Physiological Adaptations to Progressive Endurance Exercise Training in Adult and Aged Rats: Insights from the Molecular Transducers of Physical Activity Consortium (MoTrPAC)

Abstract While regular physical activity is a cornerstone of health, wellness, and vitality, the impact of endurance exercise training on molecular signaling within and across tissues remains to be delineated. The Molecular Transducers of Physical Activity Consortium (MoTrPAC) was established to characterize molecular networks underlying the adaptive response to exercise. Here, we describe the endurance exercise training studies undertaken by the Preclinical Animal Sites Studies component of MoTrPAC, in which we sought to develop and implement a standardized endurance exercise protocol in a large cohort of rats. To this end, Adult (6-mo) and Aged (18-mo) female (n = 151) and male (n = 143) Fischer 344 rats were subjected to progressive treadmill training (5 d/wk, ∼70%–75% VO2max) for 1, 2, 4, or 8 wk; sedentary rats were studied as the control group. A total of 18 solid tissues, as well as blood, plasma, and feces, were collected to establish a publicly accessible biorepository and for extensive omics-based analyses by MoTrPAC. Treadmill training was highly effective, with robust improvements in skeletal muscle citrate synthase activity in as little as 1–2 wk and improvements in maximum run speed and maximal oxygen uptake by 4–8 wk. For body mass and composition, notable age- and sex-dependent responses were observed. This work in mature, treadmill-trained rats represents the most comprehensive and publicly accessible tissue biorepository, to date, and provides an unprecedented resource for studying temporal-, sex-, and age-specific responses to endurance exercise training in a preclinical rat model.


Introduction
2][3][4] Among its many benefits, regular exercise helps maintain independence later in life, 5 reduces morbidity risk for over 26 chr onic lifestyle-r elated diseases, [6][7][8] and decr eases all-cause mortality. 9 , 10While the beneficial effects of exercise are believed to extend across organ systems, only a few tissues, usually skeletal muscle and heart, have been studied in detail. 3 , 11-15Thus, r emarka b l y, the collecti v e impact of exer cise tr aining on molecular signaling across a broad range of tissues and, by extension, how re gular exer cise promotes health and reduces disease risk, is not well-defined. 16 , 17To address these gaps, through support from the National Institutes of Health Common Fund, the Molecular Transducers of Physical Activity Consortium (MoTr-PAC) w as esta b lished to dev elop an inte gr ated molecular map of the adapti v e r esponse to exer cise tr aining across the lifespan.The primary goal is to provide a pub licl y av aila b le tissue biobank and multiomics data resource to support hypothesisdri v en r esear c h. 18 To better define the impact of exercise throughout the body, the Preclinical Animal Studies Sites (PASS) were established as one of the two exercise testing arms of MoTrPAC 18 to complement its clinical study sites.Specifically, the objectives of the PASS were to (1) develop a standardized exercise protocol for the c har acterization of physiological adaptation to exercise and (2) collect an expansi v e gr oup of tissues/organs for the cr eation of a pub licl y accessib le tissue bior e positor y and m ultiomic anal ysis database.To meet these objecti v es, the F isc her 344 (F344) rat was chosen as the model organism.The rat has long been utilized to study the impact of endurance exercise training on biology and health. 19By way of its size, the rat also provides the capability to study a broad range of tissues, which have sufficient mass to allow molecular phenotyping on multiple platforms thus maximizing quality control and inte gr ation capabilities.F inall y, gi v en the genetic, physiological, and metabolic similarities between rats and humans, 20 rats are a useful model of human phenotypic responses.To this point, rats have skeletal muscle fiber type distributions and glycogen utilization patterns more similar to humans than mice. 21 , 22The F344 rat strain is translationally relevant as it displays a pr ocli vity tow ard insulin r esistance and ectopic lipid deposition that increases with age, [23][24][25] mimicking common aging phenotypes in humans impacted by endurance training.
Here , w e describe the study design, physiological adaptations, and tissue acquisition after 1, 2, 4, or 8 wk of endurance exer cise treadmill tr aining at ∼70%-75% VO 2 max in a large cohort ( n = 294) of male and female F344 rats that were 6 or 18 mo of age at the initiation of the study.Results from this study are designed to be used as a readily accessible database and bior e positor y r esource for the scientific comm unity to couple with current [26][27][28][29] and future molecular profiling, thereby facilitating development of an inte gr ative map of systemic adaptations to endurance training.Demonstrating the utility of this resource to the resear c h community, we have recently undertaken m ultiomic anal yses on 18 differ ent tissues and the b lood from a subset of the 6 month-old cohort of rats, 28 including generating detailed insight into the molecular response in white adipose tissue, 26 and the mitochondrial 29 and nuclear transcription factor 27 response across tissues.

Animals
Male and female F isc her 344 (F344) inbred rats were obtained from the National Institute on Aging (NIA) rodent colony in cohorts of 20-30 rats.A total of 160 adult (3-5 mo of age) and 160 middle-aged (15-17 mo of age) rats were received at the animal test site (Uni v ersity of Iow a).Ther e wer e fiv e experimental groups for each age: sedentary control (SED) or 1, 2, 4, or 8 wk of treadmill training (1, 2, 4, and 8 W).To account for a potential effect of aging on outcome variables, an additional SED group was matched to the 1 W 18 mo group.The overall experimental design is outlined in Figure 1 (A).The experiment was designed so that r ats be gan exer cise tr aining at either 6 or 18 mo of age and were all housed at the test site for similar amounts of time ( ∼12 wk).We examined the response to training in these age groups because they represent adult rats with mature sexual and musculoskeletal organ systems and late middle-aged rats that have a low incidence of cancer and sarcopenia.Consequently, to meet the experimental design r equir ements, r ats w er e deli vered to the test site at differ ent a ges ( Figur e 1 A; Ta b le S1 ).Upon arri v al, rats were placed into a reverse dark-light cycle housing with lights off at 9:00 am and lights on at 9:00 pm for a minimum of 10 d prior to familiarization to the treadmill.This allowed for training of the rats during their normal acti v e period (dark phase), while also allowing for training to occur during normal working hours.This period of time was designed to provide sufficient time for the intrinsic circadian clocks across all tissues to entrain with the new light cycle.During this time, the rats were handled daily by the resear c h staff to minimize stress.Rats of the same sex were housed two per cage (146.4 in 2 of floor space) in ventilated r ac ks (Thoren Maxi-Miser IVC Caging System) with Tekland 7093 Shredded Aspen bedding.Rats were fed a standardized pellet diet (Lab Diet 5L79) consisting of 64% carbohydrates, 21% protein, and 15% fat and given ad libitum access to food and water.Both the bedding and diet used are standard for this NIA rodent colony.Daily cage activity and food consumption were not measured.The animal housing room was monitored daily and maintained at 68 • F-77 • F and 25%-55% humidity.Red lights were used during the dark cycle to provide adequate lighting for staff to perform routine housing tasks, rodent handling, and exercise training; no standard lighting was used during the dark phase.All animal pr ocedur es wer e appr ov ed by the Institutional Animal Care and Use Committee at the Uni v ersity of Iowa.

Treadmill Familiarization and Training
Treadmill exercise was performed on a Panlab 5-lane rat treadmill (Harvard Instruments, Model LE8710RTS).All animal handling and exercise was performed during the acti v e phase (dark cycle) for nocturnal rodents.Following the initial acclimation period, rats went through a 12-d treadmill familiarization protocol (outlined in Table 1 ) to expose them to the treadmill and to identify noncompliant rats.Those rats that successfully completed the 12-d familiarization protocol and were judged to be compliant (score of 2-4) were entered into the MoTrPAC database and randomized into an experimental group.Noncompliant rats (score of 1) were removed from the study.The number of rats received, randomized into an experimental group, and completed the exer cise tr aining ar e pr ovided in Ta b le S1 .
Exer cise tr aining be gan at 6 or 18 mo of age and lasted for a duration of 1, 2, 4, or 8 wk.Rats were exercised on a motorized treadmill 5 d/wk using a progressive training protocol designed to elicit an intensity of ∼70%-75% of VO 2 max, 30 The starting treadmill speed was based on VO 2 max measurements obtained following familiarization and 7-8 d prior to training in the compliant rats.Training was performed under red lights during the dark cycle and started no earlier than 10:00 am and no later than 5:00 pm over 5 consecutive days per week followed by 2 d of rest.Training was initiated with the treadmill set at a grade of 5 • and a duration of 20 min.As illustrated in Figure 1 (A), the duration of exercise was increased by 1 min each day until day 31 of training (start of week 7) when a final duration of 50 min was reached.The treadmill grade was increased from 5 • to 10 • at the start of week 3 and stayed at 10 • for the remainder of the training.The starting treadmill speed varied as a function of sex and age and increased at the start of weeks 2 and 4-7.At the start of week 7, speed, gr ade , and dur ation w ere fixed and maintained for the

Body Composition
The minispec LF90II Body Composition Rat and Mice Analyzer (Bruker; 6.2 MHz Time-Domain Nuclear Magnetic Resonance [TD-NMR] system) was used for in vivo measurement of body fat, lean tissue (ie, fat-free mass, which includes skeletal muscle), and fluid in conscious animals.Pretraining body composition was determined 13 d prior to the start of training in all T able 1. T r eadmill Familiarization Pr otocol

1-2
Rat was placed on the treadmill at a speed of 0 m/min for 10 min to familiarize it to the treadmill.The shock grid was blocked to pr ev ent the rat from sitting on the grid.

3-5
Rat was placed on the treadmill with the shock grid blocked and ran at a speed of 6 m/min for 10 min.A pen with a dull point was used to gently prod the rat or turn its head to make it walk forward.

