Uric acid formation is driven by crosstalk between skeletal muscle and other cell types

Hyperuricemia is implicated in numerous pathologies, but the mechanisms underlying uric acid production are poorly understood. Using a combination of mouse studies, cell culture studies, and human serum samples, we sought to determine the cellular source of uric acid. In mice, fasting and glucocorticoid treatment increased serum uric acid and uric acid release from ex vivo–incubated skeletal muscle. In vitro, glucocorticoids and the transcription factor FoxO3 increased purine nucleotide degradation and purine release from differentiated muscle cells, which coincided with the transcriptional upregulation of AMP deaminase 3, a rate-limiting enzyme in adenine nucleotide degradation. Heavy isotope tracing during coculture experiments revealed that oxidation of muscle purines to uric acid required their transfer from muscle cells to a cell type that expresses xanthine oxidoreductase, such as endothelial cells. Last, in healthy women, matched for age and body composition, serum uric acid was greater in individuals scoring below average on standard physical function assessments. Together, these studies reveal skeletal muscle purine degradation is an underlying driver of uric acid production, with the final step of uric acid production occurring primarily in a nonmuscle cell type. This suggests that skeletal muscle fiber purine degradation may represent a therapeutic target to reduce serum uric acid and treat numerous pathologies.


Introduction
Increased serum uric acid is an independent risk factor for the development and poor prognosis of many pathologies, including cancer (1), cardiovascular disease (2,3), chronic kidney disease (4), type 2 diabetes (5), and liver disease (6).It is also positively associated with disease risk factors, such as components of metabolic syndrome (7) and inflammation (8).The production of uric acid probably plays a direct role in the pathology since inhibition of xanthine oxidase (the enzyme that synthesizes uric acid) has been shown to alleviate contributing disorders such as endothelial dysfunction (9), insulin resistance (10), hepatic steatosis (11), and muscle atrophy (12)(13)(14)(15).Surprisingly, the mechanisms underlying increased uric acid production and its cellular source are poorly understood.
Uric acid is produced exclusively by the enzyme xanthine oxidoreductase (XOR), which catalyzes the irreversible oxidation of hypoxanthine and xanthine in the final step of purine degradation in humans.Hypoxanthine and xanthine are produced from the breakdown of purine nucleotide monophosphates (AMP, GMP, IMP, and XMP), which are generated from the catabolism of ATP/GTP, DNA/RNA, or de novo purine nucleotide synthesis.XOR is heterogeneously expressed in mammalian tissues, being most abundant in vascular endothelial cells, intestinal epithelial cells, hepatocytes, and adipocytes (16)(17)(18)(19).In cell types that do not express XOR, such as skeletal muscle fibers (16,19), purine nucleotide degradation is presumed to culminate in release of hypoxanthine and xanthine, which may undergo oxidation to uric acid in a XOR-expressing cell type or by circulating XOR (20,21).Thus, because XOR-expressing cells can oxidize their own purines and those released by non-XOR-expressing cells, the predominant cellular source(s) supplying purine precursors for uric acid formation remain unknown.
Several lines of evidence implicate skeletal muscle fibers as a major source of purine precursors used for uric acid formation during muscle wasting conditions.Diseases and conditions associated with Hyperuricemia is implicated in numerous pathologies, but the mechanisms underlying uric acid production are poorly understood.Using a combination of mouse studies, cell culture studies, and human serum samples, we sought to determine the cellular source of uric acid.In mice, fasting and glucocorticoid treatment increased serum uric acid and uric acid release from ex vivo-incubated skeletal muscle.In vitro, glucocorticoids and the transcription factor FoxO3 increased purine nucleotide degradation and purine release from differentiated muscle cells, which coincided with the transcriptional upregulation of AMP deaminase 3, a rate-limiting enzyme in adenine nucleotide degradation.Heavy isotope tracing during coculture experiments revealed that oxidation of muscle purines to uric acid required their transfer from muscle cells to a cell type that expresses xanthine oxidoreductase, such as endothelial cells.Last, in healthy women, matched for age and body composition, serum uric acid was greater in individuals scoring below average on standard physical function assessments.Together, these studies reveal skeletal muscle purine degradation is an underlying driver of uric acid production, with the final step of uric acid production occurring primarily in a nonmuscle cell type.This suggests that skeletal muscle fiber purine degradation may represent a therapeutic target to reduce serum uric acid and treat numerous pathologies.

