Multiple timescales in bacterial growth homeostasis

Abstract

Microbial growth and division are fundamental processes relevant to many areas of life science. Of particular interest are homeostasis mechanisms, which buffer growth and division from accumulating random fluctuations. This classic problem has seen recent advance in both theory and experiments, but still much remains unclear and debatable. Because there are many coupled processes inside the cell, it is not a-priori clear which variables are under regulation and which are stabilize by their coupling to regulated variables. Here, we address this question by distentangling homeostatic set-points -- estimated as temporal averages across individual lineages -- from fluctuations around them. Applying variance decomposition to individually trapped bacteria (mother machine data), we find that phenotypic variables estimated from cell-size measurements (inter-division time, growth rate, added size and more), exhibit a range of different behaviors. Some have flexible set-points that vary significantly between lineages - "sloppy" variables, while others are tightly fixed - "stiff" variables. Analyzing pairs of trapped lineages (sisters machine data), reveals that the primary source of sloppiness is a high sensitivity to the trap environment, while lineage-specific factors contribute only a small fraction. We find strong long-term correlations between sloppy set-points, and identify a control manifold in the space of growth and division variables. Cell size is a sloppy variable, whose set-point is uncoupled to growth variables; such correlations appear only on the short, single-cycle timescale. We discuss these results in light of recent models of cell size control, and point to new avenues of research.

Universality of phenotypic distributions in bacteria

“…Previous work has argued for Gaussian-like division time distributions that collapse by scaling (24,28,29), and can be reconstructed by models with deterministic exponential accumulation rates (25,39). While this framework holds for some bacterial data-sets, others exhibit a much broader variability in accumulation rates (11,(18)(19)(20)37). Our analysis suggests that different experiments, possibly due to slightly different conditions or culture details, span a range of behaviors from highly uniform to highly varying accumulation rates.…”