To assess the ability of AdipoRon to influence skeletal muscle aging, AdipoRon (1.2 mg/kg) or vehicle (PBS) were administered intravenously to 25-month-old C57BL/6J male mice three times per week for 6 weeks. AdipoRon treatment resulted in a decline in body weight in aged animals, which did not reach statistical significance (Figure 1A). Body composition analysis revealed significant loss in fat mass in AdipoRon treated animals, although there was no difference in lean mass compared to Controls (Figure 1B–C). AdipoRon treated animals had significantly lower fasting glucose levels (Figure 1D) and a trend toward decreased fasting insulin levels (Figure 1E). In line with these findings, the HOMA-IR index of insulin resistance was significantly lower in AdipoRon treated animals. Circulating levels of triglycerides were not different between Control and AdipoRon treated animals. In metabolic chamber assessments, food consumption, spontaneous activity, energy expenditure (EE), or respiratory exchange ratio (RER) did not differ significantly between Control and AdipoRon treated animals over a 24-hr period (Figure 1F; Figure 1—figure supplement 1–S1). Parsing these data into smaller increments revealed differences in EE and RER at discrete time points only (t-test; unadjusted p<0.05). In gastrocnemius, a mixed fiber muscle in the lower leg with Type II fibers predominating, PGC-1a protein levels were higher from AdipoRon treated mice than from controls (Figure 1G), suggesting that activation of PGC-1a could be mediating the effects of AdipoRon in aged mice. Levels of VDAC (voltage-dependent anion channel), an index of mitochondrial content, were not different between control and AdipoRon treated mice, indicating that mitochondrial biogenesis was not induced by AdipoRon in gastrocnemius.

The therapeutic potential of AdipoRon was evaluated via in vivo and ex vivo functional measures. Baseline measures for rotarod did not differ between the groups; however, after 6 weeks of treatment, the AdipoRon treated animals tended to spend more time on the rotarod (p=0.08) suggesting a trend toward improved muscle function and agility (Figure 1H). Ex vivo contractility measurements in extensor digitorum longus (EDL, predominantly glycolytic) and soleus (predominantly oxidative) muscles were performed to determine the effects of AdipoRon treatment on whole muscle contractile behavior. Muscle maximal specific force generation and fatigability are known to be dependent on the type and quantity of expression of myosin heavy chains isoforms (Geiger et al., 2000). Consistent with this, the decline of tetanic force after initial simulation (fatigue) observed in Type II fiber-enriched EDL was greater than that of Type I enriched soleus muscle in untreated aged mice. Force generation after repeated tetanus stimulation was significantly greater in the EDL from AdipoRon treated animals compared to controls, signifying a greater resistance to fatigue (Figure 1I). Electrical stimulation-induced force generation after a 20-min recovery period was also significantly higher, indicative of greater muscle endurance following AdipoRon treatment. Interestingly, the beneficial effects of AdipoRon on fatigability were detected in EDL but not soleus muscle that had numerically greater values for both parameters at baseline compared to EDL (Figure 1J). These data suggest that oxidative muscles (soleus) are less sensitive to AdipoRon treatment compared to muscles with more reliance on glycolytic metabolism (EDL). PGC-1a is a key regulator of fiber type switching (Lin et al., 2002) and is engaged downstream of adiponectin receptor signaling (Iwabu et al., 2010).

A change in muscle fiber type distribution within muscle is an outcome of chronic exercise (Mishra et al., 2015; Yan et al., 2011), where composition is adapted to functional demand. Differences in fiber type distribution within muscle are associated with distinct contractile properties (Zierath and Hawley, 2004). Gastrocnemius muscle was analyzed as representative of a predominantly glycolytic muscle (Type IIb and Type IIx muscle fibers). Immunohistological detection of myosin isoforms in gastrocnemius muscle cross-section (Figure 2A) revealed a significantly greater proportion of Type IIa myosin expressing fibers in muscle from AdipoRon treated mice compared to untreated control mice (Figure 2B). Histochemical detection of cytochrome c oxidase activity revealed an increase in intensity of COXIV activity staining in Type IIb fibers (Figure 2C). Interestingly, AdipoRon treatment was associated with a very small decrease in fiber size in Type IIb fibers (Figure 2D), indicating there may be structural adaptation to changes in mitochondrial activation status. Quantitative PCR confirmed Type IIb myosin isoform was by far the dominant fiber type in gastrocnemius. AdipoRon treated mice had a modest reduction in expression of all myosin isoforms compared to controls although none of the differences were significant, and overall fiber type distribution was not altered (Figure 2E). Soleus muscle is a predominantly oxidative muscle, and RT-PCR confirmed enrichment for Type I and Type IIa myosin isoforms (Figure 2—figure supplement 1). Although functional differences were not detected ex vivo in soleus, AdipoRon treatment significantly increased expression of Type I and Type IIx myosin, and numerically increased expression of Type IIa and Type IIb, although here too the overall distribution of fiber types was not altered (Figure 2—figure supplement 1).

