Session 4: Mitochondria in metabolism

Neurons can utilize a variety of fuel sources for their bioenergetic, biosynthetic and redox needs, which can significantly affect neuronal activity and overall health. A remarkable example of this is the seizure protective effects of the ketogenic diet, which reduces glucose utilization and increases ketone body (KB) consumption in the brain. However, the stringent nature of the diet makes it difficult to apply and maintain as a line of therapy. As such, understanding and tapping into diet-independent mechanisms to reduce glucose metabolism and/or increase KB utilization can offer alternative strategies to produce seizure resistance and “reverse engineer” the ketogenic diet. We previously reported that modification of a single gene, Bad, is sufficient to trigger a glucose-to-KB fuel switch in neurons that manifests in diet-independent seizure resistance. In recent studies, we have used this genetic model combined with integrative metabolomics and proteomics analyses to uncover the minimal mechanism that is necessary and sufficient to control neuronal glucose versus KB fuel decision and attendant changes in neuronal excitability. I will discuss findings from these integrative analyses, which point to a previously unappreciated regulatory role for the oxidative pentose phosphate pathway in neuronal fuel patterns.  

Mitochondria have retained their own genome, but rely on the cell for the nucleotides required to replicate it. How mitochondria regulate their uptake and use of cytosolic nucleotides is little understood. Because invading microbes perturb cytosolic nucleotide metabolism which mitochondria are dependent, we reasoned that microbial infection would serve as a model to study the dynamics of mitochondrial nucleotide metabolism. In a CRISPR screen to identify host pathways that regulate the growth of the human parasite Toxoplasma gondii, we found that mitochondrial enzymes involved in nucleotide metabolism including Deoxythymidylate Kinase (DTYMK) restricted parasite replication, unlike those in the cytosol that promoted Toxoplasma growth. Genetic ablation of DTYMK promoted Toxoplasma growth whereas chemical inhibition of host de novo pyrimidine synthesis restricted Toxoplasma growth. Infection led to an increase in levels of mtDNA, but not their corresponding RNA or protein products. We pinpointed the changes in mtDNA levels to a host transcription factor that is activated by infection and required to restrict parasite growth. Our results support the possibility that mitochondria are actively storing nucleotides, thereby restricting their availability to invading microbes. 

Non-alcoholic steatohepatitis (NASH) is a global health concern without treatment. The challenge in finding effective therapies is due to the lack of good mouse models and the complexity of the disease, characterized by gene-environment interactions. We tested the susceptibility of 7 mouse strains to develop NASH. The severity of the clinical phenotypes observed varied widely across strains. PWK/PhJ mice were the most prone to develop NASH, while CAST/EiJ mice were completely resistant. Levels of transcripts and proteins present in the mitochondria as well as mitochondrial function were robustly reduced in the liver of PWK/PhJ mice, suggesting a central role of mitochondrial dysfunction in NASH progression. Importantly, the alterations in gene expression observed in PWK/PhJ mice were the closest to the human NASH. Our study exposes the limitations of using a single mouse genetic background in metabolic studies and describes a novel NASH mouse model that closely resembles the human disease.

Many human diseases are caused by mutations that perturb metabolism, particularly mitochondrial metabolism, and result in tissue dysfunction. Some metabolic perturbations result in disease by interrupting canonical metabolic functions, and others interfere with processes beyond the metabolic network, including cellular signaling and gene expression. Understanding these pathological states of metabolic perturbation may help us develop rational approaches to normalize metabolism and restore health. We study two types of diseases characterized by metabolic dysfunction: inborn errors of metabolism and cancer. I will discuss ongoing work related to inborn errors that capitalizes on the incredible metabolic heterogeneity within these diseases to probe the human metabolic network directly in patients, then uses experimental models to explore disease mechanisms and propose potential therapies. I will emphasize methods in metabolomics and stable isotope tracing that allow us observe metabolic phenotypes in intact systems relevant to physiology and disease, highlighting disorders with complex phenotypes affecting neurodevelopment.