The purpose of this project is to determine the role that mitochondria have in the regulation of enzymes responsible for maintenance of the epigenome. The role of mitochondria in generating ATP and reactive oxygen species (ROS) is well recognized. However, less appreciated is the fact that these organelles are also involved in various biochemical pathways in the cells that give rise to a diverse range of metabolic products, including co-factors of proteins that epigenetically regulate the nuclear genome. For instance, mitochondria participate in the metabolism of S-adenosyl-methionine (SAM), which is the substrate used by DNA and histone methyltransferases to methylate CpG dinucleotides and histones, respectively, in the nucleus. Likewise, the production of acetyl-CoA and NAD+ occurs primarily in mitochondria, and these are co-factors of histone acetyltransferases (HATs) and deacetylases (HDACs), respectively, to modify histones. ATP is used by various protein kinases to phosphorylate substrates, including histones, which can change the composition of nucleosomes. Alpha-ketoglutarate, a metabolite from the tricarboxylic acid (TCA) cycle is a co-factor for the Ten-Eleven Translocation (TET) family of hydroxylases involved in hydroxyl-methylation of cytosines. Finally, mitochondrial-generated ROS can inhibit the jumonji (Jmj) demethylases leading to global histone hypermethylation. As modulation of the epigenome regulates gene expression, it follows that environmental agents that target the mitochondria may alter the regulation of gene expression by changing mitochondrial metabolism. Existing evidence indicates that mitochondrial dysfunction can lead to altered DNA methylation patterns in nuclear DNA and hyper-methylation of histones. Mitochondrial impairment can also affect gene expression. However, it still needs to be established whether epigenetic-driven changes in gene expression in the nucleus are a consequence of environmentally-mediated changes in mitochondrial function. In order to determine whether environmental agents that target mitochondria also impart their effects through alteration of the epigenome and gene expression, we first characterized whether changes in mitochondrial function results in impacts to the epigenome in the nucleus. To this end we have been using two genetic-based cell culture models of mitochondrial dysfunction. These systems rely on chronic loss of mitochondrial DNA based on ethidium bromide treatment of an osteosarcoma cell line (143B) or on acute loss of mtDNA in HEK293 cells. In the latter, which we call the HEK293DN system, a dominant negative mitochondrial polymerase (polG) is ectopically expressed in an inducible fashion, which results in a progressive loss of mtDNA over a period of 9 days. The cells without mtDNA are termed rho0 and can survive under cell culture conditions that support ATP generation through glycolysis. With the acute inducible system we analyzed 4 time points of acute mitochondrial dysfunction (days 0, 3, 6 and 9). The cells are effectively rho0 at day 9. This allows evaluating progressive effects on the epigenome during the time-dependent loss of mtDNA. The chronic 143B system relies on a comparison of control cells with their full complement of mtDNA (rho+) with cells that are chronically devoid of mtDNA (rho0) or with cells that harbor a mtDNA genome that contains an inactivating mutation in cytb, which was originally derived from a Parkinsons patient (rho-). These rho- cells have a defect in complex III of the electron transport chain. In previous years we characterized several biochemical parameters in the 143B and HEK293DN systems, including levels of different metabolites (ATP, ROS, acetyl-CoA, NAD and TCA intermediates) in addition to epigenetic markers (enzyme activities and bulk histone changes). The picture that emerged in both systems is that mtDNA depletion, whether acute or chronic, results in a decrease in the levels of acetyl-CoA, which is associated with diminished HAT activity and histone acetylation. In the HEK293DN model, we showed that rescue of the TCA cycle by re-establishing electron transport in the absence of mtDNA allowed for restoration of citrate levels and normalization of histone acetylation marks. In the 143B system, our data showed that pharmacological supplementation of the mitochondrial