6-12
Rat was placed on the treadmill with the shock grid blocked and run at 10 m/min for 10 min.If the rat was unable to run at 10 m/min, the speed was reduced to 6 m/min and then increased to 8 mm/min for 5 min once the rat was able to run forward pr operl y.On the next day the treadmill speed was set to 10 m/min.If the rat was unable to run at 10 m/min, the speed was reduced to 8 m/min.For those rats that were not able to run continuously at 10 m/min, a light shock was used to entice them to run.Rats were not allowed to sit on the shock grid.For those rats that liked to walk backwards, a pen was used to turn their head and prod them to run forward .11  After the 10 min familiarization session, the rats were run for 2 min at a treadmill grade of 10 • and a speed of 12 m/min.12  Eac h r at w as run on the tr eadmill at 0 • incline and 10 m/min for 5 min after whic h the gr ade w as incr eased to 10 • and speed to 12 m/min for 5 min.Upon completion of the run, eac h r at was assigned a score ranging from 1 to 4, with 4 being the highest score.
Scoring criteria: 4: rat is acti v e on the treadmill the entire activity session without assistance.
3: rat r equir ed minimal assistance, defined as assistance for less than 25% of the time of the activity session.
2: rat r equir ed m uch assistance, defined as assistance for greater than 25% of the time of the activity session.1: rat was noncompliant and failed to complete an activity session.Those rats that were unable to run on the treadmill for 5 min at a speed of 10 m/min and grade of 0

Maximum Oxygen Consumption (VO 2 max)
VO 2 max testing was performed 7-8 d prior to the onset of training in all rats and during the last week of training for the 4-and 8wk exercise groups.Maximum running speed (MRS), which was defined as the highest recorded speed, was recorded in all rats during the VO 2 max test.After completing the treadmill familiarization period, rats were acclimated to a single-lane enclosed treadmill (Columbus Instruments Metabolic Modular Treadmill) 2 d prior to testing.On the first day of acclimation, eac h r at was placed in the enclosed treadmill for 10 min at 0 m/min to acclimate them to the environment.On the second day of acclimation, the rat was placed in the enclosed treadmill for 10 min and ran at a speed of 10 m/min.On the test day, the rat was placed in the treadmill, and testing began once oxygen consumption had stabilized.Testing began with a warm-up for 15 min with the treadmill set at a speed of 9 m/min and 0 • incline.Following the warm-up period, the incline was increased to 10 • and treadmill speed was increased by 1.8 m/min ev er y 2 min 30 ; the protocol used differed by sex and training duration, with each protocol overviewed in Figure S1(A) -(F) .During the test, shock was used sparingl y, and onl y when the rat stopped running and sat on the shock area.Testing stopped when the rat sat on the shock ar ea thr ee consecuti v e times and did not r espond to incr eased shock.Upon cessation of the test, the rat was removed from the enclosure and blood drawn from the tail vein to measure lactate.Criteria for reaching VO 2 max during this graded treadmill test was a plateau in oxygen uptake despite increased workload, a r espirator y exchange ratio ≥ 1.05, and/or a nonhemolyzed blood lactate concentration ≥ 6 mmol/L (Lactate Plus). 30In Adult rats, VO 2 max and MRS was calculated at baseline for SED and

Tissue Collection
Tissues were collected from all rats 48 h following the last training session.The duration between the last training session and tissue collection was chosen to focus on the cumulati v e effects of steady state treadmill training and to limit potentially confounding effects of the last acute exercise bout.On the day of collection, food w as r emov ed at 8:30 am , 3 h prior to the start of dissections, which occurred between 11:30 am and 2:30 pm (in the dark cycle) ( Figure 1 B).For each experimental cohort, dissections occurred during this 3 h window in 5-6 animals per day and over a period of 3-5 d.This design was chosen to limit potential effects of time-of-day and circadian oscillations.
An overview of the workflow for tissue dissection is provided in Figure 2 (A).Specifically, body weight was measured and then r ats w ere placed in an induction box and sedated with inhaled isoflur ane (3%-4%); r ats w ere maintained in the dark until they were anesthetized, after which they were exposed to standard lighting.Once sedated, the rat w as mov ed to a nose cone and contin uousl y sedated with isoflurane (1%-2%).Blood was drawn via cardiac puncture followed by dissection of the right soleus (SOL), gastrocnemius, and plantaris (PL) muscles, right lateral subcutaneous (inguinal) white adipose tissue, right lobe of the li v er, v ena cav a, and finall y the heart and lungs.Removal of the heart was recorded as time of death.Immediately following r emov al of the heart, a guillotine w as used for decapitation.The brain w as r emov ed fr om the skull and the hypothalamus, right and left hippocampus, and right and left cerebral cortex were collected, in order.Following decapitation, specific organs wer e r emov ed fr om the bod y in the follo wing or der: right kidney, right and left adrenal, spleen, brown adipose (between shoulder blades), small intestine (jejunum), colon (transverse  The av era ge time between tissue r emov al and fr eezing or death (heart r emov al) and fr eezing ar e pr ovided for each tissue in Figure 2 (B)-(E).All tissues wer e subsequentl y shipped in dry ice to the bior e positor y at the Uni v ersity of Vermont for long-term storage and distribution.

Muscle Fiber-Type Distributions and Fiber Size
F iber-type per centages (based on m y osin heavy c hain expression) and fiber-type specific cr oss-sectional ar ea (CSA) wer e determined in the SOL, PL, LG, and MG muscles for the SED and 8 wk training groups of both sexes and ages (48 animals per muscle).The SOL is a pr edominantl y slow m uscle in the rat with > 85% type I fibers, while the PL, MG, and LG are muscles that express all four fiber types with a predominance of fast (type II) fibers. 31 , 32Specifically, a portion of each fr ozen m uscle was cut from the mid-belly, mounted on cork in embedding medium (OCT), and frozen.Care was taken during the blocking of the tissue to ensure that the muscle remained frozen.

CD31/PECAM 1-Capillary Contacts
Capillary contacts, the number of capillaries surrounding a single fiber, were determined in the SOL, PL, LG, and MG muscles for the SED and 8 wk training groups of both sexes and ages (48 animals per muscle).Serial cross-sections were cut and processed as described a bov e .For eac h muscle , all slides w er e pr ocessed on the same day.Following the initial b locking ste p, samples were incubated with CD31/PECAM1 (R&D System AF3628) and dystr ophin (H5) primar y antibodies conjugated with Alexa Fluor 488 (Santa Cruz sc-365954, 1:80), diluted in BS, and incubated at 4 • C overnight.On the following day, samples were washed 3 times for 5 min with PBST and then incubated for 50 min in BS at RT with the secondary antibody (555 IgG donkey antigoat (A21432, 1:500).Samples were then washed 3 times for 5 min in PBST and cov er ed with Pr oLong TM Diamond Antifade Mountant (P36930; Thermo Fisher Scientific).
For eac h muscle , the entir e cr oss-section w as digitall y scanned at 10X objecti v e using a Zeiss LSM710 confocal microscope using the Tile Scan tool.Images were collected within 5 d of staining.Images were quantified using ImageJ analysis softw ar e v ersion 1.53r (National Institutes of Health, USA). 34Muscle fibers were analyzed from 2 to 3 regions of the cross-section: 3 regions from the LG, and 2 regions from the MG, PL, and SOL (scaling).Regions (0.83 μm × 0.83 μm) were selected to capture a di v erse set of fiber types within each muscle.The number of capillaries surrounding each fiber (capillary contacts) was counted man uall y by a single individual who was blinded to the age, sex, and group of the sample.The mean ± SD number of fibers sampled per muscle was 367 ± 72 for the SOL, 390 ± 67 for the PL, 815 ± 89 for the MG and 1110 ± 184 for the LG.

Citr a te Synthase
Citrate Synthase (CS) activity was assayed using a modified pr otocol fr om Sr er e et al. 35 The assay buffer (200 μL final volume) contained monobasic and dibasic potassium phosphate buffers ( 36

Glycogen
Gl ycogen w as assessed using Glycogen Assay kit (Sigma, MAK016).Briefly, muscles were homogenized in 100 μL of water, boiled for 5 min, and centrifuged at 13 000 g for 5 min to r emov e debris.A volumef 10 μL of the supernatant were used in the assay following the kit protocol.An endpoint absorbance was measured at 570 nm.Results were analyzed by doing a backgr ound corr ection and normalized to milligr am of tissue .

Plasma Clinical Analytes
Using all Adult samples and only those Aged samples that were selected for multiomic analysis, [26][27][28][29] a set of nine common clinical anal ytes w as measur ed in plasma: glucose, lactate, gl ycer ol, total ketones, nonesterified fatty acids (NEFA), glucagon, insulin, leptin, and corticosterone.The first five were measured using a Beckman DxC 600 clinical analyzer with reagents from Beckman (Brea, CA) and Fujifilm Wako (Osaka, Japan; total ketones and NEFA), while the others were measured in immunoassays using commercial kits from Meso Scale Discov er y (Roc kville , MD) and Alpco (Salem, NH; corticosterone).Catalog numbers are provided in Ta b le S2 .