Results
Skeletal muscle uric acid efflux is increased after fasting and dexamethasone treatment.Increased serum uric acid levels are common among diseases and conditions in which muscle atrophy is prevalent (22).To determine if atrophying skeletal muscles have increased purine nucleotide degradation, we measured release of purine nucleotide degradation products (hypoxanthine, xanthine, uric acid) from incubated mouse extensor digitorum longus (EDL) muscles after fasting, a well-characterized model of accelerated protein degradation and atrophy (27)(28)(29).Fasting for 48 hours reduced body and EDL weight (Figure 1, A and B) and increased serum uric acid (Figure 1C).Incubated EDL muscles from fasted mice produced greater uric acid compared with nonfasted controls (Figure 1D).No differences were detected in release of hypoxanthine or xanthine.Further, no differences in protein expression of the adenine nucleotide-degrading enzymes AMPD1, AMPD3, cytosolic 5′nucleotidase 1 (NT5C1A), or XOR were detected in the EDL muscles (Figure 1E).Notably, we did detect an increase in AMPD3 protein levels in fasted tibialis anterior (TA) muscles (Figure 1F).Unfortunately, TA muscles are too large to allow sufficient oxygen perfusion during ex vivo incubations (30); hence, purine efflux measures as done here are not feasible in TA muscles.
Increased muscle protein degradation in response to fasting is partly mediated by increased circulating glucocorticoids (31).To determine if increased glucocorticoid exposure is sufficient to increase muscle purine efflux, mice were injected with the synthetic glucocorticoid, dexamethasone (DEX), daily for 5 days.DEX treatment did not cause a significant reduction in EDL muscle weights in male or female mice (Figure 2A) but did cause a 17% increase (P = 0.051) in serum uric acid (Figure 2B).Additionally, DEX increased 3-MH release from incubated EDL muscles (Figure 2C), indicating that myofibrillar protein degradation rate was increased (32,33).Incubated EDL muscles from DEX-treated mice also released more uric acid compared with vehicle treated (Figure 2D), leading to a positive correlation between uric acid and 3-MH release (Figure 2D).Importantly, no differences were found between serum uric acid or release of 3-MH, hypoxanthine, xanthine, or uric acid between males and females (Supplemental Figure 1; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.171815DS1).Therefore, the sexes were combined for the above analysis.No differences in protein expression of purine nucleotide-degrading enzymes were detected in DEX-treated EDL muscles (Figure 2E).However, similar to fasted mice, an increase in AMPD3 protein was observed in DEX-treated TA muscles (Figure 2F).
Purine nucleotide degradation and purine efflux are increased in atrophying muscle cells.Muscle tissue comprises numerous cell types, such as muscle fibers, endothelial cells, macrophages, fibro-adipogenic progenitor cells, and satellite cells.Immunohistochemical staining of muscle has shown XOR expression is overwhelmingly localized to vascular cells (16,17).To determine if purine nucleotide degradation is specifically upregulated in muscle cells during atrophy, purine release was measured from C2C12 myotubes during DEX and/or serum starvation (S.S.) treatments.After 48 hours, myotube atrophy was confirmed by quantifying myosin heavy chain (MyHC) area (Figure 3A).Compared with vehicle, myotubes treated with 10 or 100 μM DEX had less MyHC area but no differences in nuclei count (Figure 3B).S.S. myotubes had the lowest MyHC area and fewest nuclei (Figure 3B).The reduction in MyHC area was preceded by increases in protein degradation rates between 6 and 24 hours after DEX and/or S.S. treatment (Figure 3C).JCI Insight 2024;9(2):e171815 https://doi.org/10.1172/jci.insight.171815 After verifying these treatments cause atrophy, we determined if purine nucleotide degradation was increased by measuring hypoxanthine, xanthine, and uric acid in the media.Treatment with 100 mm DEX caused a significant increase in hypoxanthine and xanthine release by 12 hours, eventually reaching peak levels at 48 hours that were 3.6-and 1.8-fold greater than vehicle, respectively (Figure 3D).The increase was similar among the different DEX doses tested (Supplemental Figure 2).S.S.