Senescent and apoptotic osteocytes and aging: Exercise to the rescue?

Abstract

Osteocytes are the most prevalent cell in the skeleton and are the master regulator of bone remodeling. Despite the understanding that osteocytes have a multiyear lifespan, and some factors induce apoptosis in osteocytes, much less is understood about the induction and consequences of osteocyte senescence. Filling these gaps in knowledge will provide novel approaches to slowing age-related bone loss and preventing fragility fractures. The purpose of this review is to examine the roles of senescence and apoptosis in osteocytes in age-related bone loss. Based on evidence that exercise can prevent senescence in skeletal muscle, we provide a novel hypothesis by which exercise can prolong skeletal health.

Introduction

Preventing bone loss with aging is of considerable public health interest because of the cost, morbidity and mortality associated with osteoporosis and fragility fractures [1–4]. Osteocytes are the primary mechanosensor of the skeleton and the master regulator of bone remodeling [5]. As such, the uncoupling of bone resorption and bone formation that leads to age-related bone loss can often be traced back to changes in osteocyte number or function. There has been a resurgence of investigations into cellular metabolism of osteoblasts, osteoclasts, and osteocytes, and their contributing roles in integrated physiology and whole body energy balance. The increased appreciation that osteocytes are metabolically active creates a need to reexamine how osteocytes adapt, senesce, or die in response to different metabolic environments across its multi-year lifespan [6–9]. By applying what has been gleaned about senescence in other cell types, reexamining the lifespan of osteocytes could yield paradigm shifts in our understanding of age-related bone loss. In muscle, there is a renewed focus on the role of age-related changes in mitochondrial number and function on the induction of senescence. Age-related changes in mitochondrial function that occur in muscle and could occur in osteocytes include loss of mitophagy, changes in fission and fusion, increased mitochondrial uncoupling and superoxide production [10–12]. Within this context, physical activity and exercise programs are often utilized as non-pharmacologic therapeutics for preserving mitochondrial number and function and preventing senescence. However, the potential consequences of senescence and mitochondrial dysfunction with aging is underappreciated in osteocytes. The purpose of this review is to examine the roles of senescence in osteocytes in age-related bone loss, and provide novel hypotheses by which exercise can prolong skeletal health.

The notion of senescence in osteocytes

Osteocytes arise from terminally differentiated osteoblasts that become imbedded in the bone matrix over time. More than 90% of the cells in bone are osteocytes, so understanding their life cycle is essential. Seminal work from the Bonewald lab over two decades has defined most of what we know about the origin and function of these cells [5]. And because osteocytes are in communication with cells on the bone surface, and are bathed in extracellular fluid, much has been made of their ‘command and control’ function within the remodeling unit. But those tasks require osteocyte survival. Most studies on osteocyte viability have focused on inducing or preventing apoptosis, which in osteocytes, is an event that promotes bone loss through RANKL activation and sclerostin up-regulation. Senescence, on the other hand, is a cell fate program that is part of a DNA Damage Response (DDR) to prevent replication of maladaptive DNA mutation [13]. Senescence occurs throughout the lifespan, but senescent cells particularly accumulate with aging in other tissues. These death resistant cells can cause significant damage in surrounding tissues through release of various cytokines. Indeed, senescent cells are detected and defined based on the presence of that pro-inflammatory secretome, often referred to as senescence-associated secretory phenotype (SASP). Accordingly, senescent vs non-senescent cells can be independently identified based on the secretion of SASP-related factors such as p53, p21, and several interleukins [14–16].

Osteocytes have been traditionally considered ‘old’ osteoblasts buried within the bone matrix; hence the notion that there is a significant population of senescent osteocytes driving bone loss through SASP bone loss is appealing. Furthermore, it is tempting to assume since cell cycle arrest is part of senescence, and osteocytes don’t replicate that a large proportion of osteocytes are senescent. Moreover, reducing the number of senescent cells has been shown to be beneficial in several age-related disorders. This can be accomplished through administration of senolytics, a class of small molecules that can selectively induce death of senescent cells.

