Phytochemical compound PB125 attenuates skeletal muscle mitochondrial dysfunction and impaired proteostasis in a model of musculoskeletal decline

Impaired mitochondrial function and disrupted proteostasis contribute to musculoskeletal dysfunction. However, few interventions simultaneously target these two drivers to prevent musculoskeletal decline. Nuclear factor erythroid 2‐related factor 2 (Nrf2) activates a transcriptional programme promoting cytoprotection, metabolism, and proteostasis. We hypothesized daily treatment with a purported Nrf2 activator, PB125, in Hartley guinea pigs, a model of musculoskeletal decline, would attenuate the progression of skeletal muscle mitochondrial dysfunction and impaired proteostasis and preserve musculoskeletal function. We treated 2‐ and 5‐month‐old male and female Hartley guinea pigs for 3 and 10 months, respectively, with the phytochemical compound PB125. Longitudinal assessments of voluntary mobility were measured using Any‐MazeTM open‐field enclosure monitoring. Cumulative skeletal muscle protein synthesis rates were measured using deuterium oxide over the final 30 days of treatment. Mitochondrial oxygen consumption in soleus muscles was measured using high resolution respirometry. In both sexes, PB125 (1) increased electron transfer system capacity; (2) attenuated the disease/age‐related decline in coupled and uncoupled mitochondrial respiration; and (3) attenuated declines in protein synthesis in the myofibrillar, mitochondrial and cytosolic subfractions of the soleus. These effects were not associated with statistically significant prolonged maintenance of voluntary mobility in guinea pigs. Collectively, treatment with PB125 contributed to maintenance of skeletal muscle mitochondrial respiration and proteostasis in a pre‐clinical model of musculoskeletal decline. Further investigation is necessary to determine if these documented effects of PB125 are also accompanied by slowed progression of other aspects of musculoskeletal dysfunction.


Introduction
The musculoskeletal system comprises bones, joints, cartilage, tendon and skeletal muscle, all of which are physically and biochemically connected (Bonewald et al., 2013;DiGirolamo et al., 2013). Age-related decline in musculoskeletal function contributes to the health burden associated with ageing (Goates et al., 2019). Moreover, obesity also causes musculoskeletal decline during periods of growth in overweight/obese children and adolescents (Krul et al., 2009;Merder-Coskun et al., 2017) and further drives musculoskeletal dysfunction across the lifespan (Wang et al., 2016). Musculoskeletal dysfunction imparts a loss of mobility and independence (Roux et al., 2005) and leads to frailty (Walston et al., 2006). It also exacerbates comorbidities including cardiometabolic disease (Baskin et al., 2015), cancer (Williams et al., 2018) and cognitive decline (Ogawa et al., 2018), and increases mortality (García-Hermoso et al., 2018). There are no established therapeutics to slow musculoskeletal decline (Yoshimura et al., 2017), including musculoskeletal deficiencies associated with osteoarthritis (OA), which affects over 30 million US adults (US Bone & Joint Initiative, 2020). Accordingly, the National Institutes of Health (NIH) identified a critical need  (https://www. grants.nih.gov/grants/guide/pa-files/par-15-190.html)) to 'accelerate the pace of development of novel therapeutics… for preventing and treating key health issues affecting the elderly' .
The lack of effective therapeutics for maintaining musculoskeletal function is partially attributable to the insidious nature of musculoskeletal decline in humans, as well as the absence of animal models that recapitulate the multifactorial processes that drive musculoskeletal decline. The Hartley guinea pig is an outbred guinea pig that develops primary (also considered spontaneous or idiopathic) OA that closely resembles disease onset and progression in humans starting at 4 months of age (Jimenez et al., 1997). By 9 months of age, these guinea pigs have diminished mobility. At 18 months of age, the severity of OA renders the guinea pigs up to 50% less mobile (Santangelo et al., 2014). Similar to humans with OA (Kemmler et al., 2015;Noehren et al., 2018), skeletal muscle fibre size and density decrease and the proportion of type I fibres increases in the gastrocnemius by 15 months in these guinea pigs (Musci et al., 2020;Tonge et al., 2013). Low skeletal muscle mass in the lower limb is associated with knee osteoarthritis (Lee et al., 2016). Accordingly, the Hartley guinea pig represents a potential model to study musculoskeletal deficiencies associated with OA in a compressed amount of time (i.e. 5-15 months of age). Indeed, another lab tested the effect of treatment with metformin and rapamycin, drugs that have beneficial effects on lifespan in mice (Strong et al., 2016), on progression of musculoskeletal dysfunction in Hartley guinea pigs. Surprisingly, the drugs had deleterious effects on OA severity  and impaired skeletal muscle mitochondrial function . Thus, there remains a strong need to investigate other therapeutic interventions that can slow the progression of OA and OA-related musculoskeletal dysfunction. Oxidative stress, mitochondrial dysfunction, and the loss of proteostasis, which are all hallmarks of ageing and contribute to other chronic diseases, likely contribute to OA and loss of skeletal muscle function. Cartilage and skeletal muscle are slow to proliferate and have low rates of protein turnover (Hall, 2012;Relaix et al., 2021;Vaananen, 1993) relative to tissues such as liver . Thus, cartilage and skeletal muscle are susceptible to accumulating oxidative damage and impaired proteome integrity. Targeting mitochondrial dysfunction likely ameliorates not just impaired ATP production and oxidative stress but also other factors that underlie musculoskeletal decline such as impaired proteostasis (protein homeostasis) . Impaired mitochondrial function is associated with, and precedes, impairments in proteostasis and decrements in skeletal muscle function (Gaffney et al., 2018;Gonzalez-Freire et al., 2018). Inversely, improvements in proteostatic mechanisms regulating mitochondrial proteome integrity would improve mitochondrial function , which may in turn alleviate the energetic constraints that impair adequate cellular function. Thus, targeting oxidative stress, mitochondrial function, and/or impaired proteostasis could similarly slow the progression of OA and skeletal muscle decline.
Nuclear factor erythroid 2-related factor 2 (Nrf2) is a transcription factor that regulates hundreds of genes involved in adaptation to stress, including those involved in redox homeostasis, mitochondrial energetics and proteome maintenance (Gao et al., 2020). Nrf2 activation leads to the upregulation of cytoprotective genes by binding to the antioxidant response element in the promoter regions of target genes, including antioxidant genes such as SOD1, NQO1 and HO1 (Katsanos et al., 2006). Nrf2 activation also has anti-inflammatory effects (Ahmed et al., 2017) and has a role in regulating mitochondrial biogenesis (Piantadosi et al., 2008). Transient Nrf2 activation through phytochemical supplementation (Donovan et al., 2012;Hybertson et al., 2019;Kubo et al., 2017;Reuland et al., 2013) is a potential therapeutic intervention that could mitigate chronic diseases (Houghton et al., 2016) and contribute to extended lifespan. Transiently activating Nrf2 targets several interconnected drivers of ageing including macromolecular damage, disrupted redox homeostasis (Fang et al., 2017;Reuland et al., 2013), inflammation (Kobayashi et al., 2016) and impaired proteostasis (Konopka et al., 2017). Indeed, treatment with the phytochemical Nrf2 activator Protandim extended median lifespan of male mice in the NIH-National Institute on Aging (NIA) Interventions Testing Program (ITP) (Strong et al., 2016).
Given the positive effects of Nrf2 activator treatment, we sought to identify the effects of months-long Nrf2 activator treatment on skeletal muscle mitochondrial respiration and protein turnover in the Hartley guinea pig. Nrf2 activation occurs by a complex set of mechanisms, and thus different Nrf2 activators have similar but not identical effects on Nrf2 activity, gene regulation and physiological endpoints (Baird & Yamamoto, 2020;McCord et al., 2021;Yu & Xiao, 2021). For this study, we utilized the dietary phytochemical combination PB125 as the Nrf2 activator due to the low electrophilic toxicity of its components and potent Nrf2 activating properties (Hybertson et al., 2011;Hybertson et al., 2019;McCord et al., 2021). We hypothesized treatment with Nrf2 activator PB125 would improve skeletal muscle mitochondrial respiration and mechanisms of proteostasis, thereby attenuating musculoskeletal declines in both male and female guinea pigs.