To investigate potential feedback from AdipoRon treatment, transcriptional expression of adiponectin receptors Adipor1 and Adipor2 was measured relative to the expression of Rn18s ribosomal RNA. In gastrocnemius muscle, levels of Adipor1 were significantly greater than Adipor2 (Figure 2F). Levels of both receptors were higher in old mice compared to young mice, corroborating previous reports that adiponectin receptor expression changes with age (Ito et al., 2018). Expression levels of both receptors were numerically lower in AdipoRon treated mice and not significantly different from the levels in young mice. In soleus muscle, Adipor1 levels were higher than those of Adipor2, but the expression was not altered with age (Figure 2—figure supplement 1). Although levels for each receptor were numerically higher in AdipoRon treated mice, neither change reached significance. The known age-associated increase in fibrosis in muscle (Wood et al., 2014) prompted an investigation of the effects of AdipoRon treatment on collagen content in aged animals. Histochemical staining for collagen showed that over this relatively short treatment period AdipoRon did not impact collagen content or fibrosis in the aged muscle (Figure 2—figure supplement 2). Taken together, these data show that chronic AdipoRon treatment mimics exercise including changes in fiber type composition with more oxidative fibers, a shift in glycolytic muscle fibers toward a more oxidative phenotype, and a correction of age-related elevation of adiponectin receptor expression.

To assess the chronic effects of AdipoRon on metabolism and muscle function in young mice, we administered AdipoRon (1.2 mg/kg) or vehicle (PBS) three times per week for 6 weeks by tail vein injections beginning at 2 months of age in male B6C3F1 hybrid mice. Chronic AdipoRon treatment did not alter body weight (Figure 3A) fasting glucose, nor insulin levels in young animals (Figure 3B–C). Ex vivo contractility measurements were performed in EDL and soleus muscles from the young Control and AdipoRon treated mice. Chronic AdipoRon treatment resulted in significantly improved ability to recover after repeated tetanus stimulation in EDL muscle suggesting higher resistance to fatigue (Figure 3D). Consistent with results from the aged mice, the effect of AdipoRon was muscle-type specific as it did not impact measures of contractility and force in the soleus muscle (Figure 3E). This confirms that AdipoRon preferentially improves function in fast muscle with predominantly Type IIb/IIx fibers independent of the age of the mice during treatment. PGC-1a protein levels were higher in gastrocnemius from AdipoRon treated mice than from controls (Figure 3F). Total transcript levels of PGC1a were higher in gastrocnemius from AdipoRon treated mice but not in soleus (Figure 3G), although transcript levels of selected PGC1a target genes (Cpt1a, Acox1, and Acadm) were not altered by AdipoRon treatment. The Ppargc1 gene produces several isoforms including PGC-1a1 and PGC-1a4 from the canonical promoter and variants PGC-1a2, and PGC-1a3 from an alternate promoter (Ruas et al., 2012). AdipoRon altered the profile of PGC-1a isoforms indicating a change in promoter preference .

To determine whether AdipoRon mediated changes in muscle function were coincident with changes in muscle composition, tissue sections from young mice were subject to immunodetection of myosin heavy chain isoforms. No changes in fiber cross-sectional area were identified, presumably because young mice are in peak physical condition (Figure 3—figure supplement 1). Histochemical detection of maximal cytochrome c oxidase (Complex IV, COXIV) and succinate dehydrogenase (Complex II, SDH) activity (Mahad et al., 2009), similarly showed no differences in young mice with or without AdipoRon treatment (Figure 3—figure supplement 1). Fiber area measurements revealed that the fiber size was inversely proportional to mitochondrial activity (Type IIb>IIx>I>IIa), which has been described previously as the muscle fiber type-fiber size paradox (van Wessel et al., 2010).