Osteocyte senescence and mechanical loading

With the introduction of senolytics, there are clear benefits to elucidating the mechanisms that induce or prevent senescence in osteocytes. This is reinforced by recent evidence indicating that senolytics could be effective for slowing bone loss [17]. There is overlap in age- or damage-related factors that can induce senescence and apoptosis, including the activation of tumor suppressor p53 through oxidative stress, gamma radiation, loss mitochondrial membrane integrity and accompanying release of cytochrome c, and nuclear or mitochondrial DNA degradation. p21, a kinase inhibitor and a major target of p53, induces cell cycle arrest leading to senescence [18, 19]. It is unclear whether senescence in osteocytes results from a development of resistance to apoptosis [20]. Apoptosis of osteocytes can be induced by both unloading and damage-inducing loading [21]. Because senescence is part of DDR, it is conceivable, based on the mechanosensing of osteocytes that damage-inducing loading can induce senescence in osteocytes.

The effect of age-related reductions in the loading environment on osteocyte senescence has not yet been tested. Aging produces changes in mechanical stimuli and the response to stimuli at a tissue and cellular level [16, 22]. Specifically, decreases in physical activity and increases in sedentary time that are common with aging engender a degree of skeletal unloading [23, 24]. Aging and physical inactivity are each associated with a decrease in lacunae that contain osteocytes, loss of directional orientation of lacunae, and the number and length of dendrites that connect through canniculi; all of which would interfere with the detection of strain or fluid-flow shear stress [25–28]. Connexin43, a gap junction protein involved in mechanotransduction, protects from osteocyte apoptosis, but also decreases with aging [29–32]. Alternatively, unloading could induce senescence indirectly through the effect of inactivity on reductions in myokines, adverse changes in whole body nutrient trafficking, and the resulting lipid accumulation in both osteoblasts and osteocytes. Further research will be needed to distinguish the role of (un)loading on senescence.

Regardless of whether unloading directly induces senescence in osteocytes, the loss of mechano-sensitivity, the resulting uncoupling of bone remodeling, and the generation of SASP may be the biggest negative consequences of presumed osteocyte senescence. If SASP promotes further senescence or apoptosis of osteocytes through a prolonged secretory phenotype as it does in other cell types, then apoptosis would be preferable because apoptotic osteocytes are providing a one-time stimulus prior to ingestion by phagocytes. The increased burden of senescent cells would then likely promote additional bone resorption and inhibition of bone formation through the secretion of pro-inflammatory cytokines. The loss of mechanically sensitive osteocytes could result in accelerated bone loss.

Targeting osteocyte senescence with exercise to slow age-related bone loss

Exercise is often recommended for maintaining and improving bone health, in part, because of the demonstrated increases in bone formation rate and bone strength in response to mechanical loading, although gain in bone mass is less apparent [33]. From a mechanocentric view, exercise prescriptions for bone health favor activities that provide higher intensity or faster rates of loading (e.g., resistance training, jumping) to induce important stimuli for existing osteocytes and promote the differentiation of MSCs into the osteoblast lineage, ultimately resulting in more osteocytes. Endurance or aerobic exercise typically provides a smaller-magnitude and rate of loading, and acutely stimulates bone resorption [34]. However, because serum calcium drops precipitously during intense exercise, and PTH rises to compensate acutely, it is likely that osteocytic osteolysis is a major driver of that physiologic compensation [35]. A readily available pool of viable osteocytes that could induce skeletal osteolysis would thus seem to be essential. As noted, aerobic exercise is not often highlighted as an important part of promoting bone density or bone strength. However, aerobic exercise could provide benefits to osteocyte viability in ways that are independent from mechanical loading, including the release of exercise-stimulated myokines, altering of macronutrient trafficking, and preservation of cellular or mitochondrial repair. Unloading, on the other hand, might lead to greater senescence because of the lack of stimulus for promoting viability (Figure 1). Certainly spaceflight studies with rodents could help identify determine if osteocyte senescence is a major component of enhanced bone resorption, particularly since an inflammatory profile is noted during unloading.

Figure 1.

Model of how exercise could slow age-related bone loss by preventing the accumulation of senescent osteocytes. Exercise could prevent osteocytes senescence by inducing the release of myokines, altering nutrient trafficking to prevent lipid overload in bone, providing mechanical loading, and directly or indirectly maintaining mitochondrial number and/or function in osteocytes.

