Hallmarks of frailty and osteosarcopenia in prematurely aged PolgAD257A/D257A mice: a preclinical model to test anabolic interventions

Frailty is a geriatric syndrome characterized by increased susceptibility to adverse health outcomes. One major determinant thereof is the gradual weakening of the musculoskeletal system and the associated osteosarcopenia. To improve our understanding of the underlying pathophysiology and, more importantly, to test potential interventions aimed at counteracting frailty suitable animal models are needed. Here, we report the relevance of a mouse model of accelerated aging (PolgA(D257A/D257A)) as a model for frailty and osteosarcopenia. The longitudinal assessment of the clinical mouse frailty index showed that PolgA(D257A/D257A) mice accumulated health deficits at a higher rate compared to wild type littermates (PolgA(+/+), WT). Concomitantly, PolgA(D257A/D257A) mice displayed progressive musculoskeletal deterioration such as reduced bone and muscle mass as well as impaired functionality thereof. Specifically, PolgA(D257A/D257A) had lower grip-strength and concentric muscle forces as well as reduced bone turnover as assessed by longitudinal micro-CT. In addition, PolgA(D257A/D257A) mutation altered the sensitivity to anabolic stimuli in skeletal muscle, muscle progenitors and bone. While, compared to WT, PolgA(D257A/D257A) caudal vertebrae were not responsive to a cyclic loading regime, PolgA(D257A/D257A) muscles were hypersensitive to eccentric contractions as well as leucine administration, shown by larger downstream signaling response of the mechanistic target of rapamycin complex 1 (mTORC1). However, myogenic progenitors cultured in vitro showed severe anabolic resistance to leucine and robust impairments in cell proliferation. Overall, PolgA(D257A/D257A) mutation leads to hallmarks of age-related frailty and osteosarcopenia as observed in humans and thus, provides a powerful model to better understand the relationship between frailty and the aging musculoskeletal system.