Body Composition and Maximal Oxygen Consumption (Post-Pre) Training Differences
For measures that were recorded both pre-and post-trainingbod y mass recor ded on the same day as the NMR body composition measures, NMR lean mass, NMR fat mass, and absolute and r elati v e VO 2 max-we fit ordinar y or weighted least squar es (WLS) r egr ession models with age, sex, group, and their interactions as predictors of the (post-pr e) differ ences.Recipr ocal gr oup variances (calculated from each combination of age, sex, and gr oup) wer e used as weights in the WLS models to account for observ ed heter oscedasticity.A few influential observ ations wer e r emov ed and noted in the results.Then, two-sided t -tests were performed to determine whether the mean of the (post-pre) differ ences fr om each gr oup w as differ ent fr om 0. That is, if ther e was a change from pre-to post-training measures.Since maximum run speed was recorded in 1.8 m/min intervals and could only take on a few distinct values, we instead performed nonparametric Mann-Whitney U tests 40 se paratel y for each combination of age, sex, and group.For all measures, P -values were adjusted across groups within each age and sex combination using the Holm pr ocedur e 41 to contr ol the famil y-wise err or r ate .Results of these analyses are provided in Table S3 .

Baseline (Pretraining) Differences:
For all measures except maximum run speed, we fit log-link Gaussian, quasi-Poisson (% fat mass), or gamma (absolute and r elati v e VO 2 max) generalized linear models 42 (GLMs) with age, sex, group, and their interactions as predictors of the baseline (pr etraining) v alues.GLMs can addr ess nonconstant v ariance observed in strictly positive data, like VO 2 max.If data was not recorded for some age and group combinations, as with 4 W VO 2 max in the Aged animals, we instead concatenated age and group to form a single factor ( age group ) and fit a model with predictors age group , sex, and their interaction to avoid inestimable r egr ession coefficients.For each Gaussian GLM, r ecipr ocal gr oup variances (calculated from the untransformed response values within each combination of age, sex, and gr oup) wer e included as weights to account for any residual heteroscedasticity.Model parsimony was achieved through ANOVA F -tests and examination of r egr ession dia gnostic plots.Then, eac h of the tr ained time points wer e compar ed a gainst their a ge-and sex-matched SED controls using the Dunnett multiple comparison procedure . 43If age gr oup was included as a predictor, comparisons wer e man uall y specified, and P -v alues wer e instead adjusted within each age and sex combination using the Holm procedure. 41Since the log link was used for all models, results are presented as ratios of trained to SED group means (fold-change).
Since maximum run speed was recorded in 1.8 m/min intervals and could only take on a few distinct values, we instead performed nonparametric Mann-Whitney U tests 40 to compare eac h tr ained group to their matc hing contr ol gr oup.P -v alues were adjusted across comparisons within each sex and age combination using the Holm pr ocedur e. 41 Results of these analyses ar e pr ovided in Ta b le S4 ("NMR & VO 2 max" ta b).

Weekly Body Mass
Weekly body mass was recorded prior to the beginning of each week from weeks 1 to 8. We filtered the data to only those observations collected from the SED and 8 W groups, since this allowed for the most weekly comparisons.Since there are longitudinal measur es fr om eac h r at, w e used the nlme:: gls 38 R function to fit a generalized least squares (GLS) model 44 with age, sex, group, w eek (cate gorical: 1-8), and their interactions as predictors of log(body mass).Since the correlations between measur ements fr om the same rat decr ease as the time la g (n umber of w eeks betw een the measurements) increases, the correlation structur e w as specified with nlme: corAR1(form = ∼ 1 | pid) , where pid is a unique identifier for each rat.Model parsimony w as achiev ed via likelihood-r atio tests.Then, w e tested whether the mean of the 8 W group was different from the mean of the SED group at each week.P -v alues wer e adjusted acr oss weeks 1-8 within each combination of age, sex, and group (SED or 8 W) using the Holm pr ocedur e. 41 Results of this anal ysis ar e pr ovided in Ta b le S4 ("Weekl y Body Mass" ta b).

Fiber-Type-Specific Measures
Cr oss-sectional ar ea : Since the SOL only consists of two of the four fiber types (types I and IIa), we first created a new categorical v aria b le called m uscle type by concatenating m uscle and fiber type to av oid inestima b le r egr ession coefficients.Since ther e ar e r e peated measur es fr om each animal, we fit a linear mixed-effects model (LMM) with age, sex, group (SED or 8 W), muscle type , and their interactions as predictors of the logtransformed fiber type-specific CSA with a random intercept for each rat.Precision weights were specified with nlme:: varIdent(form = ∼ 1 | muscle type) to account for heteroscedasticity.Then, we tested whether the mean of the 8-wk-trained group w as differ ent fr om that of the SED contr ol gr oup for each combination of age, sex, group, and muscle type .P -values were Holmadjusted 41 across all 2 (SOL only) or 4 fiber types within each age, sex, and muscle combination.Since the response was logtransformed, results are presented as ratios of 8 W to SED group means (fold-changes) in Ta b le S4 ("Mean Fiber Area" tab).
Fiber-type distribution: We performed compositional data analysis (CoDA) [45][46][47] of the fiber counts, which we believe is more appropriate than common statistical methods for assessing fiber -type distribution.CoD A is appr opriate for positi v e data carr ying r elati v e , r ather than absolute , information, and it is used extensi v el y in the geosciences (eg, analysis of mineral compositions). 48Additionally, since each set of fiber-type proportions ar e deri v ed fr om the same animal and m ust necessaril y sum to 1, a change in one fiber-type proportion would affect the remaining proportions.This violates the independence of observations assumption of classic anal ysis appr oaches like ANOVA, necessitating a different approach-CoDA.
Data pr e par ation be gan b y con v erting the fiber counts fr om eac h r at to proportions using the total number of fibers per muscle.Then, we applied the isometric log-ratio (ilr) transformation, 49  to values along the real number line R avoids singularity of the v ariance-cov ariance matrix, which would pr esent pr ob lems for the statistical analyses that we will describe, though we necessarily sacrifice some interpretability of the results.The matrix of column vectors (the balances) follow a m ulti v ariate Normal distribution, so, for each muscle, we fit a m ulti v ariate m ultiple r egr ession model with categorical v aria b les a ge, sex, gr oup, and their interactions as predictors of the 1 (SOL) or 3 (LG, MG, and PL) dependent variables.
For all muscles, we utilized t -tests to compare the mean of each balance from the 8 W group to the corresponding mean fr om the SED gr oup (eg, b1 8 W − b1 SE D ).The r esulting P -v alues wer e Holm-adjusted 41 acr oss the m uscles within each combination of age and sex.The differences between balances are not easil y interpr eta b le, so r esults ar e pr esented as a shift between specific fiber types ( Muscle Fiber Type Distribution Results).Results of these analyses are provided in Table S4 ("Fiber Count" tab).

Muscle Morphology and Biochemistry
For each muscle-specific measure-capillary contacts, CS activity, glycogen, mean CSA, and terminal mass-we examined their mean-v ariance r elationship.Informed by these r elationships, we fit LMMs with log-transformed or squar e-r oot-transformed (gl ycogen onl y) de pendent v aria b les.Varia b les a ge, sex, gr oup, muscle, and their interactions were included as predictors with a r andom inter cept for eac h r at.LMMs w er e utilized because ther e ar e r e peated measur es for eac h r at, whic h violates the independence assumption of ordinary linear r egr ession, while an appropriate v ariance-sta bilizing transformation w as applied to each r esponse to addr ess heter oscedasticity.If heter ogeneity of the r esiduals w as still observ ed, w eights w ere included with nlme:: varIdent .Finally, model parsimony was achieved through ANOVA F -tests and examination of r egr ession dia gnostic plots.Next, we compared each of the trained timepoints against their age-and sex-matched SED controls using the Dunnett multiple comparison pr ocedur e 43 within each m uscle.Since gl ycogen did not use a log-tr ansform, w e instead estimated the marginal means on the squar e-r oot scale , bac k-tr ansformed to the original glycogen concentration scale while adjusting for bias, and then logtransformed these values before setting up the contrasts.In doing so, we are able to present results as ratios of trained to SED group means (fold-changes), as with the other muscle measures.Results of these analyses are provided in Table S4 ("Muscle Measur es" ta b).

Clinical Analytes
For each of the plasma clinical analytes, we first examined their mean-v ariance r elationship and fit an appr opriate log-link GLM assuming the data followed a Gaussian (corticosterone, glucose, insulin, lactate, and leptin), quasi-Poisson (glucagon), gamma (gl ycer ol and total ketones), or inverse Gaussian (NEFA) distribution.GLMs can address the issue of nonconstant variance typicall y observ ed in strictl y positi v e data.Recipr ocal gr oup v ariances (calculated from each combination of age, sex, and group) were included as weights in the Gaussian GLMs to account for r emaining heter oscedasticity.Age, sex, gr oup, and their interactions were included as predictors, and model parsimony was achiev ed thr ough ANOVA F -tests and examination of r egr ession diagnostic plots.Then, we compared each of the trained timepoints against their age-and sex-matched SED controls using the Dunnett multiple comparison procedure. 43Since the log link was used for all models, results are presented as ratios of trained to SED group means (fold-changes) in Table S4 ("Plasma Anal ytes" ta b).

Calculation of Percent Change
We do not perform r egr ession anal yses on the post/pre values for each rat, the results of which could easily be converted to % change with (post/pre-1) × 100, because the use of a ratio in r egr ession anal yses can lead to incorr ect or misleading inferences. 51Instead, % change from pre-to post-training was calculated by dividing the differences in means ("(Post-Pre) Training Differences" Methods) by the corresponding pretraining means ("Baseline (Pretraining) Differences" Methods) and multiplying by 100%.For the trained vs. SED comparisons, % change from SED to trained was calculated by subtracting 1 from the ratios and m ultipl ying by 100.