To understand the differences among muscle groups in terms of PGC-1a expression and adiponectin receptor distribution, gene expression analysis was conducted in extracts from gastrocnemius and soleus muscles from young male B6C3F1 hybrid mice (6 weeks old). Endogenous levels of Ppargc1a expression were significantly higher in soleus than gastrocnemius muscle (Figure 4A), which is to be expected due to the higher dependence of soleus muscle on oxidative metabolism (Koves et al., 2005). Interestingly, the distribution of isoforms was largely equivalent between the two muscle types, with Ppargc1a1 being the dominant isoform, lesser contributions from Ppgargc-1a4 and Ppargc1a2, and lowest levels detected for Ppargc1a3 (Figure 4B). The possibility that differences in responsivity to AdipoRon between the gastrocnemius and soleus muscle types observed in the chronic treatment studies could be due to differences in adiponectin receptor expression was investigated. Gene expression analysis revealed that in young mice, there was no overt difference in the expression levels of Adipor1 and Adipor2 between gastrocnemius and soleus, though AdipoR1 expression was notably higher than Adipor2 for both muscle types (Figure 4C).

To investigate the early response to AdipoRon in skeletal muscle, the impact of a single intravenous injection of AdipoRon (1.2 mg/kg BW) was investigated in muscles from young male B6C3F1 hybrid mice (6 weeks old) collected 90 min postinjection. AdipoRon treatment increased the gene expression of Ppargc1a in gastrocnemius but not soleus muscle (Figure 4D). Exercise training has been reported to induce the expression of the canonical promoter-driven Ppargc1a1 and Ppargc1a4, with a lesser effect on isoforms expressed from the alternate promoter. Acute treatment with AdipoRon resulted in an upward trend of transcript levels of most highly expressed isoform Ppargc1a1, a significant increase in the Ppargc1a4 isoform in gastrocnemius muscle, although increases in expression of Ppargc1a2 and Ppargc1a3 did not reach significance (Figure 4E). No changes were seen with the expression of Ppargc1a isoforms in soleus muscle, and gene targets of PGC-1a involved in fatty acid metabolism (Cpt1a, Acox1, and Acadm) were increased in response to acute AdipoRon treatment in gastrocnemius but not in soleus muscle (Figure 4F). Exercise training has also been linked to changes in adiponectin receptor expression, where an 8-week exercise regimen increased bicep femoris muscle mRNA expression of Adipor1 but had no effect of Adipor2 (Huang et al., 2006). Within our study, AdipoRon similarly increased the expression of Adipor1 in gastrocnemius muscle but had no effect on levels of Adipor2 (Figure 4G). AdipoRon had no effect on the expression levels of either adiponectin receptor in the soleus muscle, indicating that the effects of AdipoRon are muscle-type specific.

Adiponectin receptor stimulation by AdipoRon has been linked to activation of AMPK and PPARa (peroxisome proliferator-activated receptor alpha) pathways in muscle and liver, including increased PGC-1a transcription and activation of expression of its gene targets (Okada-Iwabu et al., 2013). Studies of renal function in a genetic model of diabetes (db/db) show AMPK and PPARa engagement by AdipoRon and have revealed changes in lipid parameters in response to AdipoRon in renal endothelial cells (Choi et al., 2018). Less is known of the functional consequence of PGC-1a activation by AdipoRon as it relates to cellular metabolism and mitochondrial function. Cultured NIH-3T3 fibroblasts were treated for 48 hr with 50 µm AdipoRon treatment. Mitochondrial membrane potential measured by TMRE assay was increased after 2 and 4 hr of 50 µm AdipoRon treatment (Figure 5A). The impact of AdipoRon on mitochondrial function was coincident with an increase in cellular oxygen consumption measured over the same time period (Figure 5B). Cellular ATP concentration was initially lower at 2 and 4 hr following AdipoRon 50 µm treatment, but subsequently increased and was significantly higher 12 hr following treatment (Figure 5C). Adaptive responses to metabolism include possible effects on cellular growth and remodeling of the mitochondrial network. The doubling time of NIH-3T3 is approximately 24 hr and our previous studies have shown that increased respiration is accompanied by slower growth in cultured fibroblasts (Miller et al., 2019b). Consistent with this, cell proliferation was negatively impacted by 24 hr exposure to AdipoRon regardless of seeding density (Figure 5D). We had previously shown that mitochondrial organization is influenced by chronic elevation of PGC-1a (Miller et al., 2019b). Mitochondria were visualized in fixed fibroblasts by immunofluorescence detection of TOMM20 following 24 hr AdipoRon treatment (Figure 5E). Treated cells showed quantitatively significant differences in fluorescent intensity and in mitochondrial network architecture as mitochondria tended toward shorter branch morphologies. Finally, immunodetection of VDAC (voltage-dependent anion channel) revealed a modest but significant increase in VDAC, an index of mitochondrial content, matching the immunofluorescence data (Figure 5F). A significant increase in PGC-1a protein was also detected. These data reveal prominent early and adaptive effects of acute AdipoRon treatment on the mitochondrial regulator PGC-1a and on mitochondrial function and organization and link these PGC-1a-associated metabolic changes to delayed proliferation and growth.