52
Although there is no universally accepted definition of frailty [1], it is considered as an age- 53 related syndrome characterized by the decline of multiple physiological functions, leading to 54 the accumulation of health deficits, and thus a higher vulnerability to adverse health outcomes 55 such as morbidity and mortality [2]. One of the most prominent components of frailty is the 56 progressive weakening of the musculoskeletal system [3-5], leading to common age-related 57 diseases such as osteopenia and sarcopenia. There is growing evidence that both diseases often 58 co-exist in frail older individuals (also termed osteosarcopenia [6, 7]), thereby further 59 increasing the risk for negative outcomes such as falls and fractures [8,9]. Although several 60 anabolic interventions such as dietary protein supplementation and mechanical stimulation are 61 known to promote muscle and bone formation in young individuals, the molecular insights 62 behind osteopenia and sarcopenia in the elderly population are lacking. 63 In the field of muscle physiology, studies in humans and rodents have shown that aged muscles 64 are less responsive to well-known anabolic stimuli such as amino acids [10-12] and muscle 65 contractions [13]; this phenomenon, termed "anabolic resistance", likely results from reduced 66 protein synthesis due to diminished intracellular signaling through the mechanistic target of 67 rapamycin complex 1 (mTORC1) pathway [14][15][16]. Next to impairments in intra-muscular 68 mTORC1 signaling, age-related sarcopenia has been associated with a decrease in number [17,69 18] and proliferation capacity [19,20] of myogenic progenitors or satellite cells. These are not 70 only instrumental for the maintenance of muscle fibers, but also for the adaptive responses to 71 exercise and regeneration upon injury [21]. In the field of bone physiology, evidence pointing 72 towards altered mechanosensitivity with age has also been shown in humans [22] and in mice 73 [23][24][25][26]. However, this effect might be site-specific as studies using a tibia-loading model 74 showed a reduced response of trabecular [23,24] and cortical [25,26] bone formation with age, 75 while bone adaptation in response to loading of the caudal vertebrae was maintained with age 76 [27]. 77 Therefore, whether and how age-related changes in the responsiveness to anabolic stimuli occur 78 remains unclear. A better understanding of the pathophysiology of osteosarcopenia will help to 79 identify interventions to strengthen the musculoskeletal system, which ultimately will be 80 beneficial for the prevention and/or treatment of frailty. 81 In order to address this, tools such as the frailty index (FI) have been established to quantify the 82 accumulation of age-related health deficits (e.g., loss of hearing, tremoring, comorbid diseases) 83 in humans [28] and more recently, also in mice [29]. Indeed, the striking similarities between 84 key features of the FI scores in humans and in mice [30] have highlighted the potential of rodent 85 frailty models to not only improve our understanding of frailty but also serve as a tool to test 86 responses to interventions designed to modify (or even prevent) frailty [31][32][33]. In this study, 87 we aimed to evaluate the PolgA (D257A/D257A) mouse (referred to as PolgA), which due to elevated 88 mitochondrial DNA point mutations and systemic mitochondrial dysfunction, exhibits an 89 accelerated aging phenotype [34,35], as a model of frailty and osteosarcopenia. While these 90 mice are known to develop multiple signs of aging (e.g., hair loss, greying, hearing loss) earlier 91 (around 40 weeks of age) than their wild type littermates (PolgA +/+ , referred to as WT), the frailty 92 phenotype has to the best of our knowledge not yet been assessed in these mice. Furthermore, 93 although several studies have reported lower muscle weights in PolgA mice compared to their 94 WT littermates [34,36,37], little is known about their muscle quality and functionality. To 95 address this, forelimb grip-strength and concentric muscle forces were measured in vivo and ex 96 vivo, respectively, in addition to the evaluation of hind limb muscle masses. Furthermore, the 97 response to acute anabolic stimuli such as eccentric contractions and the leucine administration 98 were assessed. With respect to the bone phenotype, only two studies have reported reduced 99 femoral bone density using X-ray densitometry [34,35]. Although this technique is still the 100 gold-standard to assess bone mineral density (i.e. bone quantity) clinically in humans, it does 101 not provide insight regarding the quality of bone tissue. Therefore, bone phenotyping in pre-102 clinical studies is commonly performed using high-resolution micro-computed tomography 103 (micro-CT) as it allows additional standardized evaluation of the three-dimensional bone  forepaws, they were pulled horizontally at the base of their tail until they let go of the bar. The 155 process was repeated 5 times to determine the average peak grip force value (gram-force) used 156 for analysis. All measurements were performed by the same experienced user.    Table S2). Homogenates were centrifuged at 10'000 g for 10 min at 199 4°C. Supernatant was collected and protein concentration was measured using the DC protein assay kit. 200 10-25 µg of total protein was loaded in a 15-well pre-casted gradient gel (Bio-rad, 456-8086). After 201 electrophoresis, a picture of the gel was taken under UV-light to determine protein loading using strain-202 free technology. Proteins were transferred via semi-dry transfer onto a PVDF membrane (Bio-rad, 170-203 4156) and subsequently blocked for 1 h at room temperature with 5% milk in TBS-Tween. Membranes 204 were incubated overnight at 4 °C with primary antibodies (listed in SM, Table S2). The appropriate 205 secondary antibodies for anti-rabbit and anti-mouse IgG HRP-linked antibodies (SM , Table S2) were 206 used for chemiluminescent detection of proteins. Membranes were scanned with a chemidoc imaging 207 system (Bio-rad) and quantified using Image lab software (Bio-rad). 208

Micro-CT imaging and analysis 209
For analysis of femoral, the right femora were harvested, placed in 70% Ethanol and scanned  Micro-CT data was processed and standard bone microstructural parameters were calculated in 221 trabecular, cortical and whole bone by using automatically selected masks for these regions as      Figure 2A). Furthermore, when the EDL was subjected to a 299 force-frequency protocol, PolgA showed a tendency towards decreased absolute force at 250 300 and 300Hz (p<0.10), while relative force was not affected ( Figure 2B,C). At 46 weeks, the grip-301 strength was 11% lower in PolgA mice compared to WT, but did not reach significance ( Figure   302 2D). The differences in forces however, were more exacerbated with a lower absolute and 303 relative force at 80 to 300Hz (absolute) and 250 to 300Hz (relative) in the PolgA (p<0.05, 304 Figure 2E,F).     we are the first to report the frailty phenotyping in the PolgA mouse model.

473
In line with increased FI with age, the musculoskeletal phenotype clearly diverged over time 474 between genotypes, with PolgA mice developing multiple signs of osteosarcopenia with age.

