγ by ligands such as oxidized polyunsaturated fatty acids promotes PPARγ association with β-catenin and induces β-catenin degradation (30,152), thereby decreasing β-catenin/TCF-mediated transcription. The inverse relationship between adipocyte and osteoblast differentiation notwithstanding, further studies are needed to demonstrate a causal role of the increased marrow adipogenesis to the decrease in bone mass with aging.

Lipid oxidation plays a critical role in the development of atherogenesis and increases with age in bone. In the process of lipid oxidation, lipoxygenases, such as Alox15, oxidize polyunsaturated fatty acids to form products that bind to and activate PPARγ and generate prooxidants like 4-HNE (153). Mice lacking Alox15 exhibit increased bone mass (154), suggesting that lipid oxidation exerts a deleterious effect on bone. In line with this, both 4-HNE, via FoxO activation, and oxidized polyunsaturated fatty acids, via PPARγ, attenuate β-catenin/TCF-mediated transcription and osteoblast generation (30). Oxidized lipids also stimulate apoptosis of osteoblastic cells and inhibit BMP-2-induced osteoblast differentiation via ROS-independent mechanisms (155–157). Thus, lipid oxidation potentiates oxidative stress and may contribute to the reduction in bone formation that occurs with aging.

Decreased Growth Factors

Growth factors like growth hormone and insulin-like growth factor 1 (IGF-1) are critical for skeletal growth, and IGF-1 is the most abundant growth factor deposited in the bone matrix throughout life (158). The bioactivity of IGF-1 is regulated by the interaction with IGF-binding proteins (IGFBP-1–6) (159). Clinical studies along with murine models have suggested that physiological levels of IGFBPs contribute to the attainment and maintenance of optimal bone mass, at least in part, by targeting IGF-1 to the skeleton (160). IGF-1 promotes the differentiation of osteoprogenitor cells through activation of a PI3K/Akt/mTOR pathway (39,161). In line with this, deletion of the IGF-1 receptor in osteoblasts decreases bone formation rate and cancellous bone volume (39,162). In addition, IGF-1 is the primary osteogenic factor released from the bone matrix via the resorptive action of osteoclast.

In humans, the levels of IGF-1 in the circulation and bone matrix decline substantially with age, most likely due to a reduction in the secretion of growth hormone. Indeed, between 20 and 60 years of age, the content of IGF-1 in human bones declines by 60% (163). Importantly, the levels of IGF-1 and IGFBP3 in the bone matrix are also correlated with the decrease in bone mineral density with aging and with increases in the total number of hip fractures (164,165). Similar to humans, the reduction in the levels of IGF-1 in bone marrow and matrix from 8 to 20 months of age in rats is associated with the decline in bone mass (39). Notably, injections of IGF-1 combined with IGFBP3, the predominant IGFBP, in the long bones of aged rats stimulates bone formation and increases cancellous bone mass. It is, therefore, possible that a decrease in IGF-1 released from the bone matrix, most likely due to a reduction in IGF-1 content along with decreased resorption, contributes to the age-associated decline in bone mass.

Other Age-Related Changes

Chronic inflammation and reduced physical activity are additional extrinsic factors that are often associated with human aging and potentially mediate some of the negative impacts of aging on the skeleton (166,167). Although inflammatory cytokines clearly play an important role in the bone loss caused by sex steroid deficiency, it is less clear that they are involved in age-associated bone loss, at least in rodents. Specifically, inflammatory bone loss is associated with elevated osteoclast formation in cancellous bone, whereas osteoclast number is low in the cancellous bone of aged mice compared with young mice (10,168). Thus the histological picture of aged bone is not consistent with a role for chronic inflammation. Similarly, biomechanical unloading causes bone loss, at least in part, via increased resorption. Thus, the histological changes that occur in the aged skeleton also do not appear to be caused by unloading. Be that as it may, it remains likely that chronic inflammation and reduced physical activity do contribute to osteoporosis in certain individuals, even though these mechanisms may not be generally involved.

Conclusions

Although a decline in bone formation and loss of bone mass are common features of human aging, the molecular mechanisms mediating these effects have remained unclear. Evidence from pharmacological and genetic studies in mice has provided support for a deleterious effect of oxidative stress in bone and has strengthened the idea that an increase in ROS with advancing age represents a pathophysiological mechanism for age-related osteoporosis. The contribution of a decline in mitochondrial function and an increase in ROS to age-related diseases and life span notwithstanding, other mechanisms including deregulated autophagy, telomere shortening, and abnormal protein folding are likely to represent additional critical effectors of aging. Furthermore, several cell extrinsic mechanisms, including a decline in sex steroids, an increase in endogenous glucocorticoids, and higher lipid oxidation are associated with the decreased bone formation seen in old age (Figure 2). Thus, the cellular changes that occur in the skeleton with advancing age appear to result from numerous independent mechanisms, although elevated oxidative stress may be a common component of many of them. It will now be important to determine whether suppression of oxidative stress in specific cell populations can ameliorate any aspects of skeletal aging.

Although the long-lived cells in bone, such as the adult MSCs and osteocytes, are more prone to suffer the negative effects of aging, short-lived cells like osteoblast progenitors and mature osteoblasts are also affected by the altered aged environment. Understanding how different anti- or pro-aging molecules exert their effects at different stages of osteoblast differentiation will be challenging but crucial for the elucidation of the signaling networks operating in aged bone. Deciphering such ageing networks, together with identification of the cells most affected by the ageing mechanisms, should provide the information necessary for the advancement and development of therapeutic strategies to prevent or treat age-related osteoporosis.

Funding

This work was supported by the National Institutes of Health (R01 AR56679, P01 AG13918, and R01 AR49794), the Department of Veterans Affairs from the Biomedical Laboratory Research and Development Service of the VA Office of Research and Development to C.A.O. (5I01BX000294), the University of Arkansas for Medical Sciences (UAMS) Translational Research Institute (UL1 RR029884), and Tobacco Settlement funds.