Husbandry
All procedures were approved by the Colorado State University Institutional Animal Care and Use Committee and were performed in accordance with the NIH Guide for the Care and Use of Laboratory Animals. Dunkin-Hartley guinea pigs were obtained from Charles River Laboratories (Wilmington, MA, USA) at 1 and 4 months of age (mo) for each treatment regimen; 14 male and female guinea pigs were utilized in each age and treatment group (total n = 112) (Fig. 1A). As mentioned, Hartley guinea pigs begin developing knee OA at 3-5 mo and have severe OA and skeletal muscle and joint phenotypes consistent with aged human musculoskeletal systems by 15 mo (Jimenez et al., 1997;Musci et al., 2020;Santangelo et al., 2014;Tonge et al., 2013). Accordingly, we chose these ages to determine if PB125 could prevent the onset (short term treatment from 2 to 5 mo) or mitigate the progression (long term treatment from 5 to 15 mo) of musculoskeletal dysfunction (Jimenez et al., 1997;Santangelo et al., 2014) and skeletal muscle decline (Musci et al., 2020) (Fig. 1A). It is important to note that, because knee OA progresses as these animals age from 2 to 15 mo, we cannot discern the effect of age itself from disease progression. Additionally, these animals are still experiencing long bone and soft tissue growth until ∼10 mo (Watson et al., 1996). Thus, any changes between 5 and 15 mo guinea pigs could be a result of disease progression or a maturation/ageing process. We acknowledge this is a limitation.
Animals were maintained at Colorado State University's Laboratory Animal Resources housing facilities and were monitored daily by veterinary staff. All guinea pigs were singly housed in solid bottom cages, maintained on a 12-12 h light-dark cycle, and provided with access to food and water ad libitum. Two control females, two PB125 females, one control male and two PB125 males required humane euthanasia prior to final analysis due to underlying issues unrelated to treatment (final n = 105). Gross necropsy findings by veterinarians did not raise significant concern as the causes of death in these cases were consistent with what would be expected in conventionally raised guinea pigs.
Measurement of PB125 in OraSweet and in guinea pig plasma using high performance liquid chromatography-mass spectrometry PB125 (Pathways Bioscience, Aurora, CO, USA) is a phytochemical compound comprising rosemary, ashwagandha and luteolin powders, which contain the three active ingredients carnosol (CRN), withaferin A (WFA) and luteolin (LUT) at a mixed ratio of 15:5:2 by mass, respectively (Hybertson et al., 2019). Prior to treatment initiation, to ensure adequate compound delivery, plasma levels of each of three primary active ingredients were measured 15, 30, 45, 60, 90 and 120 min post-dosing at concentrations of 8, 24 and 48 mg/ml ( Fig. 1B-D), which corresponds with a dosage of 250, 750 and 1250 ppm. The concentration of each drug was reported as μg/ml. Because we prepared weekly allotments of PB125 in OraSweet, we verified the stability of PB125 suspended in OraSweet stored at both room temperature and 4°C for 1 week (Fig. 1E). Compound was kept at 4°C. Reference standards of LUT and WFA were purchased from Sigma-Aldrich (St Louis, MO, USA). CRN was purchased from Cayman Chemical Co. (Ann Arbor, MI, USA). All other reagents were purchased from Thermo Fisher Scientific (Waltham, MA, USA). HPLC grade methanol was used for preparation of all solutions. Samples were analysed at the Nathan Shock Core Analytical Pharmacology Core at the University of Texas Health Medical School.
The liquid chromatography tandem mass spectrometry (LC/MS/MS) system consisted of a Shimadzu SIL 20A HT autosampler, LC-20AD pumps (2), and an AB Sciex API 4000 tandem mass spectrometer with turbo ion spray. The LC analytical column was an ACE C8 (50 × 3.0 mm, 3 μm) purchased from Mac-Mod Analytical (Chaddsford, PA, USA). Mobile phase A contained 0.1% formic acid dissolved in water. Mobile phase B contained 0.1% formic Figure 1. Study design, PB125 plasma concentration and anthropometric data A, there were two cohorts of guinea pigs in this study. The first cohort was treated with PB125 or vehicle control from 2 to 5 mo, an age range during which knee OA begins developing. The second cohort was treated from 5 to 15 mo, an age range after the onset of knee OA and during which detectable declines in musculoskeletal quality arise. In the final 30 days of each study, a bolus I.P. injection of 2 H 2 O was administered and 2 H 2 O was mixed in drinking water for measurement of protein synthesis. A portion of the soleus was harvested for mitochondrial respirometry assessments. Another portion of the soleus as well as a portion of the gastrocnemius was harvested for isotopic measurements. Comparisons between 5 and 15 mo guinea pigs were made between the cohorts at the day of harvest. Longitudinal weight and mobility data were acquired from the second cohort. B-D, the plasma concentrations of luteolin, carnosol and withaferin A after dosing guinea pigs with 8 (B), 24 (C), or 40 (D) mg/kg of PB125 from 0 to 120 min. The plasma concentrations of luteolin and withaferin A are on the primary y-axis and carnosol is on the secondary y-axis. E, the stability of the phytochemical components suspended in OraSweet vehicle at room temperature and in 4°C. F and G, growth charts for male (F) and female (G) guinea pigs (n = 28 for each sex). H, there was a significant effect of disease/age (P < 0.0001) and sex (P < 0.0001) on guinea pig body mass (n = 88). I, there was no effect of disease/age in relative soleus mass (mg of muscle/g of body mass) (n = 86). J, however, there was a significant (P = 0.005) negative effect of disease/age on the gastrocnemius (n = 86). [Colour figure can be viewed at wileyonlinelibrary.com] J Physiol 601.11 acid dissolved in 100% HPLC grade acetonitrile. The LC gradient was: 0-2 min, 25% B; 2-6 min, linear gradient to 99% B; 6-10 min, 99% B; 10-10.01 min, 99% to 25% B min; 10.1-12 min, 25% B. LUT and CRN were detected in negative mode using these transitions: 285 to 132.9 m/z and 329 to 285 m/z, respectively. WFA was detected in positive mode at the transition of 471 to 281 m/z.
LUT, CRN and WFA stock solutions were prepared in methanol at a concentration of 1 mg/ml and stored in aliquots at −80°C. Working stock solutions of each drug were prepared each day from the super stock solutions at a concentration of 100, 10 and 1 μg/ml which were used to spike the calibrators.
Dosages of PB125 in OraSweet were diluted 1000× in 70% ethanol. Calibrator samples were prepared daily by spiking blank OraSweet to achieve final concentrations of 0, 30.4, 152, 760 and 2280 μg/ml. The calibrators were then diluted 1000× in 70% ethanol. The samples were transferred to injection vials and 10 μl was injected into the system. Each drug was quantified by comparing the peak area ratios for each dosage sample against a linear regression of calibrator peak area ratios.
LUT, CRN and WFA were also quantified in guinea pig plasma. The transitions used were the same as the OraSweet dilutions. Calibrator samples were prepared daily by spiking blank plasma to achieve final concentrations of 0, 5, 10, 25, 50, 100, 500, 1000, and 5000 ng/ml. Calibrators were left to sit for 5 min after spiking. Briefly, 0.1 ml of calibrator and unknown plasma samples were mixed with 1.0 ml of chilled ethanol, vortexed vigorously, and then centrifuged at 17,000 g for 5 min at 25°C. The supernatants were transferred to 1.5 ml microcentrifuge tubes and dried to residue under a nitrogen stream. The residues were then redissolved in 60 μl of 50/50 mobile phase A/mobile phase B and were centrifuged 5 min at 17,000 g. The samples were transferred to injection vials and 15 μl was injected into the LC/MS/MS. Each drug was quantified by comparing the peak area ratios for each unknown sample against a linear regression of calibrator peak area ratios. The concentration of LUT, CRN and WFA were expressed as ng/ml plasma ( Fig. 1B-D).

