AMPK (5′-AMP-activated protein kinase) is an energetic sensor for metabolic regulation and integration. Here, we employed CRISPR/Cas9 to generate non-activatable Ampkα knock-in (KI) mice with a mutation of the threonine 172 phosphorylation site to alanine, circumventing the limitations of previous genetic interventions that disrupt the protein stoichiometry. KI mice of Ampkα2, but not Ampkα1, demonstrated phenotypic changes with increased fat-to-lean mass, impaired endurance exercise capacity, and diminished mitochondrial maximal respiration and conductance in skeletal muscle. Integrated temporal multi-omic analysis (proteomics/ phosphoproteomics/ metabolomics) in skeletal muscle at rest and during exercise establishes a pleiotropic yet imperative role of Ampkα2 T172 activation for glycolytic and oxidative metabolism, mitochondrial respiration, and contractile function. Importantly, there is a significant overlap of skeletal muscle proteomic changes in Ampkα2 T172A KI mice with those of type 2 diabetic patients. Our findings suggest that Ampkα2 T172 activation is critical for exercise performance and energy transduction in skeletal muscle and may serve as a therapeutic target for type 2 diabetes.

LC-MS/MS Proteomics & Phosphoproteomics

Muscle tissue samples from male and female WT and KI mice were collected before exercise (i.e., sedentary control, SED), after 10-min running (10 min), and after exhaustion (Exh.) and stored at -80ºC until undergoing processing. Sample processing for global proteomics and phosphoproteomics analysis was conducted as previously described (71) with several modifications. Briefly, frozen muscle tissue samples were minced into small pieces on a pre-chilled aluminum tray with dry ice. ~ 25 mg of tissue were collected and homogenized in cold lysis buffer (50 mM Tris, 8 M urea, 75 mM NaCl, 1 mM EDTA, 2 µg/mL aprotinin, 10 µg/mL leupeptin, 1 mM PMSF, 10 mM NaF, 1% phosphatase inhibitor cocktail 2, 1% phosphatase inhibitor cocktail 3, pH 8.0) on a pre-chilled bead beater using 2-min cycle. Samples were then centrifuged at 13,000xg at 4 °C for 10 min to remove debris. The extracted protein was quantified by the CA assay. The same quantity of protein was aliquoted and reduced with 5 mM DTT for 1 Ampk shaking on a thermomixer at 37 °C, 1000 rpm. Proteins were then alkylated with 20 mM iodoacetamide with shaking in the dark at room temperature, 1000 rpm for 45 min. Samples were diluted 4-fold, then digested with trypsin (enzyme to protein ratio = 1: 50) for 1 hour, shaking at 1000 rpm, 37°C. A fresh aliquot of trypsin (enzyme to protein ratio = 1: 50) was added after the samples were diluted 2-fold. Then the digestion was conducted at 37°C, 1,000 rpm overnight. The digest was desalted by Sep-Pak C18 SPE cartridges (Waters, Milford, MA). Clean peptides were quantified by the BCA assay. 250 µg of peptides from each sample was aliquoted and concentrated in a speedvac to completely dry, to be used for TMT labeling.

Peptides were resuspended in 500 mM HEPES (pH 8.5) at 5 μg/μL. Three replicates of male or female muscle tissue samples from each condition (WT-SED, WT-10 min, WT-Exh., KI-SED, KI-10 min, KI-Exh.) were labeled with two TMT18 plexes. TMT reagents were resuspended in anhydrous acetonitrile at 20 μg/μL and added to each sample at a 1:2.5 (peptide: TMT) ratio. Labeling was conducted at 25 °C, 850 rpm for1 hourr shaking on a thermomixer. Then the reaction was quenched by hydroxylamine. Samples from each plex were combined and concentrated in a SpeedVac,c followed by C18 SPE cleanup. The clean TMT-labeled samples were then fractionated into 12 fractions using high pH reversed-phase separation. 5% of each fraction was used for global proteomics analysis; the remaining fractions were subjected to immobilized metal affinity chromatography (IMAC) phosphoenrichment using freshly prepared Fe³⁺-NTA-agarose beads.