In cultured muscle cells, AdipoRon activates AMPK and increases the expression of PGC1a and its gene targets (Ito et al., 2018; Okada-Iwabu et al., 2013). The degree to which AdipoRon action is conserved among other cell types and among species remains an open question. Adipor1 and Adipor2 are expressed in diverse cell types, including murine NIH-3T3, differentiated mouse myoblast C2C12 cells, and nonhuman primate peripheral blood mononuclear cells (PBMCs), although there is cell type specificity in levels and relative ratios of the two adiponectin receptors (Figure 6A). Our data revealed functional changes in mitochondria in response to AdipoRon treatment in NIH-3T3 fibroblasts and we sought to investigate whether the molecular underpinnings for these observations were linked to AMPK and to PGC-1a. NIH-3T3 fibroblasts were treated with AdipoRon treatment (10 or 50 µm) for 10 min. Increased levels of activating phosphorylation of AMPK at Thr172 were detected in extracts of AdipoRon treated cells compared to controls for both the lower and higher doses (Figure 6B). These data align with prior reports of AdipoRon directed AMPK activation in C2C12 myoblasts and glomerular endothelial cells (Choi et al., 2018; Ito et al., 2018; Okada-Iwabu et al., 2013). Increased levels of Ppargc1a transcript were detected after 90 min of AdipoRon treatment at 50 µm but not 10 µm concentration (Figure 6C). Although the transcript of PGC-1a itself was not increased at the lower dose, expression levels of known targets of PGC-1a (Cpt1a, Acox1, and Acadm) were significantly increased at 10 µm and at 50 µm, indicating that PGC-1a activation at the lower dose of AdipoRon may occur at the posttranscriptional level. On day 7, differentiated C2C12 myotubes AdipoRon significantly induced phosphorylation of AMPK at Thr172 (Figure 6D). Ppargc1a gene expression was induced at the higher dose only, and PGC-1a target Acox1, but not Cpt1 or Acadm, was also induced at the 50 µm dose (Figure 6E). Mitochondrial membrane potential was also increased in response to AdipoRon in the myotubes (Figure 6F). Myosin isoform profiling revealed enriched expression of isoforms that are primarily associated with an oxidative phenotype that in vivo is less responsive to AdipoRon treatment (Figure 6G). The response to AdipoRon was next investigated in a diverse cell population derived from nonhuman primates. PBMCs were isolated from blood taken from adult male rhesus monkeys. The cells were cultured overnight and subsequently treated with AdipoRon. Activating phosphorylation of AMPK at Thr172 was increased after 10 min of AdipoRon treatment at 50 µm but not 10 µm concentration (Figure 6H). One hour following AdipoRon treatment of PBMCs significant increases in transcript levels of Ppargc1a and PGC-1a gene targets were detected at both the lower and higher doses (Figure 6I). These data demonstrate that the molecular response to AdipoRon is conserved among cell types, although there is cell type specificity in the expression of adiponectin receptors and in the sensitivity of response through AMPK signaling and activation of PGC-1a transcriptional co-activation.