+Veh also caused an increase in hypoxanthine release, which peaked at 24 hours and then remained unchanged (Figure 3D).S.S.+DEX resulted in the greatest hypoxanthine release among groups, and similar to S.S.+Veh, remained unchanged after peaking at 24 hours (Figure 3D).Interestingly, S.S. negatively affected xanthine release, with S.S.+Veh groups releasing less xanthine than Veh treated, and S.S.+DEX groups releasing more than S.S.+Veh and Veh, but less than DEX (Figure 3D).Importantly, unlike its precursors hypoxanthine and xanthine, uric acid was undetected in all groups until 36 hours, when it was detected in DEX groups at a miniscule 0.4 nmol/well (Figure 3D).By 48 hours it had risen to 2 nmol/well in the DEX group and 0.5 nmol/well in the S.S.+DEX group, respectively (Figure 3D).These increases in purine efflux in response to DEX coincided with 4-fold increased AMPD3 protein expression (Figure 3, E and F).In contrast, XOR was undetectable except for a faint, but unquantifiable, band in the S.S.+DEX condition (Figure 3F).
Increased FoxO3 activity is sufficient to increase purine nucleotide degradation and AMPD3 protein expression in myotubes.Glucocorticoids and fasting induce protein degradation and muscle atrophy, in part, by increasing the activity of FoxO transcription factors (28,29,34).Therefore, to determine if increased FoxO activity is sufficient to stimulate muscle purine nucleotide degradation, we infected C2C12 myotubes with an adenovirus encoding a constitutively active FoxO3 (caFoxO3) (28,35).As expected, myotubes expressing caFoxO3 had less MyHC area but similar nuclei count compared to GFP controls (Figure 4, A and B).Moreover, caFoxO3 increased hypoxanthine and xanthine release, while uric acid was undetected (Figure 4C), demonstrating that increased FoxO activity is sufficient to stimulate purine nucleotide degradation in muscle.As with DEX treatment, the elevated production of hypoxanthine and xanthine with caFoxO3 expression coincided with demonstrable upregulation of AMPD3 (Figure 4D).
The AMPD3 promoter region contains a consensus FoxO binding site 124 bases upstream of the transcriptional start site.To determine if increased AMPD3 protein is related to FoxO-dependent activation of its promoter region, myoblasts were transfected with custom luciferase reporter plasmids containing 1.1 kb of the mouse AMPD3 proximal promoter region, with or without a mutation in the FoxO binding site (ΔFoxO mutant).CaFoxO3 expression was sufficient to increase AMPD3 promoter activity by 3.9-fold, which was prevented by mutating its binding site in the AMPD3 promoter region (Figure 4E).Moreover, AMPD3 promoter activity was increased by DEX and S.S. treatment by a FoxO-dependent mechanism (Figure 4F).
Purines released from atrophying myotubes drive uric acid formation by XOR-expressing cells.Since whole muscle tissues release uric acid (Figures 1 and 2), but uric acid production by muscle cells is extremely slow (Figures 3 and 4), we next sought to verify that xanthine oxidase is expressed by some muscle-resident cell types, but not muscle fibers, in whole muscle.Immunofluorescence staining for xanthine oxidase was performed on cross sections of TA muscles of male and female mice.As expected, XOR was not detectable in the muscle fibers but was detectable in numerous other cells, many of which expressed the endothelial cell marker CD31 (Figure 5A).Importantly, not all XOR-positive cells were CD31 positive, suggesting that endothelial cells are not the exclusive source of XOR in whole skeletal muscle.