Results
Baseline Phenotypes Across Cohorts: Given the scale of this study, rats arri v ed at the facility in different shipments.For insight into cohort matching, we first confirmed that the rats within each a ge, gr oup, and sex were well-matched by comparing baseline (ie, pretraining) phenotypic parameters for VO 2 max and/or MRS, and body composition between the four training groups (1, 2, 4, and 8 W) and the SED controls.
Baseline VO 2 max and MRS: In Adult female rats, absolute VO 2 max was the same between all groups ( Figure S2A ).Relative VO 2 max was not different between the SED and trained groups, with the exception of an 8% lower r elati v e VO 2 max in 1 W relati v e to SED Adult females ( Figure S2B ).Further, baseline MRS was similar across all groups ( Figure S2C ).In Adult male rats, absolute VO 2 max was 8% higher in 1 W compared to SED rats, with no significant differences in the other trained groups.The mean relati v e VO 2 max w as 6% higher in 2 W r elati v e to SED, but the other gr oups wer e not significantl y differ ent fr om SED ( Figur e S2D -E ).
Relati v e to SED, Adult male rat MRS was modestly higher in the 1 W and 2 W groups ( Figure S2F ).In Aged rats, VO 2 max was measured in the SED and 8-wk-trained groups, only (see "Methods").In Aged female rats, the mean r elati v e pr etraining VO 2 max was 7% higher in the 8 W group relative to SED, though neither absolute VO 2 max nor MRS were different between the SED and 8 W groups at baseline ( Figure S2G -I ).In Aged male rats, no differ ences wer e observ ed in a bsolute or r elati v e VO 2 max or MRS ( Figure S2J -L ).Together these findings suggest the cohorts of rats w ere w ell-matc hed.
Body Mass and Composition : Overall, body mass and body composition were well-matched across all cohorts upon arrival and prior to the beginning of training ( Figure S3 ).In female and male Adult rats ( Figure S3A -J ), the greatest differences were found between the 1 and 2 W groups (which arri v ed as a single cohort) and the SED group, with total body mass in males and females being significantly higher in the 1 and 2 W groups r elati v e to SED (6% and 5%, r especti v el y) ( Figur e S3A ).In Adult females, the greater body mass was due to increases in both whole-body fat and lean mass ( Figure S3B -E ).In Adult males, the greater body mass was largely due to an increase in fat mass ( Figure S3F -J ) at the beginning of the study.In Aged females and males ( Figure S3K -T ), body mass was modestly, but significantly, lower in female and male 1 W ( −8% and 4%, respecti v el y) and 2 W (both, −4%), which arri v ed as a single cohort, and 4 W ( −6% and 4%, r especti v el y), which arri v ed as a separate cohort, as compared to SED.The lo wer bod y mass in 1, 2, and 4 W Aged females was accompanied by a significantly lower lean mass ( Figure S3M ) and % lean mass ( Figure S3O ); there were no differences in fat mass (except for 1 W [ −19%]; Figure S3L ) or % fat mass (except for 4 W [ −9%]; Figure S3N ) in the trained groups, as compared to SED.Despite the changes in body mass, in Aged males, there were modest statistical differences in body composition between the five trained groups and SED; fat mass ( Figure S3Q) and % fat mass ( Figure S3S ) were lower in 4 W ( −7% and −5%, r especti v el y), while % lean mass ( Figure S3T ) was higher in 1 W ( + 2%) and 2 W ( + 3%), as compared to SED.
Overall, these data show that while rats of both sexes arri v ed in separate cohorts, at different ages and across a period of 6 mo within eac h age , the major phenotypic physiological par ameters wer e v er y w ell-matc hed acr oss all gr oups prior to beginning the treadmill training intervention.

Pr ogr essive Endurance Exercise Training Protocol
To estimate running intensity during training, which was designed to target an exercise intensity of ∼70-75% VO 2 max, blood lactate concentration was measured from the tail vein at the completion of the first training bout eac h w eek.The week 4 values for the 4 W training group animals are the endof-w eek postexer cise blood lactate (da y 20), since the V O 2 max testing was performed at the start of that week.In line with a targeted intensity of 70%-75% VO 2 max, the mean ± SD of the blood lactate concentration was 4.

Effects of Training on VO 2 max and MRS
Tr eadmill testing w as performed in pr e-and post-training in the SED, 4 W, and 8 W groups of Adult and Aged rats to assess the adaptation to the pr ogr essi v e tr eadmill training.
Adult group.In the SED groups of female and male rats, absolute VO 2 max did not change over the 8-wk training period, though r elati v e VO 2 max decr eased by a mean of 6.9 ( −9%) and 2.7 ( −6%) mL/kg/min in females ( Figure 3 A and B) and males ( Figure 3 D and E), respectively.MRS also decreased significantly in the SED females and males ( Figure 3 C and F, respecti v el y).In females in response to training, absolute VO 2 max was higher only in 8 W ( + 20%; ( Figure 3 A), while r elati v e VO 2 max increased in both 4 W (1.8 mL/kg/min, + 2%) and 8 W groups (10.2 mL/kg/min, + 14%) ( Figure 3 B).MRS also increased in 4 and 8 W females ( Figure 3 C).Male rats displayed similar impr ov ements in absolute ( Figure 3 D) and relative ( Figure 3 E) VO 2 max at 4 wk (absolute: + 4%, r elati v e: + 4%) and 8 wk (absolute: + 11%, r elati v e: + 17%), as well as a r obust impr ov ement in MRS ( Figure 3

F).
Aged group.In Aged rats, neither absolute nor r elati v e VO 2 max changed significantly in the SED groups of either females ( Figure 3 G and H) or males ( Figure 3 J and K).Following 8 wk of training, both absolute and relative VO 2 max increased in Aged female (absolute: + 15%, relative: + 18%; Figure 3 G and H, respectively) and male rats (absolute: + 6%, relati v e: + 18%; Figure 3 J and K, r especti v el y).MRS w as measur ed at both 4 and 8 wk of training and increased significantly in both females ( Figure 3 I) and males ( Figure 3

L).
The mean change in absolute (mL/min) and relative (mL/kg/min) VO 2 max as the result of training are plotted for all gr oups in Figur e 4 and demonstrates that increases in both absolute and r elati v e VO 2 max with training were similar across age and sex.Descripti v e statistics (mean, SD, minim um, maxim um, r ange , and coefficient of variation) for all groups for pre-and post-training are provided in Table S5 (absolute VO 2 max) and Ta b le S6 (r elati v e VO 2 max).

Changes in Body Mass and Composition With Training
In both Adult and Aged rats, pre-and post-training measures of body composition by NMR were taken in SED, 4 W, and 8 W rats.In addition, body mass was assessed at the beginning of each tr aining w eek in all cohorts.
Adult group.In female rats, total body mass increased significantly in SED and the 4 W and 8 W training groups ( Figure 5 A).In SED females, such changes were accompanied by an increase in lean ( + 5%) and fat ( + 39%) mass ( Figure 5 B and C, r especti v el y); accounting for changes in body mass, this translated to a decrease in % lean mass ( −5%) and an increase in % fat mass ( + 11.5%; Figure S5A and S5B , r especti v el y).With training, lean mass ( Figure 5 B) and % lean mass ( Figure S5B ) incr eased significantl y in 4 W ( + 6% and + 2%, r especti v el y) and 8 W females ( + 7% and + 2%, r especti v el y).No training-induced changes in fat mass were observed in females ( Figure 5 C; Figure S5A ).In SED males, body mass ( Figure 5 D), lean mass ( Figure 5 E), and fat mass ( Figur e 5 F) wer e incr eased ( + 7%, + 6%, and + 17%, r especti v el y).Accounting for changes in body mass, % fat mass increased by 9% ( Figure S5D ), though there were no changes in percentage lean mass ( Figure S5E ).In trained males, % lean mass increased by 5% at both 4 W and 8 W time points ( Figur e S5E ), though ther e w as no statisticall y significant change in absolute lean mass ( Figure 5 E).Overall, males decreased total fat mass ( Figure 5 D) and % fat mass ( Figure S5D ) at 4 W (total: −18%; %: −16%) and 8 W (total: −38%; %: −36%), whilst total body mass onl y decr eased in 8 W ( Figure 5 D).Though we did not assess changes in body composition in the 1 W and 2 W groups, we did assess changes in body mass.In both 1 and 2 W females ( Figure S5C ) and males ( Figure S5F ), the terminal body mass was 1%-4% lower than pretraining body mass (which was measured on the NMR day).
The absolute changes in body mass, fat mass, and lean mass as the result of training are plotted for all groups in Figure 6 .The data highlight differential responses of Adult males and females to exer cise tr aining, with males losing total body and fat mass; in contr ast, tr ained females maintained a constant total body and fat mass and did not gain body mass like the SED group.In Aged rats, in contrast to Adults, trained males and females both display decreases in total body and fat mass.Descripti v e statistics (mean, SD, minim um, maxim um, r ange , and coefficient of variation) for all groups for pre-and post-training are provided in Ta b le S7 (body mass), Ta b le S8 (lean mass), and Ta b le S9 (fat mass).

Weekly Monitoring of Body Mass
Adult group.In Adult females, body mass incr eased ov er time in the SED group and remained fairly constant in the 8 W gr oup ( Figur e S6A ).Conv ersel y, the body mass of SED males  remained constant and decreased over time in the 8 W group ( Figur e S6B ).Compar ed to their a ge-matched SED counterparts at eac h w eek, the mean body mass of 8 W females was ∼4% lower starting at week 5 and remained ∼4% lower through week 8 ( Figure S6E ).In males, mean body mass of the 8 W animals was 5% lower than their age-matched SED counterparts starting at week 4; body mass continued to decrease with training duration until the difference in body mass was ∼9% at week 8; ( Figure S6E ).
Aged group.In Aged females, body mass remained constant over 8-wk in the SED group ( Figure S6C ).In the 8 W group, body mass decr eased initiall y at the start of training befor e r eturning to week 1 values at the beginning of week 8 ( Figure S6C ).In SED males, there was an initial decrease in body mass followed by a plateau from weeks 3 to 8 ( Figure S6D ).In 8 W males, howev er, ther e w as an immediate and consistent decr ease in body mass that stabilized starting at week 7 ( Figure S6D ).Compared to their age-matched SED counterparts at each week, the mean body mass of 8 W males was lower at the beginning of week 4, with similar decreases in body mass in Aged and Adult males in response to training ( Figure S6E ).In Aged females, while not statistically significant, there was a 4% decrease at the start of week 4 that persisted through week 8, much like what was observed in the Adult females ( Figure S6E ).