Muscle-bone units have been studied for over 40 years as mechanical units, but in the past few years has evolved into focusing on the biological cross-talk through myokines and osteokines. Several candidate myokines have the potential to influence aspects of bone turnover, as recently reviewed by Bonewald [36]. β-aminoisobutyric acid (BAIBA) and irisin are among the exercise-induced myokines with the potential for directly slowing age-related osteocyte senescence by protecting mitochondrial integrity [37, 38]. Irisin is a peptide cleaved from Fndc5, a muscle surface protein. It is found in the circulation in nanogram concentrations (i.e. in mice and humans) and its levels are increased during and following exercise. It was first noted to be an inducer of thermogenic programs in white adipose tissue. However recent work has shown that irisin can induce sclerostin in osteocytes, as well as RANKL[38]. The receptor for irisin an integrin, alphaV beta 5, was recently discovered and characterized in osteocytes. Moreover, irisin prevents apoptosis of IDG-SW3 osteocytes [38]. As such it appears the major skeletal target for irisin is the osteocyte, although preliminary evidence from our lab suggests irisin may have a direct effect on osteoclasts. Whether irisin prevents osteocyte senescence is unknown. However tantalizing preliminary evidence suggests irisin may have a neuroprotective role.

Fat oxidation and osteocyte viability

Aerobic exercise improves fat oxidation and redirects nutrient trafficking through more energetically expensive pathways [39, 40]. This could translate into preventing excess lipid trafficking to osteocytes, but studies are needed to directly test this hypothesis. Accumulation versus oxidation of lipid is dependent on mitochondrial capacity. Mitochondria are active in MSCs, particularly during early differentiation, osteoclasts, and osteocytes, and both glycolysis and oxidative phosphorylation contribute to the synthesis of ATP [6, 41]. A critical gap in knowledge is whether aerobic or resistance exercise has direct or indirect effects on mitochondrial number, fission, fusion or capacity in osteocytes. If so, exercise could target senescence in osteocytes through the preservation of mitochondrial content, coupling, and capacity, through preservation of biogenesis, mitophagy, fusion, and fission capabilities [42–45]. In muscle, exercise induces mitochondrial biogenesis to a lesser extent with aging, but is still effective in stimulating increases in mitochondrial enzymes and oxidative capacity [10–12].

Further, lifelong physical activity slows the age-related decline in mitochondrial number and capacity [46, 47]. There is less oxidative damage to muscle mitochondrial membranes in response to acute exercise and improved antioxidant capacity with aging in those who chronically exercise [48–50]. There are initial clues to suggest that osteocyte mitochondrial activity is responsive to exercise and/or aging. For instance, when applying mechanical loading in vivo, osteocyte calcium responses to the load, which were used as an indicator or osteocyte recruitment, were related to applied strain magnitude and frequency [51]. Recently, in situ imaging was used to demonstrate that mitochondrial activity is higher in osteocytes close to the periosteal surface and decreases as cells become closer to the endocortical surface. Osteocytes closest to the endocortical surface had greater numbers of nonfunctional mitochondria. This seems counterintuitive to the notion of marrow fat being a source of lipid during periods of fasting,[41] but this location-based variance in mitochondrial content may indirectly reflect either the effect of aging or adaptations to local energetic needs. Genetic deletion of the growth hormone receptor leads to longer lifespan in mice, but we recently showed that mitochondrial function in osteocytes was reduced and ROS was increased [52]. These data would imply that aging itself may result in either damaged osteocytes or senescent osteocytes characterized by impaired mitochondrial dynamics. Whether exercise can prevent those changes remain to be determined.

Summary

Understanding the induction and removal of senescent cells in bone is an important step forward in discovering new targets for slowing age-related bone loss. Much of what is currently hypothesized about senescence in osteocytes, or the potential for exercise to prevent senescence in osteocytes, is derived from decades of experiments in skeletal muscle. Exercise that involve high strains or rates of strain are highly touted for maintaining bone health and reducing fracture risk across the lifespan. However, more work is needed to understand the potential metabolic and anti-senescent benefits of endurance/aerobic exercise or reductions in sedentary behavior on osteocyte function. Further, it is known that mitochondria are important for energy production in all bone cells, and are intricately linked to senescence in other cell types. Determining the osteocytic mitochondrial adaptations to exercise should help bridge the metabolic and mechanical factors to promote bone health throughout the lifespan.

Highlights.

• Osteocyte apoptosis and senescence independently contribute to age-related bone loss.

• Exercise may prevent osteocyte senescence by preserving mitochondrial function.

• Endurance and resistance exercise may have distinct benefits for osteocyte viability.