Two cohorts of mice were used. Six-week-old male B6C3F1 hybrid mice were obtained from Jackson Laboratories (Bar Harbor, ME) and 24-month-old male C57BL/6J mice from the NIA aged mouse colony. The animals were housed under controlled pathogen-free conditions. The animals were fed ad-libitum Purina LabChow 5001 diet with free access to water and placed on a 12-h light/dark cycle. AdipoRon (10 mg) (Adipogen, CA) was reconstituted with 500 µl of DMSO (stock) and stored at –20°C. For the acute experiment, the animals were given either a single intravenous dose of AdipoRon @1.2 mg/kg BW in PBS (n=5) or an equivalent volume of DMSO in PBS (n=3, Controls) in a total volume of 100 µl. EDL and soleus muscle groups were harvested after 90 min. For the chronic experiments in young B6C3F1 mice (n=12) and old C57BL/6J mice (n=20), the animals in each age group were divided into two groups and received either AdipoRon @1.2 mg/kg BW in PBS or an equivalent volume of DMSO in PBS via tail vein injections three times per week (Monday, Wednesday, and Friday) for 6 weeks. Body weight and food intake measurements were recorded once a week. Body composition was measured at the beginning and at the end of the treatment period using EchoMRI Body Composition Analyzer (Houston, TX). Rotarod balance measures were performed on a Columbus Instruments Rotamax system. During the test, the instrument speed increased from 5 to 60 rpm over a 3-min interval. Animals were trained on the instrument 1 day prior to the test and were allowed three attempts during test. Time on the rotarod and the speed when the mice falls are recorded. At the end of the treatment period, animals were sacrificed. EDL, soleus, and gastrocnemius muscle specimens were mounted in OCT oriented to present fiber cross-section for tissue sectioning or snap-frozen in liquid nitrogen and stored at –80°C. All animal protocols were approved by the Institutional Animal Care and Use Committee at the William S. Middleton Memorial Veterans Hospital.

Overnight fasting blood samples were collected for measures of glucose (OneTouch Ultra Blue glucometer), and insulin levels (Ultra-Sensitive Mouse Insulin ELISA Kit (CystalChem, IL)). HOMA-IR (homeostasis model assessment of insulin resistance) index was calculated as [fasting serum glucose×fasting serum insulin/22.5].

Metabolic parameters and activity were measured (Columbus Instruments Oxymax/CLAMS metabolic chamber system (Columbus, OH)) in mice acclimated to housing in the chamber for approximately 24 hr, followed by 24 hr continuous data collection period.

EDL and soleus hindlimb muscles were dissected and perfused with oxygenated (95% O₂, 5% CO₂) Tyrode’s solution at room temperature (RT; NaCl 145 mM, KCl 5 mM, CaCl₂ 2 mM, MgCl₂ 0.5 mM, NaH₂PO4 0.4 mM, NaHCO₃ 24 mM, EDTA 0.1 mM, and Glucose 10 mM). Isolated muscles were attached to a contractile apparatus capable of measuring force (Aurora Scientific) and were electrically stimulated using parallel platinum electrodes. Maximal twitch force and tetanic force were determined by adjusting the length of the muscle until maximal twitch force is reached, defining the optimal length. A 15-min period of rest allowed muscle to equilibrate to the new environment. Fatigability was defined as the decline in tetanic force following 10 min of continuous stimulation. Muscles were tetanically stimulated at 100 Hz for 500 ms every 5 se at a voltage that generates the maximal force. Following 10 min of fatiguing stimulation, a 20-min recovery period was allowed, where there was no stimulation in an oxygenated Tyrode’s perfusion. After the recovery period, muscle was electrically stimulated again to quantify the amount of recovery force. A recovery force higher relative to fatiguing force was taken as an indication that the decline in force during fatiguing stimulation is reversible. The amount of fatigue and recovery force was represented as a percentage of the initial force (before fatiguing stimulation). Following these measurements, fatigued muscles were weighed, quick-frozen in liquid nitrogen, and stored at –80°C.

