# pH-Resolved ATP Synthesis in Skeletal Muscle: Concept, Implementation, and Assessment Using Dynamic 31P Magnetic Resonance Spectroscopy at 7T

## TL;DR

This study uses advanced MRI to show how pH levels in muscle tissue affect ATP production during and after exercise.

## Contribution

A new multi-pool model reveals pH-dependent ATP synthesis and metabolic heterogeneity in skeletal muscle.

## Key findings

- Pi recovery is slower at lower pH levels, showing a strong pH dependence.
- Aging slows Pi recovery under acidic conditions but not at neutral or alkaline pH.
- The multi-pool model provides direct evidence of metabolic heterogeneity in muscle.

## Abstract

Background/Objectives: Dynamic changes in inorganic phosphate (Pi), phosphocreatine (PCr), and pH during post-exercise recovery reflect underlying muscle energetics and mitochondrial ATP synthesis, but the conventional single-pool model assuming uniform metabolic response fails to address myofiber composition and pH-dependent metabolic heterogeneity in skeletal muscle. This study aimed to characterize the interplay between pH, Pi, and PCr, and to develop an analytical method for assessing pH-resolved ATP synthesis using 31P MRS. Methods: Five healthy subjects underwent dynamic 31P MRS scans during plantar flexion exercise. ATP synthesis was evaluated from post-exercise PCr and Pi recovery time courses using the single-pool model, and from Pi recovery time courses using a multi-pool model in which the Pi signal lineshape was segmented into four pH-specific pools: alkaline (pH 7.3 ± 0.2), neutral (pH 7.0 ± 0.1), mildly acidic (pH 6.8 ± 0.1), and moderately acidic (pH 6.6 ± 0.1). Results: The single-pool model showed that during exercise, Pi increased proportionally to PCr depletion, and both Pi and PCr recovered monoexponentially immediately after exercise with τPi 33±9 s<τPCr (40 ±9 s); ATP remained stable while pH exhibited a “heart-beat” pattern, characterized by an initial alkalization followed by neutralization during exercise, a post-exercise acidic undershoot, and a subsequent slow recovery (τpH≫τPCr). The four-pool model demonstrated a pronounced pH dependence of Pi recovery, with slower recovery at lower pH (τPi: 19 ± 6 s at pH 7.3, 25 ± 7 s at pH 7.0, 32 ± 11 s at pH 6.8, and 46 ± 17 s at pH 6.6). Pi recovery is slowed with aging under acidic conditions, with little or no effect observed at neutral or alkaline pH. These results provide new insights into skeletal muscle metabolic heterogeneity, reflecting how different myofiber microenvironments modulate ATP synthesis. Conclusions: By overcoming the constraints of the single-pool model, the proposed multi-pool framework uncovers pH-dependent ATP synthesis and provides direct evidence of pronounced metabolic heterogeneity in skeletal muscle during exercise and recovery.