Treatment, euthanasia and tissue acquisition
Based on the analysis conducted at the NSC Analytical Pharmacology Core (Fig. 1B-D), we selected a dosage of 8 mg/kg of body weight, which corresponds to 250 ppm, about 2.5× the dose of PB125 given to mice in the NIA ITP (https://www.nia.nih.gov/research/dab/interventionstesting-program-itp/compounds-testing) and the lowest dose at which changes in plasma concentration were detected. After a 1-month acclimation to housing conditions, male and female guinea pigs in each age group (2 or 5 months) were randomized to receive a daily oral dose of 8.0 mg/kg body weight of PB125 suspended in OraSweet (Perrigo, Dublin, Ireland) or an equivalent volume of OraSweet only (CON). Following established protocol, guinea pigs were given a subcutaneous injection of 0.9% saline enriched with 99% deuterium ( 2 H 2 O) equivalent to 3% of their body weight 30 days prior to euthanasia (Musci et al., 2020). Drinking water was enriched to 8% 2 H 2 O for the purpose of maintaining 2 H 2 O enrichment of the body water pool during the 30-day labelling period. At the time of harvest, the guinea pigs were 5 mo (after 3 months of treatment) or 15 mo (after 10 months of treatment). In accordance with the standards of the American Veterinary Medical Association, animals were anaesthetized with a mixture of isoflurane and oxygen; thoracic cavities were opened and blood was collected via direct cardiac puncture. Whole blood was centrifuged (1200 g, 4°C, 15 min) to separate plasma, which was frozen at −80°C until further analysis. After blood collection, the anesthetized animals were transferred a chamber filled with carbon dioxide for euthanasia.
Upon euthanasia, the right soleus muscle of the guinea pig was promptly excised and a portion (∼40 mg) was harvested and placed in BIOPS preservation buffer (2.77 mM CaK 2 -EGTA, 7.23 mM K 2 -EGTA, 20 mM imidazole, 20 mM taurine, 50 mM K-MES, 0.5 mM dithiothreitol, 6.56 mM MgCl 2 , 5.77 mM ATP, and 15 mM phosphocreatine, adjusted to pH 7.1) containing 12.5 μM blebbistatin to inhibit muscle contraction (Pesta & Gnaiger, 2011). The rest (∼70 mg) of the soleus was frozen in liquid nitrogen and used for other analyses. After excision of the soleus, at least 70 mg of the right gastrocnemius was collected and frozen immediately in liquid nitrogen. Both soleus and gastrocnemius muscles were trimmed of tendons and connective tissue and weighed. Bone marrow was also harvested in saline from the humeri.

Mitochondrial respirometry
After the soleus was placed in BIOPS, the muscle fibres were prepared for high resolution respirometry as follows. Mechanical permeabilization occurred on ice using forceps to separate the fibres. After mechanical permeabilization, fibres underwent chemical permeabilization for 30 min in BIOPS with 12.5 μM blebbistatin and 50 μg/ml saponin, followed by a 15 min rinse in BIOPS. Approximately 2.0 mg (wet weight) of muscle fibres was placed in mitochondrial respiration medium (MiR05, 0.5 mM EGTA, 3 mM MgCl 2 6H 2 O, 20 mM Taurine, 15 mM Na 2 -phosphocreatine, 20 mM Imidazole, 0.5 mM dithiothreitol, and 50 mM K-MES at pH 7.1) in an Oxygraph-2k (O2K) (Oroboros, Innsbruck, Austria) for high resolution respirometry. To control for oxygen flux at higher concentrations of oxygen, each morning of respirometry analysis, we conducted high oxygen concentration calibrations at 450, 350, 250 and 167 (i.e. concentration of room air) nmol/ml O 2 (Pesta & Gnaiger, 2011). During the experiments, oxygen concentrations were maintained between 225 and 450 nmol/ml O 2 . High resolution respirometry measurements were performed in duplicate using two different protocols. A detailed explanation of the protocols is given in Table 1.
The first protocol (Suit 1) was an ADP titration protocol to determine ADP sensitivity (K m ) and maximal oxidative capacity (V max ) under complex I-supported respiration. We measured complex I-supported leak respiration (State 2 (PGM) ) with the addition of 10 mM glutamate, 0.5 mM malate and 5 mM pyruvate. Upon acquisition of State 2 (PGM) , we titrated progressively greater concentrations of ADP from 0. 1,0.175,0.25,1,2,4,8,12,20,to 24 mM (State 3 (PGM) ), awaiting steady-state oxygen flux prior to adding the subsequent titration to determine complex I linked ADP V max and apparent K m (i.e. ADP sensitivity). After the ADP titration was completed, we added 5 mM cytochrome c to test mitochondrial membrane integrity. After cytochrome c addition, we added 10 mM succinate to acquire maximal complex I and II-supported coupled respiration (State 3 (PGM+S) ). We then added 0.5 μM carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP) sequentially until there was no increase in respiration to determine the capacity of the electron transport system to consume oxygen, or maximal uncoupled respiration (ETS (CI-CIV) ). Finally, we added 5 μM rotenone to measure maximal uncoupled respiration with the inhibition of complex I (ETS (CII-CIV) ), followed by 2.5 μM antimycin A to measure residual oxygen consumption (ROX). The respiratory control ratio (RCR: State 3/State 2), which is an index of mitochondrial efficiency, was also evaluated.
The second protocol (Suit 2) measured oxygen consumption while simultaneously measuring ROS production by using the fluorometer attachment of the O2K (Robinson et al., 2019) and addition of 10 μM Amplex Red, 1 U/ml horseradish peroxidase and 5 U/ml superoxide dismutase. We then measured fatty acid-supported leak respiration by adding 10 mM glutamate, 0.5 mM malate, 5 mM pyruvate and 0.2 mM octanoylcarnitine (State 2 (PGM+Oct) ) and 10 mM succinate (State 2 (PGM+Oct+S) ). After stimulating maximal leak respiration, we added submaximal boluses of ADP (0.5 mM: State 3 (Sub+0.5D) and 1 mM: State 3 (Sub+1.0D) ), followed by a saturating bolus of ADP (6.0 mM: State 3 (Sub+6.0D) ). We added 5 mM cytochrome c to test mitochondrial membrane integrity. We set a cytochrome c control factor threshold of 0.25. We set this threshold based on the presence of a negative linear relationship between the cytochrome c control factor and State 3 respiration in the Suit 2 protocol. Upon eliminating respirometry trials that had a cytochrome c control factor of greater than 0.25, the negative linear relationship no longer existed and all samples included in analysis were not biased by overpermeabilization, which is what the cytochrome c control factor approximates (Pesta & Gnaiger, 2011) ( Fig. 2A-C). After the addition of cytochrome c we did not make any further assessments of ROS emission as cytochrome c is a redox substrate. We added 5 μM rotenone to determine maximal coupled respiration in the absence of complex I (State 3 (Sub+D-CI) ) followed by sequential titrations of 0.5 μM FCCP until respiration no longer increased to determine maximal fatty acid-supported uncoupled respiration (ETS (Sub+D-CI) ) and added 2.5 μM antimycin A to measure ROX.