Equal amounts of skeletal muscle protein extract (45 µg) were separated on Mini-Protean TGX precast protein gels (Bio-Rad, CA) and transferred to a PVDF membrane using a Trans-Blot semi-dry transfer system (Bio-Rad, CA). The membranes were blocked using 5% BSA in TBST for phospho-antibodies or 5% non-fat milk for other antibodies for 1 hr at RT. The following primary antibodies were used overnight: pAMPK Thr172 (Cell Signaling Technology #2535, Beverly, MA), AMPK (Cell Signaling Technology #2532), GAPDH (Cell Signaling Technology #2118), PGC1a (H300, Santa Cruz Biotechnology #sc-13067), and Beta-Actin (Sigma-Aldrich, #A1978), washed in TBST, and incubated with the respective HRP-conjugated secondary antibodies (Vector Labs, #PI:1000 and #PI:2000, Burlingame, CA) for 1 hr at RT. Proteins were detected (Supersignal West Pico or Femto Chemiluminescent substrate solutions [Thermo Fisher Scientific]) and digital images acquired (GE ImageQuant Gel Doc Imaging system). Densitometric analysis was carried out using Fiji software.

RNA was extracted in TRIzol reagent and isolated (Direct-zol RNA Miniprep kit [ZymoResearch, Irvine, CA]). cDNA was synthesized from 1 µg of RNA (High-Capacity cDNA Reverse Transcription Kit [Applied Biosystems]). Real-time PCR reactions were carried out using iTaq Universal SYBR Green dye mix (Bio-Rad) with 10 ng cDNA per reaction. The treatment effect was analyzed by 2^∆∆Ct method (Livak and Schmittgen, 2001) and presented either as ‘relative expression’ compared to Rn18s mRNA, or as ‘fold change’ where relative expression of control samples was normalized to 1. Primer sequences for measured transcripts may be found in the supplementary info (Supplementary file 1 Table 1).

Serial OCT mounted cryostat sections (10 µm in thickness) were cut at –14°C (Leica Cryostat (Fisher Supply, Waltham, MA)). Freshly cut sections were stained for cytochrome c oxidase activity and digital images were captured the same day as described previously (Pugh et al., 2013). Briefly, sections were air-dried at RT for 10–15 min, incubated in a solution of 0.1 M phosphate buffer, pH 7.6, 0.5 mg/ml DAB (3,3′-diaminobenzidine), 1 mg/ml cytochrome c, and 2 µg/ml catalase at RT for ~30 min, washed with PBS, dehydrated, cleared and mounted under a glass coverslip (Permount; Thermo Fisher Scientific). Similarly for succinate dehydrogenase activity staining, the frozen muscle sections were air-dried for 10–15 min and incubated in SDH reaction mixture containing 1.5 mM nitroblue tetrazolium (NBT) and 50 mM di-sodium succinate in 0.2 M PBS (pH 7.6) at RT for ~20 minutes, washed with PBS, dehydrated, cleared, and mounted under a coverslip with Permount (Punsoni et al., 2017). Muscle fibrosis was determined by trichrome staining (Abcam, MA). For stain intensity and/or stain area analysis in digital images (n=5 per tissue), see below.

Fiber types (I, IIa, IIb, and IIx) were quantified in gastrocnemius muscle. Sections were air-dried for 30 min and then rehydrated with PBS for 5 min. Sections were then blocked with 0.5% BSA 0.5% Triton X-100 in PBS for 30 min at RT, then incubated with a cocktail of primary antibodies purchased from Developmental Studies Hybridoma Bank (DSHB, University of Iowa): BA-D5 (IgG2b, supernatant, 1:100 dilution) specific for MyHC-I, SC-71 (IgG1, supernatant, 1:100 dilution) specific for MyHC-IIa and BF-F3 (IgM, supernatant, 1:7.5 dilution) specific for MyHC-IIb for 1 hr at RT. After three washes with PBS (5 min each), the sections were incubated with secondary antibody cocktail (Invitrogen) to selectively bind to each primary antibody: goat anti-mouse IgG1 conjugated with Alexa Fluor488 (Invitrogen, #A21121); goat anti-mouse IgG2b conjugated with Alexa Fluor350 (Invitrogen, #A21140); goat anti-mouse IgM conjugated with Alexa Fluor594 (Invitrogen, #A211044) for 1 hr in the dark at RT. After three washes with PBS (5 min each) and a brief rinse in water, the sections were mounted in 85% glycerol in PBS for imaging. Type IIa fibers will appear green, IIb as red, I as blue, and the fibers that are not stained by these antibodies will appear black and are classified as Type IIx.