To determine if myotube purines could be oxidized to uric acid by XOR-expressing cell types, first, XOR protein expression was confirmed in bovine aortic endothelial cells (BAOECs) (36), 3T3-L1 adipocytes (18), and AML-12 hepatocytes (18) by Western blot (Figure 5B).Next, we tested whether these cells could oxidize physiological levels of exogenous purines.All 4 cell types (C2C12, BAOEC, 3T3-L1, and AML-12) consumed hypoxanthine, as demonstrated by progressive reductions over time in hypoxanthine media concentrations (Supplemental Figure 3, A, D, G, and J).However, only XOR-expressing cells consumed xanthine (Supplemental Figure 3, B, E, H, and K).In cell types expressing XOR, hypoxanthine and xanthine both caused dramatic increases in uric acid production, which was prevented by the XOR inhibitor allopurinol (Figure 5C and Supplemental Figure 3,  C, F, I, and L).We then tested whether coculturing myotubes with BAOECs would enable the oxidation of purines released  from myotubes and whether a myotube atrophy stimulus would augment uric acid formation.Prior to coculturing, myotube purines were labeled with 13 C 15 N-glycine, which supplies 2 carbons and 1 nitrogen atom to the purine ring during de novo purine synthesis (Figure 5E).At the start of the coculture, 13 C 15 N-glycine medium was replaced with nonlabeled medium containing vehicle or DEX, then added to wells containing C2C12 only, BAOEC only, or C2C12+BAOEC.As before in C2C12 cells only (Figure 3D), DEX increased hypoxanthine, xanthine, and uric acid in the media (Figure 5D).However, the coculture of C2C12+BAOEC led to 3.5-and 9.5-fold greater uric acid in C2C12+BAOEC versus C2C12 or BAOEC only (Figure 5D).In addition, heavy-labeled uric acid (170 Da) was only detected in wells containing myotubes and was 2.8-fold greater in C2C12+BAOEC than C2C12 after DEX (Figure 5F).Despite the substantial increase in uric acid production in C2C12+BAOEC cocultures, the percentage of heavy-labeled uric acid per total uric acid was similar between C2C12 and C2C12+BAOEC after DEX (Figure 5G).This indicates that the increased uric acid produced in the cocultures is derived solely from the myotube purines, not from greater purine degradation in the BAOECs.
Serum uric acid is greater in individuals with low physical performance.Deficits in contractile performance are known to precede muscle mass loss during aging and atrophy-inducing conditions (37)(38)(39).We showed earlier that muscle purine release was elevated prior to increased protein degradation and myotube atrophy and was equally increased in myotubes given DEX doses insufficient to increase protein degradation and cause atrophy (Figure 3 and Supplemental Figure 2).Based on these findings, we questioned whether serum uric acid levels would be higher in individuals with lower muscle performance.Since serum uric acid increases with lean body mass and may differ by age (40,41), female participants were matched for age, height, weight, body mass index (BMI), and estimated lean body mass (appendicular mass/height 2 ) (Table 1).Low performers had lower max grip strength, longer time to complete 5 sit-to-stands, fewer sit-to-stands completed in 30 seconds, and a trend (P = 0.06) for shorter 6-minute walking distance (Table 1).Serum uric acid levels were significantly greater in low (3.59 ± 1.04 mg/dL) versus average (2.86 ± 0.49 mg/dL) performance groups (Figure 6), but no differences in serum creatinine or creatine kinase activity (markers of kidney function, muscle mass, and muscle damage) were detected (Table 1).
Serum 3-MH, an amino acid found predominantly in myofibrillar actin and myosin (32), can be used as a marker of muscle protein degradation.However, serum 3-MH is also influenced by dietary meat intake (32).Therefore, the ratio of 3-MH to 1-MH (which is uniquely found in nonhuman animal muscle; ref. 32) can be used to more accurately estimate muscle protein degradation during nonfasted states (42).Serum 3-MH/1-MH trended higher in the low-performance group (P = 0.06, Figure 6), suggesting accelerated muscle protein degradation in the low performers.