Sample preparation and analysis via gas chromatography-mass spectrometry: proteins
Protein subfractions were hydrolysed in 6 M HCl for 24 h at 120°C after which the hydrolysates were ion-exchanged, dried in vacuo, and then resuspended in 1 ml of molecular biology grade H 2 O. Half of the suspension was derivatized with 500 μl acetonitrile, 50 μl 1 M K 2 HPO 4 and 20 μl of pentafluorobenzyl bromide and incubated at 100°C for 60 min. Derivatives were extracted into ethyl acetate and the organic layer was transferred into vials which were then dried under nitrogen. Samples were reconstituted in ethyl acetate (200-700 μl).
The derivative of alanine was analysed on an Agilent 7890A gas chromatograph coupled to an Agilent 5977A MS (Agilent Technologies, Santa Clara, CA, USA) as previously described Groennebaek J Physiol 601.11  Respiration associated with ATP production. NADH is generated through the oxidation of PGM, donating electrons to complex I, and creating a proton gradient where ADP is the limiting factor for ATP production. Oxygen consumption is a consequence of ADP being phosphorylated to ATP. Initial titrations are sub-saturating concentrations of ADP increasing to maximal ADP stimulated respiration State 3 (Cytc) Previous step + 5 mM cytochrome c During permeabilization, the outer membrane of mitochondria can be damaged, resulting in loss of cytochrome c. An increase in oxygen consumption is indicative of cytochrome c release during mechanical or chemical permeabilization State 3 (PGM+S) Previous step + 10 mM succinate (S) Same as previous step but now both NADH and FADH 2 are generated through the oxidation of complexes I (PGM) and II (S) substrates yielding maximal ADP stimulated respiration ETS (CI-CIV) Previous step + 1.0 mM FCCP FCCP is a protonophore allowing hydrogen ions through the inner-mitochondrial membrane, uncoupling the proton gradient from ATP production through ATP synthase. This reflects the maximal or reserve capacity of complexes I-IV, unconstrained from ATP synthase activity ETS (CII-CIV) Previous step + 5 μM rotenone Rotenone is an inhibitor of complex I. Therefore, uncoupled oxygen consumption is indicative of complex II-IV capacity ROX 2.5 μM antimycin A Antimycin A inhibits complex III shutting down the electron transfer pathway. ROX is remaining oxygen consumed from other pathways independent of complex IV Suit 2: a protocol measuring coupled and uncoupled respiration as well as ROS emission simultaneously in the presence of fatty acids State 2 (PGM+Oct) 5 mM pyruvate (P), 10 mM glutamate (G), 0.5 mM malate (M), 0.2 mM Oct LEAK respiration in the presence of complex I and II substrates including both carbohydrates and fatty acids. The amount of oxygen consumption as a consequence of protons leaking across the membrane State 2 (PGM+Oct+S) Previous step + 10 mM succinate (S) Same as previous step with the addition of the complex II substrate, succinate (S). This is considered to stimulate maximal LEAK respiration as well as generation of ROS State 3 (Sub+0.5D) Previous step + 0.5 mM ADP (D) Respiration associated with ATP production. NADH and FADH 2 are generated through the oxidation of PGMS and octanoylcarnitine donating electrons to complexes I and II, and creating a proton gradient where ADP is the limiting factor for ATP production. The amount of oxygen consumed is from ADP being phosphorylated to ATP with a sub-saturating bolus of ADP State 3 (Sub+1.0D) Previous step + 0.5 mM ADP (D) Final concentration of ADP is higher but still sub-saturating stimulated respiration State 3 (Sub+6.0D) Previous step + 5 mM ADP (D) Same as previous step, however, the final concentration of ADP (6.0 mM) is a saturating does, maximally stimulating mitochondrial respiration and ATP production State 3 (Cytc) Previous step + 5 mM cytochrome c During permeabilization, the outer membrane of mitochondria can be damaged, resulting in loss of cytochrome c. An increase in oxygen consumption is indicative of cytochrome c release during mechanical or chemical permeabilization. After the addition of cytochrome c, no more assessments of ROS emission were measured (Continued)  The newly synthesized fraction of proteins was calculated from the true precursor enrichment based upon plasma analysed for 2 H 2 O enrichment and adjusted using mass isotopomer distribution analysis (Busch et al., 2005). Protein synthesis of each subfraction was calculated as the fraction of deuterium-labelled over unlabelled alanine proteins over the entire labelling period (30 days) and expressed as the fractional synthesis rate (FSR). Thus, we divided fraction new by our labelling period (30 days) and multiplied by 100 to express FSR as %/day. Our isotope approach and analysis followed the established procedures detailed in this Core of Reproducibility in Physiology publication (Miller et al., 2020).

Sample preparation and analysis via gas chromatography/mass spectroscopy: body water
Eighty microlitres of plasma was placed into the inner well of an o-ring cap that was screwed to tube and inverted on a heating block overnight at 100°C. After incubation, 2 μl of 10 M NaOH and 20 μl of acetone were added to the samples and 2 H 2 O standards (0% to 20%) and capped immediately, vortexed and incubated at room temperature overnight. Samples were extracted with 200 μl hexane and the organic layer was transferred through pipette tips filled with anhydrous Na 2 SO 4 into gas chromatography vials and analysed via EI mode using a DB-17MS column.

Protein content
Western blotting was used to measure relative content of Nrf2, haem oxygenase (HO-1) and OXPHOS proteins in a subset of tissues. Fifty-to seventy-milligram portions of gastrocnemius and 30 mg portions of soleus (n = 9 per treatment group) were powdered under liquid nitrogen and homogenized in a Bullet Blender with zirconium beads and 1.0 ml of radioimmunoprecipitation assay (RIPA) buffer (150 mM NaCl, 0.1 mM EDTA, 50 mM Tris, 0.1% sodium deoxycholate, 0.1% SDS, 1% Triton X-100, pH 7.50) with HALT protease inhibitors. Samples were reduced (50 μl of β-mercaptoethanol) and heated at 50°C for 10 min. Approximately 10 μg of protein was loaded into a 4-20% Criterion pre-cast gel (Bio-Rad Laboratories, Hercules, CA, USA) and resolved at 120 V for 120 min. milk on a shaker overnight in 4°C. Membranes were rinsed and then incubated with appropriate secondary antibodies (Santa Cruz 2004 (Nrf2 and HO-1) and 2005 (OXPHOS)) diluted to 1:10,000 in 5% BSA for 45 min at room temperature. Protein carbonyls were measured by following the protocol in the commercially available OxiSelect Protein Carbonyl Immunoblot Kit (Cell Biolabs, San Diego, CA, USA, STA-308) as previously performed (Konopka et al., 2015;Konopka et al., 2017). After the membranes were rinsed, SuperSignal West Dura Extended Duration Substrate (Thermo Fisher Scientific, 34075) was applied and the membranes were subsequently imaged using a FluorChem E Chemiluminescence Imager (Protein Simple, San Diego, CA, USA). Analysis of densitometry was completed using AlphaView SA software (Protein Simple). Units are expressed as density of primary antibody relative to density of ponceau staining.

Mobility
Before the onset of the study, animals were acclimated over a 2-week period to an open circular field behaviour monitoring system (ANY-Maze TM , Wood Dale, IL, USA) to assess voluntary physical mobility. Animals' activities were recorded, and data were collected for 10 consecutive minutes on a monthly basis throughout the study. Mobility assessments were conducted in the mornings to minimize variance due to time of day. Additionally, cohorts containing both CON and PB125-treated guinea pigs were evaluated on the same day and time. To evaluate the effect of PB125 treatment on the time to event noncompliance, we employed a Kaplan-Meier curve, frequently used to visually summarize time-to-event data (Ranstam & Cook, 2017). The 'event' was defined as number of weeks into the study until an animal selected to remain stationary when exposed to the open circular field (i.e. zero distanced travelled in enclosure). Remaining individuals that maintained voluntary mobility throughout the entire 40-week study duration were censored at the 40-week study endpoint.

Statistics
For mitochondrial respirometry, in line with best practices, technical replicates were averaged. The average variability between these technical replicates in this study was 18%, which is standard according to the literature (Jacques et al., 2020). Apparent K m and V max values were determined using Michaelis-Menten kinetics in Prism 9.0 (GraphPad Software Inc., San Diego, CA, USA). For evaluating growth rates, a non-linear Gompertz growth line was fit to the change in body mass over time (i.e. the rate of growth). The rate of growth, k, was compared between treatment and control within in each sex. For respirometry, isotopic measures and Western blots, three-way ANOVA was used to measure the main effects of sex, musculoskeletal disease progression/age (referred to as disease/age-related effects from here forward), and PB125 treatment. Post hoc analyses were performed using Bonferroni's post hoc test.
To determine the effect of PB125 on disease/age-related changes in mitochondrial respiration and protein synthesis when a significant effect of disease/age was detected, we conducted a subset analysis using a one-way ANOVA with Dunnett's post-hoc test comparing 15 mo treated and untreated guinea pigs to 5 mo untreated guinea pigs.
Because this study was a secondary project within a larger study with a different primary outcome, we did not design this study to be powered to detect differences in mitochondrial respiration and protein synthesis at a P-value <0.05. Moreover, prior to this study, there were no published data on mitochondrial respiration in permeabilized skeletal muscle fibres of guinea pigs. While we set statistical significance a priori at P < 0.05, we also report differences with P < 0.10 and corresponding effect sizes to highlight potential directions for future studies.
For measuring sustained mobility, we employed Kaplan-Meier curves combined with log rank and Gehan-Breslow-Wilcoxon tests to assess the probability of sustained voluntary mobility throughout the 10-month study period of the 'long term' study.
Data are presented as means ± SD. All statistics were performed in Prism 9.0. Cohen's d was calculated by calculating the difference between group means relative to the variance of either group (Sullivan & Feinn, 2012).

Results
Growth rate (k-curves) of the group treated with PB125 did not significantly differ from the control group (CON) as measured by changes of body weight throughout treatment (P > 0.70 for both sexes) and at euthanasia (P > 0.70) (Fig. 1F-H). Moreover, there were no differences in relative skeletal muscle mass between PB125 and CON ( Fig. 1I and J).

