When a cell divides, all components in the cell need to go into either one of the daughter cells. The cell has developed mechanisms to make sure that some components, for example chromosomes, are correctly segregated. How does this work for mitochondria? Does the cell have active mechanisms to push equal amounts of mitochondria in each daughter cell, or do the mitochondria randomly move into one of the daughters?
In this paper, they try to answer these questions in fission yeast S. Pombe.
It was shown before that mitochondria are pushed to both of the cell poles before cell division, which would suggest that this is a mechanism of segregating the mitochondria.
However, in this paper they show that in the last 15% of the cell cycle, mitochondria spatially re-equilibrate themselves throughout the cell. When the cell divides, the mitochondrial volume in a daughter cell tracks the cytoplasmic volume of that daughter cell. For example, if the two daughters get 60% and 40% of the cytoplasmic volume, they will on average also get 60% and 40% of the mitochondrial volume.
The same is true of mitochondrial nucleoids, which contain the mitochondrial DNA; they too segregate in proportion to the cytoplasmic volume.
However, the errors made in nucleoid segregation are smaller than you would expect from passive mechanisms. Using the example from above, passively one would expect that each nucleoid has a probability of 0.6 of ending up in the larger daughter (the one with 60% of the cytoplasm) and a probability of 0.4 of ending up in the other daughter. This is called binomial partitioning and has a certain error size associated with it. The actual errors that are made in S. Pombe (i.e. the deviations from perfect partitioning) are smaller than binomial errors.
A model that would explain these sub-binomial nucleoid segregation errors is to assume that the nucleoids are regularly spaced out within the mitochondria. It is not known how this regular spacing is accomplished by the cell.
The authors also find that the number of nucleoids that are produced from beginning to end of the cell cycle does not depend on the initial amount that was present. This suggests that S. Pombe does not not use feedback control to produce its nucleoids (or other mitochondrial proteins). Feedback control is energetically expensive. Rather, nucleoids are randomly added throughout the cell cycle, following a Poisson distribution.