The ketogenic diet (KD) has been utilized as a dietary therapeutic for nearly a century. One experimental model particularly responsive to the KD is the BTBR^T+tf/j (BTBR) mouse, which displays phenotypic characteristics of autism spectrum disorder (ASD) and insulin resistance. Recently, the study of impaired mitochondrial function has become a focal point of research investigating the pathophysiology of ASD. As highly dynamic organelles, mitochondria undergo constant fluctuations in morphology, biogenesis, and quality control in order to maintain cellular homeostasis. An important modifier of mitochondrial dynamics is energy availability. Therefore, the aim of this study was to examine the impact of a KD on mitochondrial dynamics in the liver and brain (prefrontal cortex) of the BTBR mouse model of ASD. Juvenile male C57Bl/6 (B6) and BTBR mice were age-matched to 5 weeks of age before being fed standard chow (CD, 13% kcal fat) or a KD (75% kcal fat) for 10-14 days. Analysis of brain tissue identified differences in mitochondrial gene expression but no correlation with protein levels. Unlike in the brain, KD led to decreased levels of mitochondrial proteins in the liver, despite increased gene expression. Consistent with decreased mitochondrial proteins, we also observed decreased mtDNA for all mice on the KD, demonstrating that the KD reduces the total amount of mitochondria in the liver. In order to explain the discrepancy between protein levels and gene expression, we investigated whether mitochondrial turnover via mitophagy was increased. To this end, we examined expression levels of the mitophagy regulator BNIP3 (BCL2/adenovirus E1B 19 kd-interacting protein 3). BNIP3 gene and protein expression were significantly elevated in liver of KD animals (p < 0.05), indicating the potential activation of mitophagy. Therefore, consumption of a KD exerts highly tissue-specific effects, ultimately increasing mitochondrial turnover in the liver, while gene and protein expression in the brain remaining tightly regulated.

Autism spectrum disorder (ASD) is a highly prevalent neurodevelopmental disorder that exhibits a common set of behavioral and cognitive impairments. Although the etiology of ASD remains unclear, mitochondrial dysfunction has recently emerged as a possible causative factor underlying ASD. The ketogenic diet (KD) is a high-fat, low-carbohydrate diet that augments mitochondrial function, and has been shown to reduce autistic behaviors in both humans and in rodent models of ASD. The aim of the current study was to examine mitochondrial bioenergetics in the BTBR mouse model of ASD and to determine whether the KD improves mitochondrial function. We also investigated changes in mitochondrial morphology, which can directly influence mitochondrial function. We found that BTBR mice had altered mitochondrial function and exhibited smaller more fragmented mitochondria compared to C57BL/6J controls, and that supplementation with the KD improved both mitochondrial function and morphology. We also identified activating phosphorylation of two fission proteins, pDRP1S616 and pMFFS146, in BTBR mice, consistent with the increased mitochondrial fragmentation that we observed. Intriguingly, we found that the KD decreased pDRP1S616 levels in BTBR mice, likely contributing to the restoration of mitochondrial morphology. Overall, these data suggest that impaired mitochondrial bioenergetics and mitochondrial fragmentation may contribute to the etiology of ASD and that these alterations can be reversed with KD treatment.

Although the functional consequences of mitochondrial DNA ( mtDNA) genetic backgrounds (haplotypes, haplogroups) have been demonstrated by both disease association studies and cell culture experiments, it is not clear which of the mutations within the haplogroup carry functional implications and which are "evolutionary silent hitchhikers''. We set forth to study the functionality of haplogroup-defining mutations within the mtDNA transcription/replication regulatory region by in vitro transcription, hypothesizing that haplogroup-defining mutations occurring within regulatory motifs of mtDNA could affect these processes. We thus screened >2500 complete human mtDNAs representing all major populations worldwide for natural variation in experimentally established protein binding sites and regulatory regions comprising a total of 241 bp in each mtDNA. Our screen revealed 77/241 sites showing point mutations that could be divided into non-fixed (57/77, 74%) and haplogroup/sub-haplogroup-defining changes (i.e., population fixed changes, 20/77, 26%). The variant defining Caucasian haplogroup J (C295T) increased the binding of TFAM (Electro Mobility Shift Assay) and the capacity of in vitro L-strand transcription, especially of a shorter transcript that maps immediately upstream of conserved sequence block 1 (CSB1), a region associated with RNA priming of mtDNA replication. Consistent with this finding, cybrids (i.e., cells sharing the same nuclear genetic background but differing in their mtDNA backgrounds) harboring haplogroup J mtDNA had a>2 fold increase in mtDNA copy number, as compared to cybrids containing haplogroup H, with no apparent differences in steady state levels of mtDNA-encoded transcripts. Hence, a haplogroup J regulatory region mutation affects mtDNA replication or stability, which may partially account for the phenotypic impact of this haplogroup. Our analysis thus demonstrates, for the first time, the functional impact of particular mtDNA haplogroup-defining control region mutations, paving the path towards assessing the functionality of both fixed and un-fixed genetic variants in the mitochondrial genome.