Terminal Muscle Mass
Terminal masses for the four collected muscles are presented in Figure S7(A) -(P) .
Adult group.In females, PL ( Figure S7C ) and SOL ( Figure S7D ) masses wer e significantl y higher in the 1 W (SOL only: 8%), 2 and 4 W groups relative to the SED controls (PL: 4W: + 8%, 8W: + 6%; SOL: 2 W and 4W: + 9%); this may relate to the fact that baseline body mass and lean mass of the 1 and 2 W cohorts were higher than the SED group ( Figure S4A and S4C ; this could be because the 1 and 2 W r ats w er e fr om a differ ent cohort than the SED rats).In trained males, the LG ( Figure S7E ), PL ( Figure S7G ), and SOL ( Figure S7H ) were not different from SED at any timepoint.In the MG, muscle mass was lower ( −7%) than SED at 4 W, only ( Figure S7F ).
Aged group.In Aged females, the MG ( Figure S7J ), PL ( Figure S7K ), and SOL ( Figure S7L ) mass in the 8 W group was significantl y gr eater than SED, while ther e w as no differ ence in the LG ( Figure S7I ); the mean % difference, as compared to SED, was 7%, 8%, and 11% for the MG, PL, and SOL, r especti v el y.In Aged males, there were no significant differences in the terminal muscle masses of any trained groups, as compared to SED ( Figure S7M -P ).

Muscle Fiber Types and FIber Type Specific CSA
Fiber type distributions .In Adult females, ther e w as a shift from type IIb to type IIx fibers in the LG with 8 wk of training ( Figure 7 A), and an increase in type IIa fibers relative to the type IIb and IIx fibers in the PL ( Figure 7 C), with no training-related differences in fiber type composition in the MG ( Figure 7 B) or SOL ( Figure 7 D).In Adult males, there were no training-related differences in fiber type composition in the LG ( Figure 7 E) or SOL ( Figure 7 H).In contrast, the ratio of type I fibers r elati v e to type II fibers was lower in 8 W versus SED in the MG, with no change in the r elati v e pr oportions of type II fibers ( Figur e 7 F).Additionall y, ther e w as an incr ease of type IIa r elati v e to the other type II fibers in 8 W versus SED, in the PL only ( Figure 7 G).In Aged females, ther e w as an incr ease in type IIx r elati v e to type IIb fibers in the LG with 8 wk of training ( Figure 7 I), and an increase of type IIa r elati v e to other type II fibers in both the MG ( Figure 7 J) and PL ( Figure 7 K).There was no change in the SOL fiber type composition in Aged females ( Figure 7 L) or males ( Figure 7 P).In 8 W versus SED Aged males, there was a higher proportion of type

Mean Muscle Fiber CSA
In Adult rats, no differences were observed in the mean muscle fiber CSA of the 8 W and SED groups of any muscles of female ( Figure S8A ) or male rats ( Figure S8B ).In Aged female rats, no changes in mean muscle fiber CSA were observed between the 8 W and SED ( Figure S8C ).In Aged males, 8 W SOL mean muscle fiber CSA was higher ( + 23%), as compared to the SED group ( Figure S8D ).In contrast, there were no differences in mean muscle fiber CSA in LG, MG, or PL ( Figure S8D ).

Capillary Contacts Per Muscle Fiber
Treadmill training for 8 wk had no effect on the mean number of capillary contacts per fiber in any muscle in Adult (Females: Figure S9A ; males: Figure S9B ) or Aged rats (Females: Figure S9C ; males: Figure S9D ).

Muscle CS Activity
Adult group.In Adult female rats, CS activity increased in all muscles with progressive endur ance tr aining ( F igure 9 A-D) and w as most pr onounced at 4 W.At 8 W in Adult females, CS acti vity incr eased r elati v e to SED in the LG ( + 67%; Figure 9 A), PL ( + 58%; Figure 9 C), and SOL ( + 50%; Figure 9 D), but not the MG ( Figure 9 B).In females, the mean fold-increase in CS activity between SED rats and apex levels in the 4 W group was-LG: + 5.73, MG: + 4.55, PL: + 3.24, and SOL: + 5. 67.The impact of training on CS activity followed a similar temporal pattern in males, with significant increases in CS activity in the LG ( Figure 9 E), PL ( Figure 9 G), and SOL ( Figure 9 H) in all training groups, whilst in the MG the incr ease w as evident in 1, 2, and 4 W only ( Figure 9 F).In the LG, PL, and SOL, the avera ge incr ease in CS acti vity in the 8 W gr oup r elati v e to SED was + 43%, + 65%, and + 54%, r especti v el y ( Figur e 9 E and G-H).In males, av era ge fold-incr eases in CS acti vity between the SED and peak activity at 4 W was-LG: + 4.52, MG: + 3.72, PL: + 3.56, and SOL: + 3. 25.

Muscle Glycogen
Adult group.In female rats, muscle glycogen content was significantly higher in the LG ( Figure 10 A), MG ( Figure 10 B), and PL ( Figure 10 C) of 8 W trained gr oups r elati v e to SED (mean fold-changes of + 6.89, + 3.60, and + 3.63, r especti v el y).Gl ycogen content was not affected by training in the SOL ( Figure 10AD ), or at other timepoints in the other muscles ( Figure

Plasma Clinical Analytes
To understand metabolic changes induced by training, we profiled a set of plasma hormones (insulin, glucagon, corticosterone, and leptin) and metabolites (glucose, lactate, NEFA, glycerol, and total ketones) that are key indicators of metabolic homeostasis.
Adult group hormones .In females, plasma insulin was not impacted by training, though glucagon concentration was 46% lower in the 4 W group relative to SED ( Figure 11 A-B).In males, plasma insulin was elevated in the 1 and 4 W groups, with no significant differences between 8-wk-trained and SED rats ( Figure 11 E).Unlike females, glucagon was unaffected by training in males ( Figure 11 F).Corticosterone was elevated in the 1 and 2 W females, peaked at 4 W (90% increase from SED), and then returned to SED levels by 8 W ( Figure 11 C).A similar pattern in plasma corticoster one w as observ ed in males, with incr eases in earl y training time points that wer e not observ ed by 8 W ( Figure 11 G).With training, plasma leptin levels decreased at 1, 2, and 8 W in females ( −32%, −39%, and −34%, r especti v el y) ( Figure 11 D).In males, leptin decreased at 2, 4, and 8 W relative to SED ( −30%, −27%, and −58% r especti v el y), with the greatest reduction in 8 W males ( Figure 11 H).
Adult group plasma metabolites.In Adult females, several plasma metabolites displayed early training responses that were atten uated with pr olonged tr aining ( F igur e 12 A-E).Namel y, lactate and gl ycer ol decr eased in the 1 and 2 W groups relative to SED (lactate 1W: −17%, 2W: −15%; gl ycer ol 1 W and 2W: −29%), and NEFA decreased by 22% at 1 W. Glucose and total ketone bodies were the only metabolites to respond at 4 wk of exercise training in females, with glucose increasing by ∼12% and ketones decreasing by ∼30% before mostly returning to SED le vels b y 8 W. In Adult males ( Figure 12 F-J), similar to Adult females, plasma glucose lev els wer e not impacted by training, while plasma lactate decreased at 1 and 2 W of training ( −17% and −15%, r especti v el y).Inter estingl y, training had opposite impacts on gl ycer ol concentrations in males-causing an increase of 31% with 1 W and 34% with 2 W of training before returning to SED levels at the later timepoints ( Figure 12 D, I).Finally, total ketone bodies displayed a similar temporal pattern in both sexes, decreasing significantly in 8 W males ( −33%) ( Figure 12 J).
Aged group hormones .The impact of training on plasma hormones in Aged rats followed similar temporal patterns as that of Adult rats for all analytes with the exception of glucagon in Aged females, which did not change with training ( Figure 11 J).Plasma insulin was unaffected by training in females ( Figure 11 I) and increased at 1 and 4 W in males ( + 27% and + 29%, r especti v el y) ( Figure 11 M).As in Adult animals of both sexes, plasma corticoster one w as elev ated at 1, 2, and 4 W training time points ( + 83, + 53%, and + 90%, r especti v el y) and then returned to SED levels at 8 W ( Figure 11 K).Leptin appeared to decrease with training, though the r esponse w as atten uated in Aged females, with marginal decreases that were only significant in the 2 W ( −28%) and 4 W ( −33%) gr oups ( Figur e 11 L).In Aged males, the leptin training response also appeared attenuated relative to Adult males, with plot trajectories suggesting a decrease beginning at 2 wk, but was not significant until 4 W ( −40%) of training ( Figure 11 P); an additional 4 wk of training did not appear to lower leptin much further ( −44% from SED to 8 W).
Aged group plasma metabolites.In Aged females ( Figure 12 K-O), glucose and NEFA r esponded similarl y to training as they did in Adult females, where glucose increased at 4 W (17%) and NEFA decreased at 1 W ( −22%).In males, glucose increased by 9% at 1 W and peaked at 2 W (13%) ( Figur e 12 P) r elati v e to SED, whereas plasma NEFA did not significantly change with training.In comparison to Adult males and females, where lactate decreased at 1 and 2 W of training ( Figure 12 B and G), lactate levels in Aged rats of both sexes trended upwards, increasing ∼30% in 8 W relati v e to SED groups ( Figure 12 L and Q).Also, unlike their younger Adult counterparts, gl ycer ol did not appear to respond to training, though the plots suggest it may have begun to decrease at later timepoints ( Figure 12 N and S).Finally, while ketones displayed nonsignificant increases at 1 and 2 W in females, levels decr eased suddenl y ar ound 4 W and contin ued until becoming 53% lower at 8 W r elati v e to SED ( Figur e 12 O).In males, ketones did not significantly change with training, although the y did displa y a do wnwar d trend similar to Adult males ( Figure 12 T).

