Mass, architecture, and shape of cortical and trabecular bone compartments show variation with sex and age.[1-4] Men and women show evidence of skeletal dimorphism that extends through life.[5] During growth, long bones increase in length and diameter, [4] with the male skeleton attaining expansion in cortical thickness and perimeter via periosteal bone accrual, while in women this involves relatively limited periosteal deposition with more endosteal bone gains.[3,5] These greater periosteal formation rates in men, yield a skeleton with wider cortical diameter and greater strength that is attributed to differences in size and geometry, rather than greater bone mineralisation density.[2,4] These structural differences appear to be linked to bone’s sexually dimorphic mechanoadaptive capacity. Thus, although strenuous load-bearing exercise increases tibia density, size, and strength in both men and women, only women exhibit increased metaphyseal trabecular volumetric bone mineral density, whilst males instead exhibit more profound cortical diaphyseal increases.[6] The extent to which sexual dimorphic mechanoadaptive behaviour impacts age-related shifts in cortical and trabecular mass and architecture in men and women remains, however, ill-defined.

Sexual-dimorphism in skeletal behaviour is also evident in bone loss processes, where resorption-related catabolism is found to be modular, with cortical and trabecular compartments even within a single bone showing different sensitivity.[7] Bone loss in men and women is evident upon aging, with endosteal resorption linked to enhanced periosteal deposition shifting force distribution across the diaphysis to improve strength.[2,4] Men, however, lose less cortical bone than women because periosteal bone formation is higher, and not because resorption rates are higher in women.[4] Thus, aging men and women show similar bone loss by endosteal resorption (40% and 46% respectively), but the diminution in cortical strength is much reduced in males, as the remaining bone is distributed further from the neutral axis than in women.[2-4] Trabecular bone also exhibits sexually dimorphic behaviour, which in females is likely linked to the metabolic requirements of reproductive function, [2,8,9] as this compartment must serve as a mineral reservoir during pregnancy and lactation.[2,8] When such resorption is sparked in females, (re) modeling processes target horizontal trabeculae to engender a decline in connectivity.[2,8] In males, trabecular resorption instead tends to decrease trabecular thickness and number rather than to reduce connectivity.[2,9] These bone loss processes in cortical and trabecular bone precipitate osteoporosis, which has a twofold and 4.7-fold greater prevalence in women aged 50 to 64 and over 65 years, respectively.[1] It is often implied that this lower cortical and trabecular bone mass in females is prima facie evidence for sexually dimorphic mechanoadaptive behaviour, but direct support for this notion is not extensive.

To characterize load responses, many studies use pre-clinical mouse models.[10-13] Their utility is partly due to their conservation of the sexual dimorphism seen in humans, with higher tibial cortical cross-sectional areas in males than females, at both young and old age.[11,12,14] The model we developed and employ herein, involves non-surgical application of cyclic compressive loads to the mouse tibia via maximally-flexed knee and ankle joints to provoke anti-resorptive and pro-formative stimuli.[11,12,15] We have found that load-related strain patterns do not directly correlate with new bone accrual, as strain magnitudes increase with age whilst osteogenesis does not.[11,14,16] Early studies in young female mice found that six load episodes applied over two weeks elicit increases in mass and shape changes in the cortex, and increases in metaphyseal trabecular thickness and number.[4,11-13,15] Earlier studies have shown that these mechanoadaptive responses in murine tibiae differ between the sexes, yet the shifts with age fail however to fully explain the sexually dimorphic bone architectures.[11-18]

Use of this pre-clinical mouse tibia loading model to yield more complete awareness of mechanoadaptive bone behaviour in male mice and how it is modified during growth, maturation, and aging will help in the elucidation of mechanisms to prevent sex- and age-specific bone loss. This study evaluates whether age confers divergent degrees of mechanosensitivity upon cortical and trabecular bone compartments in male mice. Having previously described such shifts in female mice, [11-13,18,19] we focus herein upon the adaptive load-induced cortical and trabecular bone responses in the tibia of male mice at three ages and find that these show modular, bone compartment-related sensitivity and that they are at least partly retained even with aging.

