Inmaculada Martínez-Reyes, Lauren P. Diebold, Hyewon Kong, Michael Schieber, He Huang, Christopher T. Hensley, Manan M. Mehta, Tianyuan Wang, Janine H. Santos, Richard Woychik, Eric Dufour, Johannes N. Spelbrink, Samuel E. Weinberg, Yingming Zhao, Ralph J. DeBerardinis, Navdeep S. Chandel
In this study, the authors investigate the effect of eliminating mtDNA from cells, on metabolism. This is often achieved by incubation with the chemical ethidium bromide, which prevents the protein responsible for mtDNA replication (POLG) from working. However, ethidium bromide is toxic and has potentially off-target effects. The authors engineered cells which, upon exposure to an antibiotic, express a dominant-negative form of POLG (DN-POLG), thereby removing potential off-target effects. Within 6 days of induction of DN-POLG, mtDNA transcription had ceased.
Cells had a greatly reduced proliferation rate (although still viable), despite having access to glucose and pyruvate: metabolites required to drive glycolysis and the TCA cycle respectively. As discussed in  (see here) mitochondria oxidise NADH to NAD+, the substrate of glycolysis. The authors introduced two non-mammalian proteins (NDl1 and AOX), which together carry electrons in a similar manner to the electron transport chain, but do not pump protons across the inner mitochondrial membrane. In this way, NAD+ can be restored, without generating mitochondrial ATP from oxidative phosphorylation. The authors show that these proteins are sufficient to drive flux through the TCA cycle, and increase the NAD+/NADH ratio. However, the authors find that these cells still have an impaired proliferation rate (at odds with the findings in ?)
The authors investigated an alternative reason for the impaired proliferation of these cells: mitochondrial membrane potential (ΔΨm). Cells without mtDNA cannot pump protons to maintain ΔΨm, but ATP synthase is still able to hydrolyse ATP from glycolysis, to maintain ΔΨm. An endogenous inhibitor of this function is the protein ATPIF1. The authors knocked out this gene, and showed that cells without mtDNA are able to maintain their ΔΨm at wild-type levels. These ATPIF1-KO cells had a, statistically significant, partial recovery in proliferation rate.
Interestingly, treating the ATPIF1-KO cells with an antioxidant, which mops up ROS (canonically considered to be the bad-guy of mitochondrial physiology, see here), reduced the proliferation rate. This shows that ROS, as well as ΔΨm, are necessary for cells to proliferate.
 Wiley, Christopher D., et al. "Mitochondrial Dysfunction Induces Senescence with a Distinct Secretory Phenotype." Cell metabolism (2015).