Discussion
While the health benefits of regular endurance exercise are widely known, 52 the separate and integrati v e effects of exercise training on molecular signaling across many tissues, and how this interrelates to health and disease risk, remains to be thoroughly defined.To address this knowledge gap, here we describe the endurance exercise training arm of the Preclinical Animal Study Sites (commonly referred to as P ASS1B) of MoTrP AC, the primary goals of whic h w ere to, (1) develop a standardized endur ance exer cise protocol for the c har acterization of physiological adaptation to exercise and, (2) collect an expansi v e gr oup of tissues/organs for the creation of a pub licl y accessib le tissue bior e positor y and m ulti-omic anal ysis data base.Specificall y, we examined ke y ph ysiological and metabolic adaptations after 1, 2, 4, or 8 wk of endurance exercise treadmill training at ∼70%-75% VO 2 max in a large cohort of male and female F344 rats.
Importantly, in relation to the goals of this work, the progr essi v e endurance training program resulted in a robust ( ∼20%) impr ov ement in cardior espirator y fitness r egardless of age or sex, with v aria b le impacts of age and sex on other phenotypic measur es.Mor eov er, extensi v e phenotypic data fr om 294 rats was collected and > 5600 total samples comprising 18 solid tissues and blood were collected and biobanked, making this the most expansi v e, pub licl y accessib le data r esource and tissue bior e positor y, to date, for studying temporal, multiomic, sexspecific, and age-specific responses to progressive endurance training.Indeed, the utility of this resource is exemplified by a recent landscape study by MoTrPAC, which investigated the multiomic response within and across tissues in a subset of the male and female Adult rats from PASS1B. 28Additionally, more focused studies lev era ged this m ultiomics data to study the tissue-wide mitochondrial 29 and transcription factor 27 response to training, or sex-and training-specific responses in subcutaneous white adipose tissue. 26 key strength of the PASS1B resource is the expansive cohort size; to our knowledge, it is the first study of such magnitude to document these progressive changes in male and female rats of two a ge gr oups.Up to now, most rodent and human molecular profiling studies have been limited to studying a single time point, sex, and/or a ge gr oup, which limits the broader application and interpretation of the findings.Nota b l y, studying the pr ogr essi v e r esponse to endurance training permits inte gr ation of physiological and -omic adaptations acr oss tissues, ther eby offering the opportunity to r ev eal nov el pathwa ys ke y in tissue remodeling.An additional unique aspect of PASS1B is the highl y contr olled experimental design, which incr eases the translata bility and r e pr oducibility of the work.For example, to aid r e pr oducibility training and tissue collection occurr ed ov er a r estricted time period and at the same time of day, which was during their normally active dark phase.To accomplish tissue collection over a narrow time period, and minimize potential circadian effects, we sta gger ed the rat tr aining sc hedule .Suc h an approac h likel y r esulted in female rats being sta gger ed thr oughout their 5 d estr ous cycle , whic h would limit biasing to one phase of the estrous cycle.Our experimental design also ensured that all animals r ecei v ed the same degree of human handling, and environmental conditions (eg, bedding, feed, and ambient conditions) that were consistent upon animal arri v al and thr oughout the study.Our tr eadmill pr otocol w as car efull y chosen to allow pr ogr essi v e training of rats at standardized workloads.
While common endur ance exer cise-based interventions in rats include v oluntar y wheel running, swimming, and treadmill running, the latter was chosen for sev eral r easons.First, compared to swimming, which primarily employs the flexor muscles, running is a whole-body exercise modality that uses hindlimb flexor and extensor muscles. 53Second, the principle of pr ogr essi v e ov erload forms the foundation of a successful exercise intervention and treadmill training allows the exercise stimulus to "progress" in a controlled manner, thereby inducing an adapti v e r esponse. 54Tr eadmill training is also continuousmimic king progr ammed exer cise in humans-while wheel running is intermittent in nature.While there are advantages and disadv anta ges to all exercise modalities, 55 the consensus of the consortium was that treadmill training was the best mode of endur ance exer cise to meet the goals of MoTr-PAC, which included the potential to overlay and translate the preclinical animal studies in F344 rats (be it after acute e xercise or e xercise training) to that of the human arm of MoTrPAC.
A key adapti v e r esponse to endur ance exer cise tr aining is an increase in cardiorespiratory fitness or VO 2 max. 56 , 57Here, our training regime followed the seminal rat training study of Wisløff et al., 30 with modifications to a lower targeted continuous moderate-to-high intensity of 70%-75% VO 2 max.This intensity not only provides translational relevance to humans, 19 , 30 , 58 , 59 it also assimilates with the training protocol and target intensity of the human MoTrPAC studies. 18Moreo ver, tar geting this intensity (at a minimum) is important for treadmill training in Aged rats, as training at lower intensities does not elicit equita b le impr ov ements in VO 2 max when compared to young counterparts. 60 , 61Therefore, we chose a treadmill speed and grade to elicit similar r elati v e oxygen consumption in Adult and Aged rats 60 , 61 and humans. 59 , 62In line with the work of Wisløff, 30 we observed a robust improvement in a bsolute and r elati v e VO 2 max after 4 wk of tr aining, whic h contin ued to incr ease with 8 wk of training, in Adult, male and female rats, pairing with previous studies in various strains of rats. 30 , 63 , 60 , 64-66Consistent with the well-defined linear relationship between VO 2 and workload, 30 maximal run speed (MRS) increased in male and female Adult rats pr ogr essi v el y between 4 and 8 wk of training.Changes in VO 2 max wer e onl y measured at the 8-wk timepoint in Aged rats and also showed substantial impr ov ements.While VO 2 max w as not measur ed at the 4-wk timepoint in Aged rats, MRS increased at 4 wk followed by further impr ov ements at 8 wk indicati v e of pr ogr essi v e impr ov ements in cardior espirator y fitness in Aged animals.Such observ ations ar e consistent with studies in F344, 60 , 64 other r at str ains, 65 , 66 and humans, 59 , 67-69 whic h demonstr ate a 1%-31% increase in VO 2 max in response to training at a similar continuous intensity (60%-80% VO 2 max).It is notable to mention that interval treadmill training is capable of inducing more r obust impr ov ements in VO 2 max in rats 30 , 63 and humans. 70hile baseline and adaptability in VO 2 max to training can differ between inbred strains of rats 66 and amongst outbred rats, 71 observations of similar percentage improvements between rats and humans substantiate our training protocol utility and r e pr oducibility.Inter estingl y, when looking at individual training responses in VO 2 max, animals with a higher baseline VO 2 max tended to have a lower improvement in VO 2 max with training as compared to those with a lower baseline VO 2 max.This is likely because each group of animals, regardless of their baseline VO 2 max, trained at the same workload (ie, 70%-75% of the average VO 2 max for the cohort); as such, those with a lower baseline VO 2 max wer e likel y training at a higher r elati v e percenta ge of VO 2 max, and thus might be expected to have a greater adaptive r esponse.Ov erall, these data validate that the training protocol developed and implemented in PASS1B promotes similar cardior espirator y adaptations in male and female Adult and Aged r ats-warr anting additional investigation of systemic responses to pr ogr essi v e endur ance tr aining.
3][74] Here, training-induced changes in body composition were influenced by both age and sex.Over the 8-wk training period, fat mass decreased in both Adult and Aged males, whilst in females it only decreased in Aged, but not Adult rats.While Adult females did not lose fat mass with training, it should be noted that exer cise tr aining prevented the Adult females from gaining fat mass, as occurred in the age-matched SED rats.Sexual dimorphism in endurance training-induced fat loss is observed in weanling rats, 75 with females potentially displaying attenuated fat loss relative to males.9][80][81] Mechanisms of attenuated fat loss in female rodents and humans may be attributa b le to incr eased compensator y food consumption 82 , 83 and other ev olutionar y conserv ed molecular mechanisms to maintain r e pr oducti v e fitness in females. 26 , 75 , 84Nota b l y, while the underlying reason for changes in body and fat mass were not investigated in this study (all animals had ad libitum access to food and we did not measure food intake or 24 h energy expenditur e), r ecent m ultiomic work by MoTrPAC in the subcutaneous W AT (scW AT) of a subset of these Adults rats identified candidate molecules and pathw ays r egulating sexuall y dimorphic responses to exer cise tr aining. 26Interestingly, despite attenuated fat loss in Adult female rats, all groups decreased plasma le ptin lev els following training, which is suggesti v e of adipose tissue r emodeling tow ard a healthier phenotype and/or a decrease in visceral fat mass. 85Both Adult and Aged males displayed gr eater r eductions in plasma le ptin, pairing with changes in total fat mass and increase in glycerol levels in Adult males at 1 and 2 wk, indicati v e of lipolysis in the early training response.It is important to note that training-associated reductions in total body mass were accompanied by decreases in total lean mass in Aged, but not Adult male rats, and Aged female rats.Despite these reductions, the relative percentage of lean to total fat mass increased in Aged male, but not Aged female rats.An important point when interpreting these body composition c hanges is whic h body compartments the whole-body NMR is measuring.For example, lean (body) mass measurement is an assessment of all lean tissues, including skeletal m uscle, li v er, lungs, kidneys and heart; it does not include bone minerals, fat, and substances which do not contribute to the NMR signal, such as hair and claws.Thus, decreases (or changes in general) in total lean mass are not necessarily reflective of a decrease in skeletal muscle mass, but could be due to changes in mass in other tissue types.To this point, while there was a robust decrease in lean mass in Aged rats, terminal masses of LG, MG, PL, and SOL were similar between SED rats and the trained gr oups, and wer e ev en incr eased in 8 W Aged females, suggesting that the lower lean body mass in Aged trained groups may be due to reduced mass of other lean tissue compartments, not skeletal muscle mass.Thus, direct measures of skeletal muscle mass(es), where possible, can be helpful when interpreting body composition data, such as that provided by whole-body NMR.
Glucocorticoids play an important role in the adaptation to a variety of homeostatic stressors that perturb homeostasis, including exer cise .All tr aining groups display ed an increase in plasma corticosterone with progressive endurance training (weeks 1-4); lev els atten uated at 8 wk following a 2-wk plateau in training intensity and volume.Similar to our findings, during the initial weeks of chronic exercise training (up to 4 wk), plasma corticosterone concentrations have been reported to be higher in the rested and post acute exercise or restraint state and to decrease during subsequent training weeks as chronic central adaptations occur. 86 , 87Potential implications for this increase include effects on metabolism through actions on multiple organs including the li v er, adr enals, brain, skeletal m uscle, and white adipose tissue. 88i v en that skeletal m uscle meta bolic adaptations ar e essential for whole-body impr ov ements in aer obic fitness with endur ance tr aining, 89 w e assessed c hanges in CS , capillarization, and glycogen in four hindlimb muscles.As CS catalyzes the first step of the Krebs cycle, its activity is commonly used as a marker of skeletal muscle oxidative capacity.To this point, our training protocol resulted in a robust increase in CS activity in both Adult and Aged rats, regardless of sex, and in multiple muscles that were tested.Overall the changes we observed are consistent with previous training studies that measured CS activity 65 or SDH activity 60 in F344 rats of similar ages.Nevertheless, the temporal dynamics of CS activity did differ between Adult and Aged rats.In Adult trained rats, CS activity peaked at 4 wk, and then decreased by 8 wk, albeit to levels higher than SED rats.Gi v en the training protocol in weeks 7 and 8 was designed to be at a steady state, the adaptations in Adult rats may reflect a plateau in mitochondrial adaptations that occur in the absence of increased intensity, as originally proposed by Dudley. 58 Importantly, this decrease did not impact the increase in VO 2 max after 8 wk of training in Adult rats.Conv ersel y, CS acti vity peaked at 8 wk in Aged r ats, whic h may be r eflecti v e of differ ences in the rate of change in training intensity between the Adult and Aged rats between weeks.Alternati v el y, the continued increase in CS activity despite a plateau in training volume and intensity in Aged rats, has also been r e ported in humans, 62 where changes in vastus lateralis respiratory capacity following endur ance tr aining at 70% VO 2 max were observed to be higher in older r elati v e to young subjects despite similar impr ov ements in aer obic fitness. 59Nev ertheless, it is important to not ov erinterpr et c hanges in mitoc hondrial content based on one mitoc hondrial enzyme .To this point, recent multi-omic analysis of skeletal muscle (and the other tissues collected from this study in Adult r ats) demonstr ated a robust improvement in multiple markers of oxidative metabolism and mitochondrial capacity that were sustained through 8 wk of training. 28 , 29hronic endur ance exer cise has been shown to induce angiogenesis and increase capillarization in both human and rodent skeletal muscle. 90 , 91Changes in capillarity hav e been measur ed by a number of methods including: capillary density, capillaryto-fiber ratio, and capillary contacts.We measured the mean capillary contacts of fibers in four muscles of variable fiber type composition (SOL, PL, MG, and LG) and activity patterns during the moderate intensity exercise training.We found no significant increase in mean capillary contacts in the four muscles studied following 8-wk of treadmill training.The lack of an increase could be related to many factors including age, intensity of training, and duration of training.Our training program was at a moderate intensity for a duration of 8-wk, with the last 2 wk maintained at steady state (constant speed, incline and duration).Many studies that have observed an increase in capillarity have been at a higher intensity and for a longer duration (10-12 wk). 92 , 93It should be noted that our training protocol produced minimal changes in fiber CSA.While we did not find an increase in capillarity as measured by capillary contacts, we did observe that the SOL, a predominantly slow, oxidative muscle, had the highest mean capillary contacts at both ages and in both sexes.Inter estingl y, ther e w as a general tr end for males to hav e gr eater mean capillar y contacts in all muscles compared to females.
Muscle glycogen levels are also indicative of skeletal muscle training adaptations, in part due to greater muscle GLUT4 abundance 94 and sar colemma tr anslocation following acute exercise, 95 and elevated fatty acid oxidation that occurs with training resulting in glycogen sparing. 96We observed increases in m uscle gl ycogen content in all training groups at 8 wk in all m uscles exce pt for the type I fiber-dominant SOL muscle.While several human studies cite sex and age-specific differences in m uscle gl ycogen content and glucose kinetics following training, 59 , 97 such differences may be impacted by timing of sampling (eg, sample collection < 48 h after the last bout of exercise or before a plateau in training intensity) and/or reduced sample size in human trials.In our study, Aged animals display ed over all higher concentr ations of muscle glycogen, which may be r eflecti v e of differences in muscle fiber type distribution, substrate pr efer ence during exer cise , and/or functional capacity with aging. 98 , 99Inte gr ation of multiomic assays performed on these rats, will help identify molecular regulators contributing to age-and sex-specific differences in skeletal muscle metabolic adaptations to training.
The effect of endurance training on the fiber type composition of a muscle is dependent on the muscle type and the intensity and duration of the exercise training. 60 , 100-103Classification of muscle fibers based on MHC expression yields four primary fiber types in rodent limb muscle (I, IIa, IIx, and IIb) and thr ee primar y fiber types in human limb m uscles (I, IIa, and IIx). 104In humans, endurance training has been shown to promote a shift in MHC expression from IIx to war d IIa in the vastus later alis muscle . 105In the current study, we found a consistent shift from type IIx/IIb to type IIa in the PL muscle of both Adult and Aged rats, regardless of sex.A shift toward more type IIa fibers was also observed in the MG and LG of Aged but not Adult rats.The greater fiber type shifts in the Aged MG/LG compared to the Adult MG/LG may reflect increased recruitment of these muscles in the Aged rats during the treadmill running.7][108][109][110] Inter estingl y, ther e w as a noticea b le differ ence in the IIx/IIb ratio in female versus male rats, regardless of age.Sexual dimorphism in the fiber type composition of jaw muscles has been r e ported, 111 , 112 but to our knowledge sexual dimorphism in the mixed hindlimb muscles of rodents has not been studied in detail and the underl ying r easons for this difference are unknown.
Gi v en that skeletal muscle mass and fiber area is an important determinant of health and mortality, especially with adv ancing a ge, 2 , 4 , 6 , 113 we also assessed changes in overall and fiber type-specific CSA.A total of 8-wk of endurance training did not significantly impact m y ofiber CSA in Adult male rats.Adult female r ats, how e ver, displa yed fiber type-specific increases in the MG (IIb and IIx) and PL (IIa).Gi v en Adult females were generally better runners than Adult males, differences on the impact of endur ance tr aining on fiber CSA could relate to differences in recruitment and external loading.Conversely, in Aged rats with training only males increased mean fiber CSA in the SOL, likely dri v en by type I fiber-specific increases in the SOL.Increased CSA of type I fibers in Aged males was also observed in the LG and MG at 8 W. Aged females did not gain body or lean mass with training, but did display an increase in PL type IIa m y ofiber CSA.Increased type IIa CSA in the PL is consistent with findings from a similar training protocol in 25-month-old female F344 rats. 102ging is associated with muscle atrophy, especially of type II fibers in both rodents and humans. 60 , 114-117Atrophy of the type II fibers w as appar ent in the LG and MG of the Aged males relati v e to the Adult males.Collecti v el y, our data supports a growing body of evidence that endurance training may attenuate ageassociated selecti v e fiber atr ophy in older indi viduals. 118ince the seminal study by John Holloszy demonstrating the effects of endurance exercise training on mitochondrial mass and function in skeletal muscle, 19 thousands of studies ha ve in vestigated the salient effects of exercise on health and biology.Among these, other important works have described the effects of training duration and intensity, for example, on skeletal muscle mitochondria by muscle group and muscle type . 58How ev er, a pr ominent limitation in adv ancing the field is the overt lack of investigation into the effects of progressive endur ance exer cise tr aining over time , in males and females, at https://motrpac-data.org/publications/data/animal/phenotyp e/full-table-endurance-training.Notably, a subset of Adult animals from this cohort ( n = 6 per experimental group and sex) have undergone extensive -omic profiling, with this publicly accessib le r esource being pub lished by the MoTrPAC Study Group. 28Those animals used in the multi-omics analyses are identified in the collated data set.