486
One mechanism commonly thought of as a contributor to sarcopenia is the age-related lowered 487 sensitivity to anabolic stimuli. This phenomenon, known as anabolic resistance, has been at 20 weeks to lower in PolgA mice with age. Interestingly though, this difference did not arise 543 due to significant bone loss in PolgA mice, but rather in the failure to achieve normal peak bone 544 mass. We and others have previously reported the absence of age-related bone loss in caudal 545 vertebrae [72, 73]. Hence, mouse caudal vertebrae may not be optimal for investigating age-546 related bone loss. As the bone morphometric parameters in the femora declined with age in 547 PolgA mice, this study suggests that age-related changes in bone micro-architecture differ 548 depending on the skeletal site that is analyzed. Indeed, a more severe age-related deterioration 549 of bone microarchitecture in long bones compared to lumbar vertebrae has previously been with our previous observation that caudal vertebrae remain mechanosensitive to cyclic 584 mechanical loading with age [27]. Interestingly, in that study, 82-week-old C57BL/6J mice 585 showed a greater anabolic response compared to the 52-week-old mice; hence, it is possible 586 that even older PolgA mice would respond differently than the ones used in this study.

587
Furthermore, the load in this study was not adjusted to account for differences in the initial bone  (Figures 5 and 6), we suspect that the bone loss occurring 597 in the loaded WT mice was due to the surgical insertion of the pins required for loading. 598 Interestingly, this decline in trabecular bone was not observed in the PolgA mice, thus 599 explaining the high BV/TV and Tb.Th in the sham-loaded PolgA mice at end-point. We have 600 previously observed similar bone loss in 52-week-old mice with loading being able to reduce 601 this bone loss [27]. Similar to our current study though, the 15-and 82-week-old mice did not 602 show any bone loss. It is possible that the prematurely aged PolgA mice behave similarly to the 603 82-week-old mice of that study. Nevertheless, the fact that PolgA bones lose mechano-responsiveness term muscle health, this might not be the case for bone health. Future studies are required to better 615 understand discrepancies and potential interactions in mTORC1 signaling between muscle and bone, 616 and whether such differences can be linked to altered mechanosensitivity with age. 617 There are a number of limitations to our study that should be mentioned. Firstly, the acute 618 response to anabolic stimuli in the muscle tissue cannot be directly compared to the long-term 619 loading regime (over 4 weeks) to which the bone tissue was subjected. Physiological resistance 620 training protocols leading to skeletal muscle hypertrophy in mice are difficult to achieve. The 621 protocols, which are available such as voluntary resistance running [85] and high intensity 622 interval running [86]  future studies should investigate the potential of longer-term resistance exercise training in 626 alleviation of mTORC1 hyperactivity in (mouse models of) progeria. To be noted though is that 627 the above mentioned exercise regimes were applied already at young (10-12 weeks) up to old 628 age thereby making it unknown whether they would have the same effect in older PolgA mice 629 already displaying signs of aging. In this respect, future studies subjecting PolgA caudal 630 vertebrae to cyclic loading regimes throuoghout life (i.e, from young until old age) would be 631 interesting to assess whether the reduction in bone mechanosensitivity with age could be 632 prevented.

633
A further limitation of our study was that the evaluation of the bone remodeling activities was 634 mainly based on time-lapsed micro-CT analysis, whereas two-dimensional (2D) histomorphometry 635 using fluorescent labels would allow for the assessment of bone formation and mineralization 636 at an even higher resolution [88]. However, owing to the lack of an appropriate marker, bone correlations between dynamic histomorphometric parameters quantified by micro-CT and conventional 642 histomorphometry, respectively [25, 39, 89]. A further advantage of micro-CT, however, is that bone 643 histomorphometric parameters can be assessed in the entire 3D compartment rather than in a limited 644 number of 2D histological sections. This not only reduces labor-intensive processing of the samples but 645 also reduces inter-and intra-observer errors associated with the analysis thereof [90,91]. The final 646 determining advantage for using in vivo micro-CT in this particular study though was the long-term 647 period of 20 weeks over which bone remodeling activities were assessed. Due to the resorption of 648 fluorochrome labels, 2D histomorphometry requires very short intervals (2-3 days) between marker 649 injections [92], and hence, is not well suited for long-term studies. 650 In conclusion, we show that PolgA mice develop multiple hallmarks of aging, such as reduced 651 bone remodeling and muscle mass early in life, which collectively can be quantified using the 652 mouse FI. We further demonstrate acute mTORC1 hyperactivity in PolgA muscle upon 653 anabolic signals, which is related to diminished satellite cell proliferation. By mimicking many 654 aspects of osteosarcopenia, the PolgA mouse provides a powerful model that facilitates our 655 understanding of the relationship between muscles and bones, and also between the aging 656 musculoskeletal system and frailty.