Disease/age-related declines in mitochondrial respiration occur in both male and female guinea pigs
Because mitochondrial respiratory capacity has never been measured in permeabilized skeletal muscle fibres from Hartley guinea pigs, we first evaluated disease/age and sex differences in mitochondrial respiration. Table 1 provides an explanation for each respiratory state mentioned below. Figure 2D-K provide summary titration data. Eleven of 210 trials were excluded due to overpermeabilization (cytochrome c control factor > 0.25; Fig. 2A  Disease/age had a negative effect on several aspects of mitochondrial respiration in both male and female guinea pigs. Fifteen-month-old male and female guinea pigs had lower coupled (State 3 (PGM+S) ) ( Fig. 3A) and uncoupled (ETS (CI-CIV) ) ( Fig. 3B) respiration compared to 5 mo counterparts (Cohen's d = 0.731, 0.680; P = 0.001, P = 0.004, respectively). There was also a disease/age-related decline (Cohen's d = 1.388, P < 0.0001) in uncoupled respiration without complex I support (ETS (CII-CIV) ) (Fig. 3C). Disease/age had no effect on fatty acid oxidation-supported respiration (Fig. 3E-G). Mitochondrial efficiency (RCR) also decreased as a result of disease/age (Cohen's d = 0.533, P = 0.012) (Fig. 3D).

PB125 has sex specific effects on mitochondrial ADP kinetics
No O2K data from the ADP titration protocol were excluded based on cytochrome c control factors as all values were below 0.25 ( Fig. 2A). We determined ADP kinetics by titrating progressively higher concentrations of ADP with saturating amounts of pyruvate, glutamate and malate. ADP V max was greater in both 15 mo male and female guinea pigs compared to 5 mo counterparts ( Fig. 3J; Cohen's d = 0.445, P = 0.049). In females, ADP V max was lower compared to males ( Fig. 3J; Cohen's d = 0.720, P = 0.001). Guinea pigs that received treatment with PB125 had a greater complex I-supported ADP V max ( Fig. 3J; Cohen's d = 0.498, P = 0.026). Post hoc comparisons indicate that PB125 improved ADP V max in 5 mo female guinea pigs (P = 0.045).
Despite ADP V max being greater in 15 mo guinea pigs, there was no effect of disease/age on the apparent K m of ADP (Fig. 3K). There were also no differences in K m between sexes. However, PB125 did significantly increase the apparent K m (Cohen's d = 0.669, P = 0.007) indicating lower ADP sensitivity, as there was an increase in ADP V max and an absence of change in respiration rates in sub-saturating amounts of ADP ( Fig. 2D-J). There was a non-significant interaction between sex and PB125 treatment (P = 0.092), indicating that the PB125-mediated increase in K m (i.e. decrease in sensitivity) may have occurred only in males.

PB125 attenuates disease/age-related declines in mitochondrial respiration
For any main effects of disease/age on mitochondrial respiration, we evaluated whether PB125 attenuated the disease/age-related changes. That is, where we identified significant differences between 5 mo CON and 15 mo CON guinea pigs, but no differences between 5 mo CON and 15 mo PB125 animals, we reported those findings as an attenuating effect of PB125 treatment on age-related changes in mitochondrial respiration. While there was a main positive effect of disease/age on ADP V max in the three-way ANOVA (Fig. 3J), there was no significant difference in ADP V max between 5 and 15 mo CON guinea pigs in the subsequent one-way ANOVA ( Fig. 4A; Cohen's d = 0.547, P = 0.109). Treated 15 mo guinea pigs, however, had a significantly higher ADP V max compared to 5 mo animals ( Fig. 4A; Cohen's d = 0.869, P = 0.007). Interestingly, this effect was only observed in males (P = 0.021) (Fig. 4B). While ADP V max was greater in 15 mo guinea pigs, 15 mo guinea pigs had a significantly (Cohen's d = 0.709, P = 0.024) lower maximal coupled respiration (State 3 (CI-CIV) ) compared to 5 mo counterparts (Fig. 4C) though this effect was observed in females only (Fig. 4D). PB125, however, prevented that disease/age-related decline (Cohen's d = 0.270 and 0.440, compared to 5 mo CON and 15 mo CON, respectively, Fig. 4C). Maximal uncoupled respiration (ETS (CI-CIV) ) ( Fig. 4E) was also lower in 15 mo CON compared to 5 mo CON (Cohen's d = 0.600), but PB125 prevented the decline (Cohen's d = 0.120 and 0.480, compared to 5 mo and 15 mo CON, respectively). Further interrogation revealed that 15 mo females had lower ETS (CI-CIV) compared to their 5 mo counterparts, which PB125 attenuated (Fig. 4F). Interestingly, when complex I was inhibited, PB125 had no effect on uncoupled respiration (ETS (CII-CIV) ) and had no effect on the disease/age-related decline in CII-CIV capacity (Cohen's d = 0.202) in 15 mo males or females ( Fig. 4G and H). While the RCR of 15 mo CON guinea pigs was lower compared to 5 mo guinea pigs, RCR was not different between 15 mo PB125-treated guinea pigs and 5 mo CON (Fig. 4I). However, this occurred only in males where there was a significant difference (P = 0.036) between 5 and 15 mo CON animals ( Fig. 4J) but no difference (P = 0.151) between 15 mo males treated with PB125 compared to 5 mo CON (Fig. 4J). There was no difference in RCR between 5 and 15 mo females (Fig. 4J). PB125 had no effect on mitochondrial content as assessed by OXPHOS western blot (data not shown), suggesting that the improvements in mitochondrial respiration are independent of mitochondrial content in skeletal muscle and may reflect improved mitochondrial quality.

Sex-, treatment-and disease/age-related effects on skeletal muscle protein synthesis
To determine whether or not PB125-mediated improvement in mitochondrial respiration was linked to improvements in components of proteostasis, we used 2 H 2 O to measure cumulative protein synthesis rates over 30 days. There were no differences in fractional synthesis rate (FSR) in either the gastrocnemius or the soleus between male and female guinea pigs (Fig. 5). There was a disease/age-related decline in the rates of protein synthesis in all subfractions in the soleus and gastrocnemius of both male and female guinea pigs (P < 0.010 for all subfractions) (Fig. 5). PB125 did not have a main effect on FSR in any of the subfractions of either muscle from 5 mo or 15 mo, male or female guinea pigs (Fig. 5). However, there was a non-significant interaction between age and treatment (P = 0.086) in the myofibrillar subfraction of the soleus of both male and female guinea pigs, J Physiol 601.11

PB125 mitigates disease/age-related declines protein synthesis
Because there was a disease/age-related decline in protein synthesis rates in all subfractions of both the soleus and gastrocnemius, we sought to determine if PB125 prevented any of those declines. PB125 attenuated the disease/age-related decline in myofibrillar FSR of the soleus in both males and females ( Fig. 6A and  B). Additionally, PB125 attenuated the decline in mitochondrial FSR in the soleus (Fig. 6C), but these significant differences were no longer detectable when evaluated in males and females separately (Fig. 6D). In the soleus, PB125 also mitigated the decline in cytosolic FSR in males only (Fig. 6F), but had no effect on the decline in collagen FSR in either sex ( Fig. 6G and H). In contrast, PB125 had no attenuating effect on the disease/age-related decline in protein synthesis in any subfraction of the gastrocnemius (Fig. 6I-P).

The effect of PB125 on downstream Nrf2 targets
To determine how long-term PB125 administration affected the Nrf2 pathway, we assessed protein content of Nrf2 and haem oxygenase-1 (HO-1), both of which are downstream targets of Nrf2 activation (Fig. 7A and  B). There was a disease/age-related increase in Nrf2 protein content (Cohen's d = 1.462, P = 0.025), but no age-related difference in HO-1 expression. PB125 had a significant effect on Nrf2 protein content (P = 0.027), with there being a significant interaction between treatment and disease/age. Five-month-old guinea pigs treated with PB125 had greater Nrf2 content compared to CON (Cohen's d = 0.832), whereas PB125-treated 15 mo guinea pigs had lower Nrf2 content compared to CON (Cohen's d = 1.285). Another target downstream of Nrf2 activation, HO-1, demonstrated a similar disease/age and treatment interaction, particularly in males where 15 mo male guinea pigs had greater HO-1 content compared to 5 mo counterparts (P = 0.010). However, PB125-treated 15 mo male guinea pigs had significantly lower protein levels of HO-1 compared to CON counterparts. There was and collagen (n = 76) (G) subfractions, respectively). PB125 attenuated the decline in the myofibrillar (A) and mitochondrial (C) subfractions, but not in the cytosolic (E) or collagen subfractions (G). B, D, F and H, PB125 attenuated (Cohen's d = 0.533) the decline in myofibrillar FSR in both males (P = 0.920) and females (P = 0.166) (n = 76) (B) and attenuated the decline in cytosolic FSR in males only (P = 0.207) (n = 74) (F). There was no disease/age-related effect on mitochondrial FSR in male or females separately (D), and PB125 failed to attenuate the disease/age-related declines in collagen FSR (H). I, K, M and O, in the gastrocnemius, 15 mo CON guinea pigs had significantly lower FSR compared to 5 mo CON guinea pigs in each subfraction (P < 0.0001 for all subfractions) (n = 76, in each subfraction). The FSR of all subfractions of the gastrocnemius in 15 mo PB125 guinea pigs was also significantly lower compared to 5 mo CON guinea pigs (P < 0.0001 for all subfractions). J, L, N and P, this pattern was observed in both male and female guinea pigs (n = 76, in each subfraction). [Colour figure can be viewed at wileyonlinelibrary.com] no differences in protein carbonyl content, a marker of protein damage, in the soleus or gastrocnemius (Fig. 7C  and D).

