J. William O. Ballard, Wen C. Aw, Michael R. Garvin, Yifang Hu, Antón Vila-Sanjurjo, Torsten Thomas, Neil A. Youngson, Sonia Bustamante, Shaun Nielsen, Russell Pickford, Gordon K. Smyth, Samuel G. Towarnicki, and Richard G. Melvin
Diet may be modified seasonally or by biogeographic, demographic or cultural shifts. It can differentially influence mitochondrial bioenergetics, retrograde signalling to the nuclear genome, and anterograde signalling to mitochondria. All these interactions have the potential to alter the frequencies of mtDNA haplotypes (mitotypes) in nature and may impact human health. In a model laboratory system, we fed four diets varying in Protein: Carbohydrate (P:C) ratio (1:2, 1:4, 1:8 and 1:16 P:C) to four homoplasmic Drosophila melanogaster mitotypes (nuclear genome standardised) and assayed their frequency in population cages. When fed a high protein 1:2 P:C diet, the frequency of flies harbouring Alstonville mtDNA increased. In contrast, when fed the high carbohydrate 1:16 P:C food the incidence of flies harbouring Dahomey mtDNA increased. This result, driven by differences in larval development, was generalisable to the replacement of the laboratory diet with fruits having high and low P:C ratios, perturbation of the nuclear genome and changes to the microbiome. Structural modelling and cellular assays suggested a V161L mutation in the ND4 subunit of complex I of Dahomey mtDNA was mildly deleterious, reduced mitochondrial functions, increased oxidative stress and resulted in an increase in larval development time on the 1:2 P:C diet. The 1:16 P:C diet triggered a cascade of changes in both mitotypes. In Dahomey larvae, increased feeding fuelled increased β-oxidation and the partial bypass of the complex I mutation. Conversely, Alstonville larvae upregulated genes involved with oxidative phosphorylation, increased glycogen metabolism and they were more physically active. We hypothesise that the increased physical activity diverted energy from growth and cell division and thereby slowed development. These data further question the use of mtDNA as an assumed neutral marker in evolutionary and population genetic studies. Moreover, if humans respond similarly, we posit that individuals with specific mtDNA variations may differentially metabolise carbohydrates, which has implications for a variety of diseases including cardiovascular disease, obesity, and perhaps Parkinson’s Disease., Author summary The detection and quantitation of mtDNA polymorphisms in populations and across whole habitats continues to be used as a central investigatory tool in evolutionary genetics. But, the approach is laden with assumptions about selection that are rarely examined. We present a series of studies that traverse the genotype to the phenotype. The studies were designed to experimentally test the interaction between diet and mitotype in Drosophila flies and provide a mechanism by which selection occurs. We start with population cage studies that include four laboratory diets and four mitotypes. We then directly compete two mitotypes (Alstonville and Dahomey) on a high protein and a high carbohydrate diet and show a flip in their relative fitness that is driven by differences in larval development. Next, we identify a single naturally-occurring point mutation, which drives the cage results. We track the ripple effects up to the level of the organelle (mitochondria), through the labyrinth of metabolic pathways and on to the phenotype. Notably, when flies were fed the high carbohydrate diet, energy metabolism was extensively remodelled in both mitotypes causing increased physical activity in Alstonville flies. These data invite an extensive experimental re-evaluation of the assumption that mtDNA inescapably evolves in a manner consistent with a strictly neutral equilibrium model. It also motivates investigation of genotype-specific dietary manipulation as an integrative treatment of human disorders involving mitochondrial metabolism and offers the potential for future therapeutic strategies.