Novel Humanized G6PD Deficient Mice Reveal Mechanisms of Increased Tolerance to Exercise As Gleaned By Multi-Omics in Models of Critical Speed

INTRODUCTION: Glucose-6-phosphate-dehydrogenase (G6PD) deficiency is the most common enzymatic deficiency in humans, affecting ~500 million people or ~6% of mankind. G6PD is the rate limiting enzyme of the pentose phosphate pathway (PPP), the main antioxidant pathway in mature red blood cells (RBCs). Activation of the PPP generates NADPH, an essential reducing equivalent for antioxidant metabolism that protects against intra- and extra-vascular hemolysis in response to oxidant stress. Impaired G6PD activity makes RBCs susceptible to hemolysis induced by oxidant challenges. Exercise-induced oxidant stress is a well-established concept in the field of physiology. Individuals with G6PD deficiency are recommended to refrain from strenuous exercise, as they are deemed to be at increased risk for hemolytic events in response to exercise-induced oxidant stress. While biochemically sound, this rationale has never been tested mechanistically, owing to ethical and technical limitations. As such, knowledge is limited regarding molecular impacts of exercise activity in G6PD deficient individuals.

METHODS: Here we generated novel humanized mouse models carrying either (i) human canonical G6PD (normal activity), or two of the most common G6PD variants (ii) African (V68M - < 10% residual enzymatic activity) or (iii) Mediterranean (S188F - <1% residual enzymatic activity). We leveraged these mouse models to assess critical speed (CS), a functional measurement of the speed-to-duration relationship. To expand the molecular resolution of genetically associated adaptations to exercise, CS tests were combined with measurements of hemodynamics - including cardiac output, imaging (e.g., scanning electron microscopy of RBCs before and after exercise) and multi-omics (mass spectrometry-based metabolomics, lipidomics and proteomics) characterization of plasma, RBCs, and multiple tissue matrices including liver, heart, kidney, lung, spleen, and muscles (gastrocnemius and soleus).

RESULTS: First of all, we confirmed a significant decrease in G6PD activity in RBCs from mice carrying the African and Mediterranean variants ( Figure 1). Unexpectedly, G6PD deficient mice showed ~10% faster CS compared to mice carrying the canonical human G6PD (p<0.0001 and p<0.02, Figure 1). While scanning electron microscopy showed differences before and after exercise, with accumulation of reversibly modified morphologies across all groups after the run, no significant differences were observed across groups. No significant changes in hemolysis markers were observed in plasma samples after exercise, neither at the metabolic (e.g., heme catabolites) or proteomics level (e.g., hemoglobin chains, RBC-derived proteins). Within hemodynamics, the hG6PD _Med- mice are showing a higher efficiency in their cardiac function through cardiac function (p<0.005 - Figure 1). Pathway analysis of metabolomics data generated from RBC and plasma fractions showed significant decreases in arginine metabolism, increased glycolysis, and elevated Krebs cycle intermediates, in addition to expected differences in PPP. Specifically, in hG6PD _A- RBC acylcarnitines accumulate after running. In hG6PD _Med-, there are clear left ventricular differences highlighted in energy metabolism ( Figure 1). Multiomics analyses identified five major tissue clusters between heart, muscle, liver, kidney, and spleen, with G6PD function having the most significant metabolic impact on liver and soleus. RBC proteomics results highlighted deceased ferroportin (key iron transporter) in both the deficient groups, with markedly significant changes in the hG6PD _Med- (p<0.0001).

CONCLUSION: Novel humanized G6PD deficient mice maintain faster CS compared to mice. Omics approaches afford the opportunity to identify metabolic correlates to measurements of performance in relevant animal models, such as the CS model, which offer expanded mechanistic understanding of molecular bases of CS and the associated impact of G6PD deficiency. Our results not only challenge the assumption that exercise-induced oxidant stress predisposes the ~500M individuals with G6PD deficiency worldwide to increased hemolytic propensity, but rather indicate a metabolic advantage of functional polymorphisms to G6PD with respect to exercise performance, an observation that could contribute explaining the prevalence of such mutation in human populations.