Energetic costs of cellular and therapeutic control of stochastic mitochondrial DNA populations

https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1007023

Hanne Hoitzing, Payam A. Gammage, Lindsey Van Haute, Michal Minczuk, Iain G. Johnston, and Nick S. Jones


Background on mitochondrial DNA dynamics and control

Mitochondria have their own genomes (mtDNAs). These genomes can mutate upon division and at any one given time, mixture of normal (wildtype, w) and mutated (m) mtDNA can exist within a cell. Heteroplasmy is defined as the fraction of mutant mtDNA molecules.

The birth and death of mtDNAs is a stochastic process, their numbers fluctuating over time. Some kind of feedback control must be present, as mtDNA numbers in normal healthy cells tend to remain within certain bounds.

Treatments exist to reduce the load of mutant mtDNAs inside cells. For example, nucleases which are targeted to the specific sequence of a mutant mtDNA can be introduced in cells. They will bind to these mutant sequences and cut the (though off-target cutting of the wildtype genomes is a problem).

Thinking about controlling levels of mtDNA gives rise to various questions:

•  What exactly is this feedback control? What is the quantity that is being controlled (e.g. is it total mtDNA copy number, or is it the overall energy level)? 
• How does the type of control influence heteroplasmy levels? Does one type of control lead to faster mutant accumulation than another?
• How does the cell choose a particular feedback control? Does it do this randomly or does it minimize some 'cost function'?
• Can we somehow interfere with the cellular feedback control to reduce mutant loads?

 
Paper results

This paper investigates these questions a bit more closely.  Some of the main findings are:

• Many different forms of feedback control (e.g. linear, quadratic, etc..) can give rise to similar mtDNA dynamics and heteroplasmy dynamics.
• What makes all the difference, however, is which quantity is being controlled (rather than how it is controlled). Is it total copy number (w + m)? Is it only the number of wildtypes (w)? Is it some more general linear combination (w + 𝛿 m)?
• The more strongly one species is controlled, the more control is lost over the other
• A mitochondrial cost function is introduced, and it is shown that it can actually be more expensive for a cell to contain a mixture of mutant and wildtype molecules, rather than only mutants!
• A control based on energy levels seems to make more sense than blindly controlling total mtDNA copy number. This means that if mutants produce less energy, the quantity being controlled is (w + 𝛿 m) with 𝛿 < 1.
• Variance of mtDNA dynamics is important! An increase in variance in mutant and/or wildtype copy numbers (which will always occur over time) can lead to an increase in cost of maintaining a tissue
• Gene therapies specifically targeting mutant mtDNAs can successfully lower heteroplasmy levels, but this becomes hard when high tissue heteroplasmy levels are caused by only a small fraction of cells (i.e. a few cells have very high heteroplasmy levels and most cells are ok). Again, it's the mtDNA variance that's important!
• Long and weak gene therapies seem to reach lower overall heteroplasmy levels compared to short and strong therapies.