The effect of PB125 on mobility
To determine whether or not improvements in skeletal muscle mitochondrial respiration and proteostasis translated to improvements in, or maintenance of, mobility, we assessed voluntary activity in an enclosed area using overhead monitoring. We used Kaplan-Meier curves ( Fig. 7E and F) to depict the probability of sustained voluntary mobility throughout the 40-week study period. There was no statistically significant effect of PB125 on maintained mobility in either male or female guinea pigs. However, CON guinea pigs selected to remain stationary more rapidly than PB125 guinea pigs ( Fig. 7E and F; grouped sex hazard ratio = 0.713, 95% CI: 0.3501 to 1.453; median ratio = 1.5, 95% CI: 0.756 to 2.976; P = 0.231). Further, PB125 males tended to have a relative maintenance of mobility compared to controls until about 32 weeks into the study. Interestingly, ∼50% of PB125 males were event non-compliant by 36 weeks, while ∼50% of CON males maintained mobility the entire 40-week study duration (remaining animals were censored at this time) (Fig. 7E). For the majority of the study, PB125 females maintained their mobility compared to CON females. Approximately 50% of control females loss mobility around 16 weeks, while PB125-treated females sustained voluntary mobility until about 28 weeks (Fig. 7F).

Discussion
In this study, we tested the effects of a novel phytochemical Nrf2 activator, PB125, on two factors that drive chronic disease in humans: mitochondrial dysfunction and loss of proteostasis in locomotor muscle. PB125 moderately improved various aspects of mitochondrial respiration. We observed that PB125 ameliorated declines in skeletal muscle mitochondrial respiration and protein synthesis in both male and female Hartley guinea pigs, over an age span during which they mature, develop OA, and have increasing mobility impairment. The relatively moderate improvements and maintenance of mitochondrial respiration and proteostatic mechanisms may also be associated with prolonged maintenance of voluntary activity in females. Collectively, this study demonstrates the potential utility of targeting Nrf2 to ameliorate musculoskeletal decline.

Sex-and age/disease-related differences in mitochondrial respiration
This is the first study to measure skeletal muscle mitochondrial respiration in both male and female Hartley guinea pigs using high resolution respirometry. At the time of data collection, there were no mitochondrial respiration data in Hartley guinea pig skeletal muscle. To date, the Konopka group is the only other lab that has published mitochondrial respirometry data in male-only Hartley guinea pig soleus. Following a similar protocol, they found slightly higher (∼10-20%) respiration and ADP V max values in 8 mo males compared to our 5 mo males . We sought to characterize differences between male and female guinea pigs at 5 and 15 months of age to determine sex differences and changes in skeletal muscle mitochondrial respiration as these guinea pigs mature and develop osteoarthritis. We found a clear sex difference in coupled and uncoupled respiration, with females having lower rates of oxygen consumption than males, accompanied by decreased fatty acid-supported respiration and ADP kinetics. Interestingly, these differences do not seem to be a consequence of mitochondrial content and may instead reflect intrinsic differences in mitochondrial function. One study in both young and old men and women determined that there was no difference in phosphocreatine recovery, a metric of mitochondrial capacity, post-exercise (Kent-Braun & Ng, 2000). However, measuring ATP production using bioluminescence revealed that mitochondria of men have greater capacity to produce ATP than those of women (Karakelides et al., 2010). Employing high resolution respirometry has revealed equivocal results; thus, it remains unclear whether or not females have greater oxidative capacity than males (Cardinale et al., 2018;Miotto et al., 2018). Regardless, it is essential to continue interrogating potential sex differences in mitochondrial J Physiol 601.11 function and changes that occur with the progression of osteoarthritis in both sexes.
Mitochondrial function declines with age and contributes to the ageing process in humans (Distefano et al., 2017;Gonzalez-Freire et al., 2015;Gonzalez-Freire et al., 2018;Short et al., 2005). However, less is known about changes in mitochondrial function during development in either healthy or overweight adolescents, though obesity seems to lead to impaired musculoskeletal function in developing adolescents (Krul et al., 2009) and may accelerate the ageing process (Salvestrini et al., 2019). We demonstrated that both male and female Hartley guinea pigs experience a decline in mitochondrial respiration as they mature from 5 to 15 mo. Given the relatively young age of these guinea pigs (15 mo; ∼10% of recorded maximal lifespan (Gorbunova et al., 2008), and ∼25% of average companion guinea pig lifespan (Quesenberry et al., 2021)), it is difficult to ascertain if these changes are a consequence of age, the underlying factors that drive OA and musculoskeletal dysfunction, or a combination of both. Other laboratory and companion guinea pigs do not exhibit these musculoskeletal phenotypes as early in their lifespans (Musci et al., 2020;Santangelo et al., 2011). Thus, it is remarkable that despite being relatively young, these guinea pigs demonstrate a decline in mitochondrial respiration. However, this Figure 7. PB125 influences Nrf2 content but does not significantly change protein carbonylation or mobility A and B, there was no effect of Sex on Nrf2 (n = 96) (A) or HO-1 (n = 72) (B) protein content in the gastrocnemius. There was significant effect of disease/age and treatment on Nrf2 protein content Cohen's d = 1.462, P = 0.025). There was a significant (P < 0.0001) interaction between disease/age and treatment on both Nrf2 and HO-1 protein content. C, there was no effect of disease/age, sex, or treatment on protein carbonyl concentration in the soleus (n = 46). D, there was no (n = 46) significant increase (P = 0.0814) in protein carbonyl concentration in the gastrocnemius. E and F, there was a greater proportion of PB125-treated male (n = 26) (E) and female (n = 24) (F) guinea pigs that maintained mobility over the course of the study period. However, there was no statistically significant effect of PB125 on the probability of maintaining mobility throughout the course of the study. * Significant difference between 5 mo CON; #significant difference between 15 mo CON. [Colour figure can be viewed at wileyonlinelibrary.com] decline could be influenced by a departure from a growth phase (essentially complete by approximately 9 mo in this strain) and/or a consequence of factors that cause or stem from the progression of OA.
Osteoarthritis is associated with impaired mitochondrial respiration and redox regulation in degenerating joints (Collins et al., 2016;Collins et al., 2018;Farnaghi et al., 2017;Loeser, 2010). Interestingly, lower mitochondrial respiration was associated with more severe OA in 9 mo Hartley guinea pigs Minton et al 2021). In the current study, both coupled and uncoupled respiration, as well as mitochondrial efficiency, declined with age/disease progression in both male and female guinea pig skeletal muscle. There was also a decline in fatty acid-supported oxidation. In contrast, ADP V max unexpectedly increased with age in both male and female guinea pigs. Given the non-uniform changes in mitochondrial complex protein content (data not shown), it is unlikely differences in mitochondrial content explain the age/disease-related declines in respiration. Moreover, we do not have mitochondrial respiration data in strains of guinea pigs that do not develop OA at these ages. However, these data, as well as other literature, support the notion that impaired mitochondrial respiration is a characteristic of this pre-clinical model of musculoskeletal decline.