Discussion
The present study sought to determine if purine nucleotide degradation and purine release are increased in skeletal muscles undergoing atrophy and whether elevated muscle purine release could stimulate uric acid synthesis.In mice, fasting-and glucocorticoid-induced protein degradation coincided with increased serum uric acid and uric acid efflux from isolated EDL muscles.In cultured myotubes, glucocorticoidand caFoxO3-induced treatment coincided with increased hypoxanthine and xanthine efflux but little to no uric acid production, which is consistent with poor XOR expression in muscle cells.Coculturing myotubes with BAOECs, which do express XOR, enabled the oxidation of myotube-released hypoxanthine and xanthine to uric acid.Last, in humans, serum uric acid was greater in individuals with below-average physical performance who also had a trend for increased serum 3-MH/1-MH ratio, an indicator of increased muscle protein degradation.Thus, our findings demonstrate that purine nucleotide degradation and purine release are increased in muscle cells undergoing atrophy, and muscle purine release stimulates uric acid synthesis by XOR-expressing cells, such as endothelial cells.Given that skeletal muscles are the most abundant bodily tissue, contain the greatest concentrations of adenine nucleotides (43), and have immunofluorescence images after staining for myosin heavy chain (MyHC) and nuclei (DAPI) in vehicle, 100 μM DEX, S.S.+Veh., or S.S.+100 μM DEX.(B) Quantifications of MyHC area and nuclei count per well.n = 6 wells/condition and 100 images/well, *=P < 0.05, **=P < 0.01, 1-way ANOVA with Tukey's multiple comparisons.(C) Protein degradation rates were determined as media accumulation of 13 C 9 15 N-phenylalanine 6-24 hours after dexamethasone treatment or after S.S.+Veh., S.S.+100 μM DEX, or untreated controls.n = 6 wells/condition, *=P < 0.05, **=P < 0.01 vs. Veh/control, 1-way ANOVA with Dunnett's multiple comparisons.(D) Media concentrations of the purine nucleotide breakdown products hypoxanthine, xanthine, and uric acid during 100 μM DEX and/or S.S. treatment.*=P < 0.05 DEX vs. Veh., # =P < 0.05 S.S.+Veh.vs. Veh., † =P < 0.05 S.S.+Veh.vs. S.S.+DEX 100 μM.n = 6 wells/condition, 2-way ANOVA with Tukey's multiple comparisons.(E and F) AMPD3 protein expression after 48 hours of DEX and/or S.S. treatments.*=P < 0.05 vs. Veh.or indicated group.

R E S E A R C H A R T I C L E
JCI Insight 2024;9(2):e171815 https://doi.org/10.1172/jci.insight.171815enzymatic capacity for de novo purine synthesis (44), our data suggest that skeletal muscle is a major precursor source of circulating uric acid.
Our incubated mouse muscle experiments agree with studies of exercise in humans demonstrating that whole muscles can degrade purines and release uric acid into the circulation (20,21).However, our series of subsequent in vitro experiments extend these findings to explore the role of different cell types.First, we found that induction of protein degradation and/or atrophy of multinucleated muscle cells (C2C12 myotubes) resulted in increased purine nucleotide degradation, which culminated predominantly in the release of hypoxanthine and xanthine, not uric acid.Further, XOR-expressing cells readily converted exogenous hypoxanthine and xanthine to uric acid, which was prevented by the XOR inhibitor allopurinol.Last, coculturing myotubes with BAOECs (which express XOR) enabled the oxidation of myotube-derived purines to uric acid, which was enhanced when the myotubes were undergoing atrophy.Therefore, the observed release of uric acid from Interestingly, the atrophying EDL muscles and myotubes had increased purine nucleotide degradation despite ample access to exogenous nutrients and oxygen, without any contractile demand.Increased skeletal muscle purine nucleotide degradation and purine efflux readily occur during hypoxia and intense contractions (21,45), during which ATP synthesis is unable to match ATP hydrolysis.This mismatch leads to increased activation of AMPD and, thus, degradation of nucleotides to maintain the cellular energy state (20,21,45).Therefore, our findings may reflect the ability of fasting, glucocorticoids, and FoxO to impair muscle bioenergetic function to an extent that resting ATP/ADP could not be maintained.This possibility is supported by NMR measurements of decreased phosphocreatine and calculated free energy of ATP hydrolysis (ΔG ATP ) in rat gastrocnemius muscle after DEX treatment (46).This

R E S E A R C H A R T I C L E
JCI Insight 2024;9(2):e171815 https://doi.org/10.1172/jci.insight.171815would be consistent with increased uric acid release without changes in AMPD1/3, NT5C1A, and XOR expression in the EDL muscles.