Both global and phosphopeptide fractions were analyzed usinga  Waters nanoAcquity UHPLC system (with 20 cm x 75 um i.d. 1.9-um column packed in-house with Waters BEH C18 silica) coupled to an Orbitrap Fusion Lumos (Thermo Scientific, San Jose, CA) with a 120-min LC gradient. Positive ion mode spray voltage was set at 2.2 kV. Full MS spectra were recorded at a resolution of 60K with a scan range of 350 – 1800 m/z. Automated gain control (AGC) value was set 4to e5. MS/MS was acquired in data-dependent mode (DDA) at a resolution of 50K, AGC of 1e5. Isolation window was 0.7 m/z. High-energy collision dissociation (HCD) with a normalized collision energy setting of 30% was used. Dynamic exclusion time was set at 45s.

Untargeted Metabolomics

Muscle tissue samples from male and female wild-type (WT) and knock-in (KI) mice were collected as described above. For extraction, tissues were thawed on ice, and approximately 750 µL of a cold chloroform: methanol (1:2) mixture was added, along with four steel balls (Fisher Brand; diameter 2.4 mm). Tubes were plunged into liquid nitrogen for 5 minutes and vigorously shaken in a Fisher Brand Bead Mill 24. Tubes were then vortexed and shaken at 900 rpm for 30 minutes at 4°C in a temperature-controlled thermal shaker. After adding 400 µL of water, samples were vortexed, and the upper aqueous phase was recovered as the metabolite mixture. Two hundred microliters of the extract were dried overnight in a SpeedVac and reconstituted in 200 µL of 0.1% formic acid (FA) containing the 100X Metabolomics QReSS™ Kit (Cambridge Isotopes, MSK-QRESS-KIT). A 10 µL aliquot from each tube was removed to create a pooled quality control (QC) sample, which was injected at the beginning and end of the mass spectrometry (MS) sequence. Additional QC samples were injected after every five sample injections.

LC-MS data acquisition was performed using a fully automated AcquireX Intelligent Data Acquisition Workflow with additional instrument parameters. The data acquisition and data processing workflow is based on a previous publication (72).  We first generated an exclusion list from a reagent blank sample to determine the background. A pooled quality control (QC) sample was injected for feature detection and component assembly to create an inclusion list. Using the inclusion list, a series of iterative data-dependent acquisition (DDA) injections were performed, with each injection informed by the previous one. Precursors from the inclusion list were fragmented, and once detected, they were automatically transferred to the exclusion list. This approach minimized redundant fragmentation and maximized relevant spectra and metabolite annotation. Samples were analyzed on a Thermo Orbitrap IDX Tribrid MS coupled to a Thermo Vanquish UHPLC. Metabolite separation was achieved using a Waters BEH C18 column (Waters Corp.; 2.1 × 150 mm, 1.7 µm) maintained at 30°C. Samples were analyzed in both positive and negative ionization modes  

AMPK Consensus Motif Mapping

Each peptide was 15 amino acids in length, centered on the phosphorylated residue (position 0). Motif detection was performed based on known AMPK substrate preferences (52). Three levels of stringency were applied: strict, relaxed, and minimal motif match. The pattern for strict match was [I/V/L/M] - X - [R/K] - X - X - [S/T] - X - X - X - [I/F/L/M], in which the −5 position must be I, V, L, or M, −3 must be R or K, and +4 must be I, F, L, or M. The pattern for relaxed match was X - X - [R/K] - X - X - [S/T] - X - X - X - [I/F/L/M]. For broader detection, a minimal motif was annotated requiring the presence of a basic residue at -3 with the pattern of X - X - [R/K] - X - X - [S/T] - X - X - X - X - X.