These findings demonstrate that sensitivity to mechanical load is evident through maturation and aging in the tibial cortical bone in male mice but that the trabecular compartment, within the same bone, in these male mice is by contrast relatively insensitive to load at all ages. As we have previously generated multiple datasets from female mice across similar ages utilizing our tibial model, and we have used a conserved loading regime herein which generates 2,000 μᶓ on the medial tibial surface, we will therefore also consider our new findings in the context of our previous results from female mice of varying ages - young, mature, and aged - to explore whether there is clear evidence for sexual dimorphism in the age-related shift in these mechanoadaptive responses.[11-13,18,19] Although these experiments were not contemporaneous, we acknowledge that the comparison can only be indirect. However, we emphasize that these data derive from a single set of researchers using the same methodology, loading machine, and protocols throughout. Nonetheless, caution is warranted in integration our new findings in males across a range of ages, especially in the context of previously published work exploring mechanoadaptive responses in young, mature, and aged females, as these data were not derived from a single experimental study.[11-13,18,19]

Our data are consistent with the hypothesis that load responses in cortical and trabecular compartments, even in a single bone, differ and that they exhibit only little age-related decline in males. Our earlier research had demonstrated that young (10/12-week-old) female mice exhibit a marked and widespread cortical bone accrual response to tibial loading, but that these load-induced responses are, however, more spatially-restricted along the tibia length in mature (22 weeks) and completely absent in aged (18-month-old) female mice.[11-13,18] The metaphyseal trabecular compartment in the tibiae of females, similarly showed architectural increases in mass in young mice, which were absent in both mature and aged female mice, [12,18] indicating a generalized, non-modular, waning in the scale of mechanoadaptation with aging in females. We find, herein, a marked increase in cortical bone formation at each of three defined locations along the tibia length in young male mice, and also in the tibiae of mature mice; although it was lacking in the most distal regions. Intriguingly, endosteal and total new cortical bone formation was also significantly enhanced in loaded tibiae (vs. left) in male 18-month-old mice, whilst loading failed to elicit any modification in trabecular mass or architecture in male mice at any age. Together, these findings suggest that sensitivity to mechanical loading is better maintained in the cortical bone of aging males yet is totally abolished in female mice. [11-13,18] On the other hand, the trabecular compartment in male mice appeared to behave modularly with relative insensitivity to load at all ages, while females exhibited an age-related decline with load-induced trabecular changes only in the young.[11-13,18]

Sexual dimorphism in mechanoadaptation has been elegantly examined before.[14,20] Meakin et al. [20] reported that mechanical loading (producing strains of 2,000 or 2,500 μᶓ) was capable of modifying cortical area at a specific location (37% of the tibia length) in young 16-week-old and 19-month-old male, as well as female C57BL/6 mice, and that only the scale of these changes was lessened by age in both sexes. Galea et al. [14] also explored sex differences across almost the entire tibia length in 19-week-old and 19-month-old mice in response to loading (engendering 2,500 μᶓ), to find site-specific sex differences in young and an abrogated load-induced increase in periosteal enclosed cortical and marrow areas in the aged mice. Our work in young males aligns with these earlier studies, yet our finding that cortical insensitivity to load fails to emerge in aged male mice, does not.[14,20] This may be due to the higher strain levels applied in these early studies or due to the analytical methods that reported percentage change instead of direct left/right (non-loaded vs. loaded) comparisons.

Unlike in cortical bone, changes in trabecular organization in response to loading (peak strain) were lacking in male mice. This juxtaposes with previous data in female mice [18,19] where robust trabecular responses to loading observed in young mice are consistent with reports of a marked osteogenic response to exercise in pre-pubertal girls. These sex differences in trabecular bone’s response to loading raise interesting possibilities related to the very divergent role of the skeleton as a mineral reservoir for reproduction purposes in females.[8] Such greater sensitivities in trabecular bone in females are supported by the studies of Ko et al. [21] which report more resorption and trabecular bone loss in growing female (6-week-old) mice than males after 2 weeks of neurectomy-induced disuse. Sexually dimorphic behavior is also evident in osteoclast-mediated resorptive targeting of trabecular bone, with males being prone to greater trabecular thinning and females showing more marked loss in connectivity due to targeted horizontal trabecular resorption.[2,8] It is important to emphasize that the human male skeleton shows the greater trabecular area, density, and connectivity than age-matched females, presumably making the trabecular compartment stronger and more ‘resistant’ to similar load-induced strains engendered on the cortical surfaces [2,5]; where they are calibrated. Whilst this may explain the observed paucity of trabecular loading response in male mice, the cellular mechanisms underpinning this modularized, compartment-specific, sexually dimorphic mechanoadaptation remain elusive.

Our data impact appreciation of the relationship between mechanical load environments and bone architecture. Thus, there has been much debate focused on estrogen’s role in maintaining bone mass, and the contribution made to postmenopausal bone frailty by its loss. Two arguments seem to dominate this debate. In the first, it is proposed that loss of estrogen (or its receptor) renders bone relatively insensitive to mechanical loading and therefore precipitates its structural demise through a failure in mechanoadaptive processes.[22,23] The second disputes this and offers a contradictory argument, which proposes that estrogen-only contributes to retaining bone mass for reproductive purposes [24]; thus, post-menopausal estrogen loss will also explain the diminished bone mass predisposing the aged murine female skeleton to fracture.

