Mitochondrial fragmentation drives selective removal of deleterious mtDNA in the germline

https://www.nature.com/articles/s41586-019-1213-4

Toby Lieber, Swathi P. Jeedigunta, Jonathan M. Palozzi, Ruth Lehmann & Thomas R. Hurd

• The authors generate mutated fruit flies by transfering mitochondria from a wild-type strain of Drosophila yakuba into a strain of Drosophila melanogaster in which the mtDNA contain a temperature-sensitive mutation in Complex IV of the electron transport chain.
• The authors designed fluorescent probes to specifically bind to the D-loop of either D. yakuba or D. melanogaster, allowing them to visualise heteroplasmy.
• At 18C, the point mutation does not affect Complex IV activity, whereas at the inhibitory temperature of 29C, Complex IV activity is greatly reduced and is selected against 
• Selection first manifests during oogenesis, where a reduction in mitofusins causes fragmentation of the mitochondrial network
• The authors identify the proteins Atg1 and BNIP3  as necessary for the selective removal of mitochondria with mutated mtDNAs
• A reduction in Atg1 or BNIP3 decreases the amount of wild-type mtDNA, suggesting a link between mitochondrial degradation and replication
• At the restrictive temperature, selection occured in the germline but not in the somatic cells which surround the germline in the ovariole, and was largely absent in the male germ line (possibly because only female mtDNA is inherited)
• Inhibiting cell death through over-expression of the cell-death inhibitor p35 did not block selection
• Expression of the alternative oxidase protein (AOX) bypasses the function of complex IV and partially blocked selection, suggesting that the selection process senses defects in oxidative phosphorylation
• The authors observed greater fragmentation in the germline mitochondria relative to the soma. Using photoactivatable GFP, the authors show that mitochondrial contents rarely pass from one mitochondrion to another, suggesting that the purpose of fragmentation is to reduce complementation so that the genotype of individual mitochondria may be sensed through their phenotype.
• Reducing Mitofusin expression in somatic cells also induced selection
• The authors inhibited the protein IF1, to allow ATP synthase to run in reverse and maintain mitochondrial membrane potential by burning ATP. In doing so, the authors did not observe statistically significant selection, which may suggest that membrane potential sensing is the mechanism by which mitochondria are selected.
• Expression of a dominant-negative form of ATP synthase caused a reduction in mtDNA copy number of both mutants and wild-types