Recent publication:(Sandler,Pearl-Mizrahi. et al, Nature 2015)
Lineage correlations of single cell division time as a probe of
Oded Sandler*,Sivan Pearl Mizrahi*,Noga Weiss,Oded Agam,Itamar Simon & Nathalie Q. Balaban
We all see ourselves as a consequence of what we have inherited and absorbed from our ancestors.
In a population of dividing cells in which each daughter cell inherits a similar genetic content and experiences
similar environment, this principle should be even more pronounced, such that along a lineage of cells,
we expect to find a boring uniformity in most traits.
However, it has been observed that the cell cycle duration (the time from cell's birth to its
division), is far from being similar in cell lineages. When a cell divides, it gives rise to two
daughter cells with similar contents and similar cell cycle durations. However the duration of
daughters' cell cycles hardly resemble that of their mother cell, suggesting that the trait of cell cycle
duration is reset in the transition from one generation to the next.
Using state-of-the-art microscopy and highly controlled experimental settings, we were able to
follow hundreds of cells' lineages. Our measurements show that the cell cycle
duration is highly similar between daughter cells but seems totally unrelated to the duration of the
mother cell. Surprisingly, cousin cells have similar cycle durations. This finding poses a paradox:
if the cards of cell cycle duration are shuffled randomly at the transition from one generation to
the next, how could cousin cells, whose relation stems from a common grand-mother, possess
similar cycle durations? Could someone distribute a card pack in an orderly fashion and at the
same time make it look like it was blown by the wind? To put it in mathematical terms, could the
variability in the inheritance of cell cycle durations be set by a deterministic process which only
appears random? To answer this question, we made use of the Grassberger Proccacia algorithm
and applied it to series of cell lineages, which revealed that the pattern of the cell cycle durations in
a lineage is, in fact, deterministic.
To resolve the paradox of deterministic inheritance, without mother-daughter correlations, we
developped a toy model postulating the existence of an inherited factor, oscillating independently
of the cell cycle. The model assumes that a cell cycle’s duration is set by the level of this factor at birth; a
cell which is born at a time when this factor level is low/high, will have a relatively short/long cell
cycle. When the cell divides, the factor level is inherited by its daughters, such that
they will have similar cell cycle durations. Consequently, both daughter cells will divide at
similar times, with similar levels of this factor, giving birth to cousin cells with similar cell cycle
duration and so on…. Yet, since the inherited factor is oscillating in time, daughter cells will not
necessarily resemble their mothers.
Taken together, we were able to reveal surprising correlations in cycle durations of cousin cells,
in the absence of mother-daughter correlations. We explain this seemingly paradoxical finding by
a non-linear deterministic process and corroborate it by a mathematical framework
which reproduces our findings. Our work may pave the way to similar analyses in other biological processes
and systematically uncover hidden deterministic relations. Specifically, if the variability in cell cycle
duration is not the inevitable consequence of randomness, but rather consists of deterministic
components, it is tempting to think that somewhere lay the keys to control it.
(see also accompanying News & Views http://www.nature.com/nature/journal/v519/n7544/full/nature14210.html)