PB125 treatment ameliorates disease/age-related declines in mitochondrial respiration
PB125-treated guinea pigs had augmented mitochondrial respiration in 5 mo females and 15 mo males as characterized by greater ADP V max and electron transport system capacity ETS (CI-CIV) . PB125 did not augment coupled respiration despite improvements in uncoupled respiration, which may reflect enhanced reserve capacity suited to handle greater electron flux. Importantly, PB125 attenuated disease/age-related dysfunction of complex I and II-supported coupled and uncoupled respiration and fatty acid oxidation in both sexes. Notably, PB125 selectively attenuated the disease/age-related decline in coupled respiration in males and uncoupled respiration in females. PB125 attenuated the disease/age-related declines in mitochondrial efficiency/coupling in males only. In humans, mitochondrial coupling decreases with age (Kumaran et al., 2005). Exercise-induced attenuation in loss of mitochondrial efficiency/coupling with age (Conley et al., 2013) has led researchers to speculate that improving mitochondrial efficiency may help attenuate sarcopenia (Harper et al., 2021). Interestingly, in the presence of rotenone, a complex I inhibitor, there was no effect of PB125, which suggests that PB125 improves mitochondrial respiration through improvements in complex I function.
Though PB125 does stimulate Nrf2 activation (as measured by a promoter/reporter assay) in vitro (Hybertson et al., 2019), the changes we observed in Nrf2 content with PB125 treatment were not consistent between young and older guinea pigs. Instead, PB125 had a significant interaction with age, increasing Nrf2 content in 5 mo guinea pigs, and decreasing Nrf2 protein expression in 15 mo guinea pigs. HO-1 content showed a similar pattern. Though we have no acute in vivo data to support this, we hypothesize that PB125 activated Nrf2, as demonstrated by greater Nrf2 content in 5 mo guinea pigs. Ten months of treatment with PB125 led to consistent Nrf2 activation, leading to greater downstream antioxidant enzyme content. With age, we hypothesize that greater basal levels of ROS led to a greater Nrf2 content in 15 mo CON guinea pig. We further posit that PB125 ameliorated the increase in age-related ROS, which led to significantly lower levels of Nrf2 and HO-1. This pattern is similar to aerobic exercise, which upregulates antioxidative capacity and protects from age-related increases in oxidative stress (Muthusamy et al., 2012). In humans, Nrf2 expression increases with age; however, aerobic exercise training reduces levels of Nrf2 (Ostrom & Traustadottir, 2020). We speculate a similar pattern occurred with PB125 treatment, though further work is required to understand how PB125 affects Nrf2 activation in vivo in both young and old organisms. Similar to many drugs and supplements, it is quite possible that PB125 has widespread effects that are not mediated through a single pathway.
A different Nrf2 activator, sulforaphane, similarly improved complex I function (Bose et al., 2020) as PB125 did in the present study. The pathways underlying the effect of Nrf2 activation on mitochondrial respiration are not entirely understood. Several studies have demonstrated that Nrf2 is a central mediator for exercise-induced mitochondrial biogenesis and improvements in mitochondrial function (D'Souza et al., 2020;Islam et al., 2020;Merry & Ristow, 2016). Interestingly, both Nrf2-related redox signalling (Safdar et al., 2010) and complex I function decrease with age in skeletal muscle (Kruse et al., 2016). Thus, PB125 may target a critical mechanism that contributes to mitochondrial dysfunction, though the specific mechanisms by which Nrf2 activation might contribute to complex I function remain to be elucidated.
As a master regulator of cytoprotective gene transcription, Nrf2 is a critical component of redox homeostasis. Skeletal muscle mitochondria of aged Nrf2 knock-out mice emit significantly more ROS than aged wild-type counterparts reflecting the role of Nrf2 in regulating redox balance (Kitaoka et al., 2019). In vitro, Nrf2 knock-out models have compromised complex J Physiol 601.11 I activity due to impairments in NADH availability (Kovac et al., 2015). Importantly, pyruvate dehydrogenase is a redox sensitive enzyme responsible for supplying NADH to complex I (Fisher-Wellman et al., 2015). Thus, age-related increases in oxidative stress may constrain the supply of NADH to complex I, which would explain age-associated decline in complex I capacity and how NAD + supplementation restores mitochondrial respiratory capacity (Kruse et al., 2016;McElroy et al., 2020). In our study, PB125 increased mitochondrial respiration, particularly in complex I, which may have been mediated by improved cellular redox regulation. However, future studies will need to more rigorously investigate the effect of PB125 on redox homeostasis.
Another potential mechanism by which PB125 enhanced mitochondrial respiration in this study is through greater mitochondrial protein turnover. However, there was a decline in mitochondrial biogenesis as these guinea pigs matured and OA developed, suggesting that, in order to maintain mitochondrial density, degradation of mitochondrial proteins (i.e. mitophagy or ubiquitin-dependent degradation of mitochondrial proteins) also declined. Impaired mitophagy contributes to mitochondrial dysfunction and disease in humans (Gouspillou et al., 2018;Newman & Shadel, 2018;Ryu et al., 2016). Importantly, PB125 attenuated the age/disease-related decline in mitochondrial protein synthesis, suggesting that declines in degradation/mitophagy may have also been attenuated, though we did not directly measure this. As such, we posit that mitochondrial protein turnover, which is essential for maintenance of overall mitochondrial function (Szczepanowska & Trifunovic, 2021), was maintained in 15 mo PB125 guinea pigs compared to 15 mo CON guinea pigs in this study. Others have also demonstrated that Nrf2 activators play a role in modulating mitochondrial protein turnover. In C. elegans the Nrf2 homolog mediated tomatidine (a Nrf2 activator)-induced mitophagy (Fang et al., 2017). Our group has demonstrated that Protandim, also a phytochemical Nrf2 activator, enhanced mitochondrial protein turnover in wheel-running rats (Bruns et al., 2018). Thus, Nrf2 activation seems to preserve mitochondrial protein turnover in 15 mo guinea pigs while turnover may have declined in 15 mo CON guinea pigs. Interestingly, the Konopka group published mitochondrial respiration data in Hartley guinea pigs treated with either rapamycin or rapamycin combined with metformin, interventions reported to have anti-ageing effects, following similar mitochondrial respiration protocols as the present study. Surprisingly, they reported that both rapamycin and rapamycin combined with metformin had deleterious effects on mitochondrial respiration  and worsened OA severity . In contrast, the PB125-induced improvements in the present study also led to ameliorating OA severity in the same guinea pigs (in preparation). Altogether, these data support that PB125 can attenuate age related declines in mitochondrial respiration, which may in turn lead to the maintenance of musculoskeletal health.