Figure 1 .
Figure 1.MoTrP AC P ASS1B Study Overview and Design.( A ) Overview of cohort intake, testing, and pr ogr essi v e endur ance tr aining protocol in male and female Adult and Aged F344 rats.Schematic displays pretraining acclimation and familiarization protocol for all rat cohorts.Note, postexercise blood lactate concentration was also measured on the first of eac h tr aining w eek.Also, in the 18 mo cohort, an additional SED contr ol gr oup (for both sexes) w as a ge-matc hed to 1 W tr aining group; this gr oup w as included to account for potential aging effects.( B ) Overview of the timeline of events on the day of sacrifice.This figure was created with BioRender.com( www.biorender.com ) and confirmation of publication and licensing rights was obtained.final 10 d of the training intervention.If a rat was unable to perform at least 4 d of training per week, it was removed from the study and euthanized.Rats assigned to the Sedentary (SED) contr ol gr oup wer e placed on the tr eadmill for 15 min/d at 0 m/min for 5 d/wk and followed a schedule like the 8-wk training group.For insight into changes in body mass over time, at the beginning of each training week (including the first training session), body mass was measured in each rat immediately prior to beginning the treadmill session; rats were not fasted.Also, immediately after completing the first and fifth training session of each week, blood lactate concentration was measured via the tail vein (Lactate Plus meter).For each cohort, rats were
and descending) and feces, right testes or ovaries, right vastus lateralis, left SOL, gastrocnemius, and PL muscles, right tibia, and right femur.All tissues, except the left hindlimb muscles and fem ur, wer e flash fr ozen in liquid nitr ogen immediatel y upon r emov al, placed in cr yovials, and stor ed at −80 • C. The left SOL, PL, medial gastrocnemius (MG), and lateral gastrocnemius (LG) m uscles wer e r emov ed, weighed, pinned to cork, frozen in c hilled isopentane , and stored at −80 • C for histological analysis.The fem ur w as placed in 70% ethanol and stored at 4 • C. The time of r emov al and fr eezing w as r ecorded for all tissues.
.5 m m and 63.5 m m , r especti v el y), EDTA (10 m m ), DTNB (0.1 m m ), acetyl-CoA (0.1 m m ), and Triton X-100 (0.1% v/v).The r eaction w as initiated by the addition of 4 μL of m uscle l ysate (8 μg) and 5 μL of oxaloacetate (10 m m ).Absorbance at 412 nm (25 • C) w as measur ed at 5 min.Values were then normalized to protein content and compared to a standard curve made with purified CS (Sigma, C3260).
which uses the sequential binary partitions { I ||| IIa || IIx | IIb } to generate balances b1, b2, and b3. 47 (pp107-108), 50 1.b1 = 1 √ 2 log I g( I I a, I I x, I I b ) or l og I I I a (SOL) 2. b2 = 2 3 log I I a g( I I x, I I b ) 3. b3 = log I I x I I b where g ( •) denotes the geometric mean (the n th root of the product of n values: a measure of centrality) of the subcomposition.These partitions w ere c hosen for the following interpretations of their balances: 1. b1 = type I compared to { IIa, IIx, IIb } fibers 2. b2 = type IIa compared to { IIx, IIb } fibers 3. b3 = type IIx compared to type IIb fibers Reducing the compositions in the simplex S 4 to ilr coordinates in R 3 and the two-component composition (SOL) in S 2 5 ± 1.6 m m and 3.4 ± 1.4 m m for Adult females and males and around 4.7 ± 2 m m and 3.5 ± 1.7 m m in Aged females and males, r especti v el y ( Figure S4A -D ).

