The hierarchy of metabolic fuel selection in resting skeletal muscle, where fat oxidation inhibits carbohydrate utilisation¹, is thought not to operate during exercise^2,3. However, previous work underpinning our understanding of skeletal muscle substrate utilisation during exercise has relied predominantly on the measurement of by-products of fat and carbohydrate metabolism in a variety of cellular and animal models. These include tracer experiments following electrical stimulation of muscle homogenates⁴ or isolated perfused whole-muscle preparations⁵, and measurement of arterio-venous differences across electrically stimulated whole-limb preparations⁶. These methods are unable to describe directly the changes in cellular metabolic flux during exercise. The aim of this project was to develop a technique that allows metabolism and functional measurements of exercising skeletal muscle in real-time, in an intact animal. A method of in vivo murine gastrocnemius muscle stimulation was established for measurement of force production and fatigue resistance. We combined this with a protocol measuring muscle metabolism using hyperpolarized [1-¹³C] pyruvate magnetic resonance spectroscopy (MRS), to follow conversion of pyruvate to downstream metabolites including lactate, alanine and bicarbonate⁷. This technique allows for the simultaneous measurement of metabolic flux and assessment of skeletal muscle force production and fatigue resistance during exercise in intact animals.

Hyperpolarized 13C MRS protocol: [1-¹³C] pyruvate was hyperpolarized using a prototype hyperpolarizer and dissolved as previously described^8,9. An aliquot of 0.2 mL of 80 mM hyperpolarized [1-¹³C] pyruvate solution was injected over 10 seconds via a tail vein cannula into an anaesthetised mouse positioned in a 7 T preclinical Varian/Agilent MR scanner. Spectra were acquired for one minute post-injection, using a 10 µs 15º hard excitation pulse (TR=1 s, 8kHz bandwidth). Signal was localised to the gastrocnemius muscle using a 13C RF surface coil. The first 30 spectra after the appearance of the pyruvate peak were summed and analysed using the AMARES algorithm in the jMRUI software package¹⁰ and results are shown as the ratio of the returned amplitude of the metabolite of interest to that of pyruvate.

In vivo gastrocnemius muscle stimulation: Three month old C57BL/6 male mice underwent an in vivo gastrocnemius muscle stimulation protocol in conjunction with two hyperpolarized scans, one before the stimulation protocol with the muscle in a resting state and one during the stimulation protocol (n=3, mean body mass 29 g). An anaesthetised mouse was placed in a bespoke Perspex cradle designed and manufactured for this protocol. A cannula was inserted into the tail vein. The sciatic nerve was isolated surgically and electrodes were placed distal to the tibial nerve branch. The knee and ankle joints were immobilised, the calcaneal tendon was attached to a force transducer via a suture thread, before a ¹³C RF saddle coil was placed over the muscle. Gradient echo ¹H localiser images were then used to obtain the cross-sectional area (CSA) of the gastrocnemius muscle. Hyperpolarized [1-¹³C] pyruvate was injected in to the tail vein and spectra acquired every second over one minute from the muscle at rest. After 30 minutes a stimulation protocol¹¹ consisting of a train of eight pulses of 100 μs at 30 Hz followed by a rest period of 1.25 seconds was repeated over a 10 minute period using a PowerLab system and induced force production measured over this time. When the gastrocnemius muscle was exercising at a steady state, around four minutes in to the exercise period, hyperpolarized [1-¹³C] pyruvate was again injected in to the tail vein and spectra acquired as before.

¹³C spectra were acquired successfully at rest and during exercise. Figure 1 shows the force produced over the ten minute stimulation period normalised to the CSA of the hindlimb. Figure 2 shows an increase in flux through pyruvate dehydrogenase demonstrated by a significant (p =0.031) increase in [1-¹³C] label incorporation from pyruvate through to bicarbonate during exercise compared to muscle at rest. A method of in vivo gastrocnemius muscle stimulation was successfully established. The protocol mimicked the cyclical contraction pattern the gastrocnemius muscle would undergo during a running motion. Combining this stimulation method with hyperpolarized [1-¹³C] pyruvate MRS facilitated the simultaneous assessment of skeletal muscle metabolism and function. Our results demonstrate that the technique is sensitive enough to distinguish the differences in metabolic flux between the resting and exercising state. This technique may provide new insights into muscle fuel utilisation during exercise, which can be complicated by disease or nutrition.