acetyl-CoA pool is associated with restoration of HAT activity in the rho0 cells while inhibition of the mitochondrial metabolism of acetyl-CoA is associated with a decrease in HAT activity. Over the past fiscal year we focused our work on better understanding how changes in mitochondrial function impact the transcriptome and epigenome in these two cell culture systems. Using several NextGen and genomic approaches (RNA-seq, ChIP-seq of histone marks, methylation bead arrays and metabolomics), we found that mitochondrial dysfunction results in specific changes to the transcriptome and epigenome that are based on the type of mitochondrial dysfunction. Mechanistically, we found that mitochondrial-derived acetyl-CoA is necessary and required to maintain histone H3 acetylation marks (H3K9, 18 and 27). Manipulating the levels of mitochondrial acetyl-CoA results in alterations in bulk histone acetyltransferase (HAT) activity and in the levels of acetylation of histone H3. We did not identify changes in HDAC activity. In the cells with mtDNA harboring the cytb mutation, histones were hyper-acetylated despite the pronounced respiratory dysfunction. We found that in this model, cells compensate for the loss of oxidative TCA-produced acetyl-CoA by glutamine reductive metabolism. No changes in H4 or in histone methylation marks were observed in either model. Both mtDNA depletion and mutation models cause changes in genome-wide DNA methylation, although the behavior is distinct between the models: the nuclear DNA is hypermethylated in mtDNA depleted cells, while in the cytb mutants it is hypomethylated when compared to controls. Analysis of gene expression in both models indicated that modulation of one carbon (1C) metabolism by folate and serine/glycine, which feed into the methionine cycle regulating the levels of S-adenosyl-methionine (SAM), is impacted by mitochondrial dysfunction. SAM is the metabolite used for methylation of DNA and proteins. Serine and 1C metabolism, while activated in the mtDNA depletion model, is inhibited in the cytb mutant cells. Over this past fiscal year we have also interrogated whether mitochondrial-driven dysfunction changes the epigenome in vivo. We exposed the viable yellow agouti mouse (Avy) to rotenone, a mitochondrial complex I inhibitor. Rotenone is a pesticide known to contaminate the environment that has been linked to the induction of Parkinsons disease both in humans and in animal models. This pesticide inhibits mitochondrial function leading to two different outcomes: (i) it slows down the TCA cycle by blocking the first site of electron flow in the ETC, and (ii) it increases mitochondrial ROS at this site by backing up electron flow to molecular oxygen. The Avy animals have been used in many studies as an epigenetic reporter; they carry a mutation in the promoter of agouti that makes it possible to examine the epigenetic status of the gene by examining the coat color of the mice. For example, when the promoter in the Avy allele is methylated, the animals have a normal agouti coat color (called pseudoagouti). On the other extreme, when the promoter is unmethylated, the coat color of the animals is completely yellow. Our data show that offspring of dams exposed to rotenone throughout pregnancy and lactation have an increased frequency in the yellow category, indicating that complex I dysfunction during development leads to increased hypomethylation of the reporter.
| Lozoya, Oswaldo A; Martinez-Reyes, Inmaculada; Wang, Tianyuan et al. (2018) Mitochondrial nicotinamide adenine dinucleotide reduced (NADH) oxidation links the tricarboxylic acid (TCA) cycle with methionine metabolism and nuclear DNA methylation. PLoS Biol 16:e2005707 |
| Lozoya, Oswaldo A; Santos, Janine H; Woychik, Richard P (2018) A Leveraged Signal-to-Noise Ratio (LSTNR) Method to Extract Differentially Expressed Genes and Multivariate Patterns of Expression From Noisy and Low-Replication RNAseq Data. Front Genet 9:176 |
| MartÃnez-Reyes, Inmaculada; Diebold, Lauren P; Kong, Hyewon et al. (2016) TCA Cycle and Mitochondrial Membrane Potential Are Necessary for Diverse Biological Functions. Mol Cell 61:199-209 |
| Carlin, Danielle J; Rider, Cynthia V; Woychik, Rick et al. (2013) Unraveling the health effects of environmental mixtures: an NIEHS priority. Environ Health Perspect 121:A6-8 |