PB125 attenuates declines in components of protein homeostasis
Decline in mechanisms to maintain proteostasis (which include not only protein synthesis and degradation, but also chaperone-mediated folding and protein trafficking; Noack et al., 2014) contributes to musculoskeletal dysfunction (Kaushik & Cuervo, 2015;Santra et al., 2019). There is limited insight on the effect of either age or OA progression on protein homeostasis in humans, though basal protein synthesis appears to be unchanged with age in humans (Brook et al., 2016;Volpi et al., 2001). Moreover, it is unclear whether or not both men and women experience a decline in proteostasis with age. While men generally have greater muscle mass than women, men also lose muscle mass faster and muscle strength to a greater degree; however, women are less fatigue resistant (thoroughly reviewed in Gheller et al., 2016). In the present study, we documented declines in protein synthesis in all subfractions of the soleus and gastrocnemius muscles of both male and female Hartley guinea pigs. There were no sex differences in fractional synthesis rates in either muscle. This is the first study to characterize declines in protein synthesis in female Hartley guinea pigs. However, these declines could be attributed to the fact that we compared skeletal muscle protein synthesis rates of 5 mo guinea pigs, while they are still growing, to 15 mo guinea pigs, which have plateaued in body weight ( Fig. 2A and B). We have previously documented that rates of skeletal muscle protein synthesis in 15 mo male Hartley guinea pigs are slower than 9 mo Hartley guinea pigs, which is the age at which growth plateaus (Fig. 1F and G) (Musci et al., 2020). Thus, the decline in protein synthesis observed in the present study may be a consequence of both reaching maturity and musculoskeletal decline consistent with either musculoskeletal disease progression or advancing age.
We documented differences in long-term protein synthesis rates to minimize the bias of proteins turning over faster (Miller et al., 2015); however, it is possible that our approach may still not identify differences in fractional synthesis rates between ages if the protein pools subject to turnover (i.e. the dynamic protein pools) are not the same between the 5 and 15 mo guinea pigs. As recently demonstrated by Abbott and colleagues, the dynamic protein pool declines with age and thus can obscure the fractional synthesis rates and biases towards aged animals having lower synthesis rates (Abbott et al., 2021). The approach the authors utilized raises important considerations for evaluating the effect of age or interventions on protein turnover in future studies. Employing such an approach may also help reconcile differences in observations on the effect of age on protein turnover between species Musci et al., 2020;Volpi et al., 2001) and more accurately describe the age-related effects on protein kinetics.
The mechanisms by which PB125 attenuated the decline in protein synthesis are not entirely clear. It is worth noting that the attenuation in protein synthesis occurred only in the soleus. It is unclear why there is a muscle-specific effect. However, the soleus muscle is primarily comprised of type I myofibres and gastrocnemius is much more mixed (Musci et al., 2020). PB125-mediated improvements in mitochondrial respiration likely alleviated energetic constraints. Protein turnover is energetically demanding, accounting for nearly 35% of basal metabolism (Bier, 1999;Rolfe & Brown, 1997;Waterlow, 1984). Impairments in mitochondrial function consequentially constrain the amount of energy dedicated to proteostasis. Mitochondrial dysfunction precedes the loss of proteostasis in skeletal muscle, which leads to declines in function (Ben-Zvi et al., 2009;Gaffney et al., 2018). Moreover, other interventions that attenuate the decline in or improve mitochondrial function also improve proteostatic mechanisms and preserve overall muscle function. For example, maintaining physical activity and caloric restriction in rodents delays declines in mitochondrial respiration as well as skeletal muscle function (Stolle et al., 2018;Zangarelli et al., 2006), a similar observation made in masters athletes (Zampieri et al., 2015). While both exercise and caloric restriction have broad effects, more targeted interventions focused on improving mitochondrial function also report a similar phenomenon: enhancing mitochondrial function delays skeletal muscle dysfunction (Campbell et al., 2019;Gaffney et al., 2018). This observation occurs in other tissues as well. Increasing mitochondrial proteostasis decreases proteotoxic amyloid aggregation in cells, increasing fitness and lifespan in C. elegans (Sorrentino et al., 2017). These studies emphasize the importance of mitochondrial respiration and the production of ATP to facilitate proteostatic mechanisms. In humans, aerobic exercise improves mitochondrial function through mitochondrial remodelling and improves skeletal muscle function (Greggio et al., 2017). Altogether, our data support the posit that PB125-mediated improvements in mitochondrial respiration alleviated constraints in energy, which led to a greater amount of ATP available to support proteostasis.
Another mechanism by which PB125 may have attenuated declines in skeletal muscle proteostasis is through the mitigation of inflammation and oxidative stress, which can have deleterious effects on protein turnover, particularly protein synthesis. Protein synthesis, at rest, appears to be no different between young and old individuals (Brook et al., 2016;Volpi et al., 2001). However, age-related inflammation and oxidative stress can blunt the anabolic response to stimuli such as exercise or feeding. This concept, termed anabolic resistance, is a contributor to age-related musculoskeletal dysfunction and appears to blunt the anabolic response to resistance exercise training (Brook et al., 2016;Cuthbertson et al., 2005;Wilkes et al., 2009). Interventions designed to mitigate age-related increases in oxidative stress or inflammation seem to improve skeletal muscle anabolic responses to exercise (Trappe et al., 2002), feeding (Rieu et al., 2009;Smiles et al., 2019) and insulin (Rivas et al., 2016). There is an increase in some systemic inflammatory mediators in these Hartley guinea pigs, which may be associated with OA (Huebner et al., 2007;Santangelo et al., 2011). Importantly, PB125 stimulates transcription of endogenous antioxidant and anti-inflammatory genes (Hybertson et al., 2011(Hybertson et al., , 2019. Thus, it is possible that PB125 treatment ameliorated oxidative stress and inflammation and improved the anabolic response to feeding. However, because we measured cumulative protein synthesis over 30 days, rather than acutely in response to an anabolic stimulus such as feeding, we cannot determine if there were any changes specifically in the anabolic response to feeding. Future studies should investigate the efficacy of PB125 on abrogating inflammation and oxidative stress and, in acute settings, determine whether targeting Nrf2 could enhance the anabolic response to feeding or activity. In the present study, there was no observed effect of treatment on ROS emission or protein oxidation. However, our lab has previously demonstrated that another Nrf2 activator increases antioxidant protein expression  and augmented protein synthesis related to proteostasis in skeletal muscle of rats in response to wheel running exercise (Bruns et al., 2018). Thus, treatment with Nrf2 activators may represent a class of interventions that augment adaptation to acute stressors .

Future directions
The improvements in mitochondrial respiration and proteostasis did not translate to statistically significant improvements in sustained mobility. However, it is important to note that our measure of mobility is only one metric of musculoskeletal function. Additionally, musculoskeletal function is not the only factor that dictates mobility. For example, we recently demonstrated that these guinea pigs exhibit hallmarks of human J Physiol 601.11 Alzheimer's disease that could explain impairments in integrative functional outcomes such as gait (Wahl et al., 2022). Thus, it is still important to assess additional and more specific components of musculoskeletal function, as mitochondrial function is a strong determinant in physical function such as gait speed and grip strength in humans (Gonzalez-Freire et al., 2018).
Another observation that warrants further investigation is the sex-specific effect of PB125. Other interventions involving Nrf2 activators have also demonstrated sex-specific effects. The Interventions Testing Program reported that treatment with the Nrf2 activator Protandim extended median lifespan in heterogeneous male mice, but not females (Strong et al., 2016). Our lab has also previously demonstrated that Protandim only improved myofibrillar proteostasis in men (Konopka et al., 2017). Other healthspan promoting interventions, such as a metformin and rapamycin, also have sex specific effects (Strong et al., 2016). We have begun interrogating these sex specific effects through the use of kinetic proteomics (Wolff et al., 2019;Wolff et al., 2021). Moving forward, it will be necessary to interrogate the sex differences in the present study and the implications they have on the efficacy of PB125 to attenuate musculoskeletal decline in men and women, as they experience similar prevalence worldwide (Shafiee et al., 2017).
A limitation of the study was comparing guinea pigs still in the growth/maturation process to those that have finished growing. While Hartley guinea pigs reach sexual maturity by 4 mo, our data demonstrate they do not cease growing until 9 mo. Nonetheless, even at a relatively young age, these Hartley guinea pigs are an established model of age-related knee osteoarthritis (Jimenez et al., 1997;Santangelo et al., 2014). Thus, some of the comparisons made could be a consequence of comparing guinea pigs that are growing to those that are not. Regardless, the phenotype exhibited by these guinea pigs still reflects age-related OA and musculoskeletal decline in humans. The aim of our study was to determine whether long term supplementation of a phytochemical compound could ameliorate the progression of age-related OA and musculoskeletal decline documented in this model Minton et al., 2021;Musci et al., 2020). Further, our group has also documented that brains of these guinea pigs exhibit hallmarks of human brain ageing (Wahl et al., 2022). Future studies should focus on comparisons in these guinea pigs after the growth phase. Such studies would reveal how the musculoskeletal system changes after maturation and while OA continues to progress, as well as provide opportunities to test potential interventions that target both ageing and musculoskeletal dysfunction. However, a significant barrier to asking such questions is that OA is often so severe in Hartley guinea pigs at 18 mo that veterinarians often recommend euthanasia. However, future studies could study interventions between 10 and 18 mo, after the growth phase has been completed.
The decline in mitochondrial respiration while these guinea pigs were still maturing also raises interesting future directions, particularly in trying to determine how mitochondrial function may play a role in adequate organismal development. Another direction is to further interrogate the decline in mitochondrial function and protein synthesis during active growth. Of note, Hartley guinea pigs experience an increase in systemic inflammation during development and OA progression (Santangelo et al., 2011) and experience an increase in adiposity (Musci et al., 2020;Radakovich et al., 2019). Accordingly, it is tempting to suggest these Hartley guinea pigs may represent a model for adolescent obesity. However, additional studies are necessary comparing the development of Hartley guinea pigs to a strain of guinea pigs not as susceptible to inflammation and OA during development (i.e. up to 12 mo).

Conclusions
Musculoskeletal dysfunction is a primary contributor to disability and loss of independence in humans. There are few existing interventions that effectively mitigate the decline in skeletal muscle function, particularly in ageing individuals and/or adolescents and younger adults with OA. In this study, we further characterized a model of musculoskeletal dysfunction measuring mitochondrial respiration and protein synthesis in both male and female Hartley guinea pigs. Moreover, we tested a potential healthspan-extending phytochemical compound, PB125, which is currently in the NIA-ITP (https: //www.nia.nih.gov/research/dab/interventions-testingprogram-itp/compounds-testing), on mitochondrial respiration and proteostasis in this pre-clinical guinea pig model of musculoskeletal decline. We found that PB125 improved mitochondrial respiration and attenuated declines in protein synthesis, mechanisms that could contribute to improvements in function and longevity. This project adds to the growing literature that supports the use of Nrf2 activators to improve organismal health. The data from this study provide mechanistic insight by which a readily translatable intervention could mitigate musculoskeletal decline in humans.