On the other hand, FoxO-mediated transcriptional upregulation of purine nucleotide-degrading enzymes could have increased purine nucleotide degradation without increases in AMP/GMP/IMP levels.This possibility is supported by our findings of increased AMPD3 protein expression in TA muscles after fasting and DEX treatment, and increased AMPD3 protein expression in myotubes after DEX treatment and caFoxO3 expression, and our previous reports showing AMPD3 overexpression is sufficient to increase adenine nucleotide degradation in TA muscles and myotubes (23,24).
The ability of DEX and caFoxO3 to increase AMPD3 protein expression likely stems from their activation of the AMPD3 promoter region and its subsequent gene transcription, which we demonstrated to depend on the presence of FoxO binding sites.This agrees with previous studies indicating the glucocorticoid receptor (47) and FoxO transcription factors (29) as regulators of AMPD3 gene expression.In humans, oral dosing of DEX caused an upregulation of AMPD3 mRNA in skeletal muscle (48).In C2C12 myotubes, Kuo et al. identified AMPD3 as one of 147 genes containing glucocorticoid binding regions in close proximity to their genomic sequence that was also upregulated by DEX treatment (47).Transgenic mice with a muscle-specific knockout of all FoxO isoforms have attenuated AMPD3 mRNA upregulation in response to fasting-and denervation-induced atrophy (29).Given that AMPD3 expression is controlled by glucocorticoids and FoxO, which are required for muscle atrophy in response to a diverse set of upstream etiologies (31), our findings suggest that an AMPD3-mediated increase of muscle purine nucleotide degradation is a common feature of muscle atrophy.
We found that women with below-average physical performance, but equal levels of lean body mass, had higher serum uric acid (UA) and a trend for increased serum 3-MH/1-MH, suggesting that these women have accelerated myofibrillar protein turnover.This agrees with a published report showing hyperuricemia is associated with poor muscle strength in Japanese men (49) and mimics our animal data showing a positive correlation between EDL 3-MH and UA release without muscle loss in DEX-treated mice.Therefore, serum UA levels may be a sensitive biomarker for accelerated muscle protein degradation.Since reductions in contractile function and increases in protein degradation often precede loss of skeletal muscle mass and sarcopenia (25,37,38), our findings also suggest that serum UA could be of clinical use for predicting risk of muscle atrophy.However, it is important to note that our measures of muscle function are not able to discern whether the weakness/poor performance is due to muscle deficiencies, neural changes, or some combination of both.Longitudinal studies comparing serum UA, muscle mass, neuromuscular function, and direct measures of muscle protein degradation rates during atrophic and nonatrophic conditions would further elucidate the utility of UA as a predictor or diagnostic biomarker for sarcopenia.
The results from this study lend insight into targeting XOR as a therapy.Treatment with XOR inhibitors, such as allopurinol and febuxostat, has been shown to have numerous therapeutic effects in humans and mice, such as improved endothelial function and reduced cardiovascular disease-related incidents (50), chronic kidney disease progression (51), cancer cachexia and mortality (52), and disuse-mediated muscle atrophy (12)(13)(14).Mechanistically, these effects often are attributed to preventing oxidative stress caused by XOR-produced superoxide anion.Our results suggest that muscle purine nucleotide degradation could be an upstream event causing increased XOR activity and various tissue pathologies.Indeed, loss of ATP may be a cause of dysfunction in skeletal muscle in numerous diseases (53).Further, this may explain why the loss of muscle strength, even if not corrected for muscle mass, is a prognostic factor in all-cause mortality (54).Another consideration is that systemic allopurinol treatment, while generally safe, has resulted in serious side effects, especially when given together with particular other drugs (55).Therefore, there is still a need for improved therapeutics to limit XOR activity.Future studies examining the contribution of skeletal muscle purine nucleotide degradation and cell-specific XOR activity on peripheral tissue pathologies may lead to developing novel therapeutic strategies.
In conclusion, skeletal muscle purine nucleotide degradation and purine release are increased in skeletal muscles with poor contractile function and increased protein degradation.Mechanistically, glucocorticoids and FoxO3 increase muscle purine nucleotide degradation by transcriptionally upregulating AMPD3 expression, which increases muscle purine release and drives downstream XOR activity and UA production by XOR-expressing cells.As such, preventing muscle purine nucleotide degradation could be a novel strategy for treating various pathologies associated with increased XOR activity and hyperuricemia.