F igure 3 .
V O 2 max and MRS.Pre-and post-training measur es of a bsolute V O 2 max, V O 2 max r elati v e to total body mass, and MRS in Adult females ( A ) -( C ), Adult males ( D ) -( F ), Aged females ( G ) -( I ), and Aged males ( J ) -( L ).Each arrow or point represents a single rat, and they span from pre-to post-training values.Arrows are colored according to the direction of change from pre to post, and individual rats are arranged in ascending order by their pr etraining v alue within each gr oup.P -v alues wer e obtained from two-sided one sample t -tests of the (post-pre) differences, and they were Holm adjusted within each combination of age and sex.

Figure 4 .
Figure 4. Delta VO 2 max.Change in absolute ( A ) and relative ( B ) VO 2 max from pre-to post-training.Horizontal lines represent the mean of each group.A colored line indicates that the mean of the group was significantly different from zero, while a black line indicates that the mean was not significantly different from zero.Exact P -v alues ar e shown in Figur e 3 .

Figure 5 .
Figure 5. Body composition.Pre-and post-training measures of body composition (body mass, lean mass, and fat mass) in Adult females ( A ) -( C ), Adult males ( D ) -( F ), Aged females ( G ) -( I ), and Aged males ( J ) -( L ).Each arrow or point r e pr esents a single rat, and they span fr om pr e-to post-training values.Arrows are colored according to the direction of change from pre to post, and individual rats are arranged in ascending order by their pretraining value within each group.P -values were obtained from two-sided one sample t -tests of the (post-pre) differences, and they were Holm adjusted within each combination of age and sex.

Figure 6 .
Figure 6.Delta body composition.Change in body mass ( A ), lean mass ( B ), and fat mass ( C ) from pre-to post-training.Horizontal lines represent the mean of each gr oup.A color ed line indicates that the mean of the gr oup w as significantl y differ ent fr om zer o, while a b lack line indicates that the mean w as not significantl y differ ent fr om zer o.Exact P -v alues ar e shown in Figur e 5 .SED in the LG ( + 25%; F igure 8 M), MG ( + 34%; F igure 8 N), and SOL ( + 24%; ( Figure 8 P), but not the PL ( Figure 8 O).

Figure 7 .
Figure 7. Mean fiber type %.Mean percentage of each fiber type (I, IIa, IIb, and IIx), determined by MHC expression, in the LG, MG, PL, and SOL of Adult females ( A ) -( D ), Adult males ( E ) -( H ), Aged females ( I ) -( L ), and Aged males ( M ) -( P ).Each donut chart summarizes measurements taken from 6 rats.Superscript numbers denote a significant difference (two-sided, two-sample t -test; Holm P < .05) between the 8 W and SED means for a particular fiber type ratio (described on the right of the figure and in the "Fiber-Type-Specific Measures: Fiber type distribution" Methods).

Figure 8 .
Figure 8. Fiber-type-specific CSA.Mean CSA of each fiber type for the LG , MG , PL, and SOL muscles from Adult females ( A ) -( D ), Adult males ( E ) -( H ), Aged females ( I ) -( L ), and Aged males ( M ) -( P ).Boxes are 95% confidence intervals for the mean CSA of each group.For each muscle and fiber type, the 8 W trained gr oup w as compar ed to SED, and P -v alues wer e Holm-adjusted across all fiber types for a gi v en combination of age, sex, and muscle .Br ac kets indicate a statistically significant difference between groups (Holm P < .05).

Figure 9 .
Figure 9. Citrate synthase activity by muscle.Citrate synthase activity in the LG, MG, PL, and SOL muscles of Adult females ( A ) -( D ), Adult males ( E ) -( H ), Aged females ( I ) -( L ), and Aged males ( M ) -( P ).Each trained group was compared against the SED group using the Dunnett test.Br ac kets indicate a significant change in CS from SED to trained (Dunnett P < .05).
10 A-C), in female rats.Glycogen content in males displayed similar training responses as females, with a few exceptions ( Figure 10 E-H).Similar to females, glycogen content was elevated in the LG ( Figure 10 E) and PL ( Figure 10 G) at 8 W by + 2.10fold and + 3.55-fold, r especti v el y, though ther e w as no change in MG ( Figure 10 F) glycogen with training.In the PL ( Figure 10 G) there also was an early increase at 1 W (2.23-fold), and in the SOL ( Figure 10 H) a 33% decr ease in the mean gl ycogen at 4 W only. Aged group.In the Aged females, glycogen content increased in the 8 W group relative to SED in the LG ( + 3.43-fold; Figure 10 I), MG ( + 4.08-fold; Figure 10 J), and PL ( + 4.44-fold; Figure 10 K).The MG in females also displayed an increase in glycogen content at 4 W by + 2.97-fold ( Figure 10 J).In the SOL, no increases were

Figure 10 .
Figure 10.Glycogen by muscle.Glycogen concentration in the LG , MG , PL, and SOL muscles of Adult females ( A ) -( D ), Adult males ( E ) -( H ), Aged females ( I ) -( L ), and Aged males ( M ) -( P ).Each trained group was compared against the SED group using the Dunnett test.Br ac kets indicate a significant change in glycogen from SED to trained (Dunnett P < .05).observ ed, with gl ycogen content being significantl y lower in the 1, 2, and 4 W groups, as compared to SED ( Figure 10 L).In males, glycogen concentration was higher in 8 W, as compared to SED, in the LG ( Figure 10 M), MG ( Figure 10 N), and PL ( Figure 10 P), with fold-increases of + 3.34, + 2.98, and + 3.67, r especti v el y ( Figure 10 M-P).

Figure 11 .
Figure 11.Systemic hormones.Levels of plasma insulin, glucagon, corticosterone, and leptin in Adult females ( A ) -( D ), Adult males ( E ) -( H ), Aged females ( I ) -( L ), and Aged males ( M ) -( P ).Each trained group was compared against the SED group using the Dunnett test.Br ac kets indicate a significant change in these hormones from SED to trained (Dunnett P < .05).Measurements were performed in all Adult rats, and only in the -omics cohort of Aged rats.

Figure 12 .
Figure 12.Clinical meta bolites.Lev els of plasma glucose , lactate , NEFA, gl ycer ol, and total ketones in Adult females ( A ) -( E ), Adult males ( F ) -( J ), Aged females ( K ) -( O ), and Aged males ( P ) -( T ).Each trained group was compared against the SED group using the Dunnett test.Br ac kets indicate a significant change in these metabolites from SED to trained (Dunnett P < .05).Measurements were performed in all Adult rats, and only in the -omics cohort of Aged rats.