These arguments fail, however, to recognize the possibility that mechanoadaptive responses and the ‘reserve’ bone reproductive capacity may be preferentially segregated across distinct cortical and trabecular compartments, respectively. Our studies demonstrate differences in the response of cortical and trabecular bone in the tibia to applied loads, and that they are not always conserved in males and females. Thus, while tibiae in males appear to express similar cortical bone responses at all ages, female mice exhibit increases in cortical bone formation in young, mature but not the aged. In contrast, loading does not appear to modify trabecular bone in males. Our studies support an alternative notion that may accommodate both arguments. We propose that trabecular bone in females fulfills a predominantly reproductive function, but its contribution to bone strength is evidenced by its sensitivity to loading; thus, pre-menopausal estrogen promotes the generation of a bone reserve for reproduction and loading acts to adapt its structure.[8] In the post-menopausal phase, trabecular bone is no longer required as a reproductive reserve and fails to be conserved by the actions of load-bearing acting alone. This aligns with our finding that trabecular bone in males is relatively insensitive to loading, even in younger reproductively active animals; a reproductive reserve is never required in males. On the other hand, cortical responses are similar in both sexes, but more subject to changes in sensitivity with aging in females. Thus, cortical bone has a predominantly mechanical role in both sexes.

Several works have explored the potential cellular and molecular mechanism that link abnormalities in Wnt ligands, and Wnt receptor signaling to bone formation.[25-28] Research has shown that deficient Wnt ligand signaling can delay skeletal ossification and osteoblast differentiation including decreased trabecular bone mass and consequently reduced bone formation.[25,26,28] The loss of Wnt receptor function results in increased osteoclastogenesis, reduced trabecular volume and cortical perimeter and can be correlated with low bone mass at birth.[25,26] Different from males, females present a sharp drop in estrogen levels with a concomitant increase in sclerotin production. This protein is an antagonist of Wnt signaling leading to bone loss.[25-27] Thus, apart from the differences in bone anatomy between the sexes, the female skeleton may also experience more deleterious effects upon disturbed Wnt signaling when estrogen levels decline.[25,26]

There are caveats to acknowledge in our work. Any comparison across sexes is indirect, as we assess our new data from males in the light of our previous findings in females. To confirm sexually dimorphic mechanoadaptation, further studies should make direct comparisons within a single, albeit very large and likely complex experiment across various ages. Despite our care to strain-calibrate applied loads, it remains possible that the divergent responses may be underpinned by higher tibial cortical mass in males.[14] Nonetheless, we consider it worthwhile making these female: male comparisons, as we have used an identical loading regimen, loading cup design, within identical conditions. This is important as outcomes are known to deviate between laboratories using similar loading protocols and seemingly identical models, [29] which makes the distinctions between males (examined herein) and females (reported previously [11-13,18,19] more valuable than they might be otherwise).

It would also have been ideal if measurements were made in the same location-matched proximal, midshaft and distal tibial regions at all ages. We reported endosteal and periosteal formation only from aged mice, however, the potential to generate these data at other ages is no longer possible, and further work may instead repeat the entire set of experiments to allow for parallel comparisons. The effect of loading was nonetheless found to extend across proximal to distal tibial portions and this declines with aging in mature mice, making it highly likely that the lack of response in the distal segment in mature males will translate to aged males. An advance made possible by our findings would be an exploration of the osteoblast or osteocyte behaviours that underpin these sexually dimorphic shifts in bone mechanoadaptation with aging. Aged bone targets endosteal bone surface with an increased number of visible resorption pits and increases osteoclastic activity.[30] As we have shown that loading enhances endocortical bone in aged animals, this could be used to target therapeutic approaches that align to the specific sites of skeletal frailty found in males and females. Earlier studies highlight this potential in their pinpointing of sex-related decreases in female osteoblasts, and no change in osteocytes in response to mechanical loading.[20] The load-induced cortical bone modifications we report may inform age-related architectural skeletal changes seen in men and women. Thus, the wider cortices found in the bones of an aging males may, at least partly, be the product of a relatively undiminished mechanoadaptive response. In contrast, raised fracture risk linked to narrower cortices in females [2] may reflect a diminished mechanoadaptive capacity in the cortical compartment of aged female bones.

Our findings indicate modular mechanosensitivity in the cortical and trabecular bone compartments in males that is subject to only modest decline with aging, which may underpin the architectural features of male skeletal aging.