Methods
Animals.C57BL/6J mice (13-15 weeks) from Jackson Laboratory were housed 2-5 mice per cage with free access to food and water.For fasting studies, female mice were placed in a clean cage at 10 am with access to water and food, or water only, for 48 hours.For DEX studies, male and female mice were subcutaneously injected once daily for 5 days with 5 mg/kg body weight veterinary grade DEX (Dexium, Bimeda) or vehicle.
Purine quantification.Media samples were diluted 1:4 with cold (-20°C) 80% methanol with subsequent centrifugation at 15,000g for 10 minutes at 10°C to remove proteins.The supernatant was lyophilized using a Vacufuge plus (Eppendorf) and resuspended in H 2 O before UPLC analysis with Waters Acquity UPLC H-Class Bio system, as we have done previously (57).
Amino acid quantification.Amino acids were measured after derivatization of amine groups using the AccQ-Tag Ultra kit (Waters 186003836).Briefly, proteins were precipitated with methanol, and the supernatant was lyophilized using a Vacufuge plus.Samples were reconstituted in water and derivatized by combining 70 μL borate buffer, 15 μL sample, and 15 μL AccQ-Tag Ultra reagent.The samples were analyzed with an AccQ-Tag Ultra C18, 1.7 μm column (Waters 186003837) on a Waters Acquity UPLC system and quantified by absorbance at 260 nm.
Protein degradation rates.We adapted our previous protein degradation assay using 3 H-labeled proteins (56,58) to use stable isotope labeling.C2C12 myoblasts were differentiated in DMEM supplemented with 2% horse serum and 13 C 9 15 N-phenylalanine (130 mg/L, MilliporeSigma 608017).Four days later, myotubes were washed twice with unlabeled DMEM, with additional washes repeated 1 and 2 hours later to allow degradation of short-lived proteins.After the second chase period, atrophy treatments were initiated, and media samples were collected over time for measurement of 13  AMPD3 promoter luciferase reporter assay.The 1.1 kb genomic promoter sequence ranging from 1,000 bases 5′ to 100 bases 3′ of the mouse AMPD3 transcriptional start site (TSS) was synthesized by GenScript and inserted in the XhoI/HindIII cloning site of the pNL1.1[Nluc]reporter plasmid (Promega).A consensus FoxO binding site, GTAAACAACTG, is located at 124 bases upstream of the TSS.To inactivate this site, a Q5 Site-directed Mutagenesis Kit (New England Biolabs) was used to mutate the original sequence to GTCCACAACTG.This is termed the ΔFoxO mutant.The accuracy of the wild-type and ΔFoxO promoter sequences was confirmed by Sanger sequencing (Azenta Life Sciences).Myoblasts were transfected 24 hours prior to differentiation.Nano-Glo luciferase activity reporter assays (Promega, N1630) were conducted according to the manufacturer's protocol.Briefly, myotubes were harvested in passive lysis buffer supplied in the assay kit, and luciferase activity was measured by a plate reader.Protein concentration of the lysate was quantified by BCA assay (Pierce, Thermo Fisher Scientific).Luminescence values were normalized to total protein.

Figure 3 .
Figure 3. Purine nucleotide degradation is increased in atrophying myotubes but culminates in the release of hypoxanthine and xanthine due to their lack of XOR expression.C2C12 myotubes were treated with dexamethasone (DEX) and/or serum starvation (S.S.) treatments for 48 hours.(A) Representative

Figure 6 .
Figure 6.Serum uric acid is greater in individuals with below-average physical fitness performance.Serum was collected from adult women with either low or average physical performance.(A) Serum uric acid levels.(B) Serum 3-MH.(C) Serum 1-MH.(D) Ratio of 3-MH to 1-MH.Box plots show the interquartile range (box), median (line), and minimum and maximum (whiskers).n = 20/group, unpaired 2-tailed t test.

Table 1 . Participant characteristics and performance measures
Mean ± SD.P values determined by unpaired 2-tailed t test.App mass, appendicular lean mass.