Digital images were captured using a 20× objective in a Leica DM4000B microscope equipped with Retiga 4000R digital camera (QImaging Systems, Surrey, BC). Background correction was conducted for all images using an unstained adjacent area of the slide. Images were converted to 8-bit format and inverted. Fiber type was classified based on immunostaining for myosin isoform and individual fibers were outlined using a freehand tool and the cross-sectional area. Stain intensity of the cytochrome c oxidase and succinate dehydrogenase was quantified using ImageJ/Fiji software. Using a custom-generated algorithm, the blue stained areas for collagen were separated and highlighted from the rest of the image and % stained area was quantified (MIPAR software). Mitochondrial analysis was performed on cells in ImageJ by applying image deconvolution, background subtraction, adaptive binarization, and segmentation algorithms, followed by particle analysis and morphology analysis with the ImageJ plugin MiNA (Valente et al., 2017) to quantify intensity and mitochondrial branching.

NIH-3T3 fibroblasts were purchased (ATCC; CRL-1658) and cultured in DMEM supplemented with 10% bovine serum and 1% penicillin/streptomycin. The cells were treated with vehicle (DMSO in media) or AdipoRon (10 or 50 µM in media) for 10 (AMPK phosphorylation) or 60 min (gene expression analysis). After treatment, the cells were washed once with PBS and then lysed in TRIzol or RIPA buffer depending on the experiment. C2C12 cells were purchased (ATCC; CRL-1772) and grown in 10% bovine serum and 1% penicillin/streptomycin. Differentiation was initiated after 2–3 days growth by exposure to differentiation media: Serum-free DMEM supplemented with 2% equine serum, 1% penicillin/streptomycin, and 172 nM insulin. Myoblasts were fed differentiation media every 24 hr for 5–7 days until myotube morphology was reached. All cell lines were authenticated by STR profiling and cleared of mycoplasma contamination.

Cells were treated with vehicle or AdipoRon (50 µM) for 2 or 4 hr. For TMRE fluorescence, cells were seeded at 5×10⁴ overnight and treated with AdipoRon (50 µm) for 2 or 4 hr. Cells were then incubated for 30 min in 100 nM TMRE dye in media and, after equilibrating for 10 min, plates were read at 530 nm excitation/580 nm emission, parallel treatment with 10 µM uncoupler FCCP (Cayman Chemical) accounted for mitochondrial oxidative phosphorylation. Data shown as fluorescent ratio of given treatment over uncoupled control. For oxygen consumption assays, cells were trypsinized and equal numbers (4×10⁵) were resuspended in respiration buffer and loaded on to the PreSens Oxoplates (Regensburg, Germany). Appropriate ambient and anoxic controls were included in the same plate.

Relative ATP levels were quantified ATPLite assay (Perkin-Elmer, Waltham, MA; INFINITE M1000 PRO microplate reader (TECAN, Grodig, Austria)) in cells cultured as above and treated with vehicle or AdipoRon (50 µM) for the prescribed times.

NIH-3T3 cell proliferation was quantified according to the CyQUANT Cell Proliferation Assay Kit (Molecular Probes, Inc, Eugene, OR) at four cell densities 1600–6700 cells/well. In brief, cells were seeded in microplate wells in growth medium at desired densities along with serial dilutions of cells for determination of a cell number standard curve. Cells were incubated for 24 hr after which culture medium was removed, and the number of cells quantified according to the manufacturer’s instructions.

Blood was collected from primates at the Wisconsin National Primate Research Center (WNPRC) with the approval of the Institutional Animal Care and Use Committee of the Office of the Vice Chancellor for Research and Graduate Education of the University of Wisconsin, Madison. PBMCs were isolated from 6 ml of blood using SepMate-50 tubes (Stemcell Technologies, Cambridge, MA). Isolated primate PBMC’s (1×10⁶) were cultured in 10 cm plates in DMEM supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. Cells were treated with vehicle (DMSO in media) or AdipoRon (10 or 50 µM in media) for 5 (AMPK phosphorylation) or 60 min (gene expression analysis). PBMCs were collected by centrifugation and the cell pellet was washed and lysed in TRIzol or RIPA buffer.

Results are expressed as mean ±pooled SEM. For cell culture and animal experiments, data sets involving two groups were analyzed by unpaired two-tailed t-tests and for data sets with three groups one-way ANOVA followed by Bonferroni’s multiple comparison test with a cutoff of p<0.05.