Robustness and Universality in Organelle Size Control

One of the grand challenges in cellular biophysics is understanding the precision with which cells assemble and maintain subcellular structures. Organelle sizes, for example, must be flexible enough to allow cells to grow or shrink them as environments demand yet be maintained within homeostatic limits. Despite identification of molecular factors that regulate organelle sizes we lack insight into the quantitative principles underlying organelle size control. Here we show experimentally that cells can robustly control average fluctuations in organelle size. By demonstrating that organelle sizes obey a universal scaling relationship we predict theoretically, our framework suggests that organelles grow in random bursts from a limiting pool of building blocks. Burstlike growth provides a general biophysical mechanism by which cells can maintain on average reliable yet plastic organelle sizes.

Figure 1

(a) Schematic of biophysical processes in model of organelle biogenesis. (b) Gillespie simulation results of organelle size versus number for 3 classes of organelle size models. (c)–(f) Average organelle size versus organelle number. (g)–(j) Average intracellular CV versus organelle size. Organelle sizes are normalized to largest organelle size of a given type. Gray data points indicate errors in CV propagated from estimated errors in organelle size and gray region is the average value of the gray CV error data points. Star indicates measured CV of diffraction limited 100 nm beads.

Figure 2

(a) Schematic of burstlike model of organelle growth. (b) Distributions of experimentally measured late Golgi, lipid droplet, and peroxisome sizes. Size of 500 voxels corresponds to $0.4\mathrm{\mu }{\mathrm{m}}^3$. (c) Histograms of late Golgi, (d) lipid droplet, and (e) peroxisome sizes rescaled by their experimentally derived organelle-specific burst sizes. (f) Histogram of organelle size distributions rescaled by both experimentally derived organelle-specific burst sizes and burst frequencies; black curve is the theoretically predicted scaling relationship from the model.

Figure 3

(a) Total organelle size $V$ as a function of the number of organelles in the limits where the pool is exhausted (blue line) versus nonexhausted (maroon line). (b) Average intracellular CV of organelle sizes versus average organelle sizes in 500 simulated cells from the burstlike model, when pool is exhausted versus nonexhausted. Total organelle size (c) and (d) average intracellular size CV of late Golgi from populations of wild-type (maroon) and $\mathrm{\Delta }ARF1$ (purple) cells. (e) Total organelle size and (f) average intracellular size CV of lipid droplets marked with Erg6-mRFP from populations of wild-type cells grown in glucose (maroon) and in medium with 0.2 percent oleic acid (blue).

Figure 4

(a) Average Golgi size versus number of Golgi from images of single induced pluripotent stem cells taken from the Allen Cell Atlas. (b) Average Golgi intracellular size CV versus average Golgi size. (c) Average mitochondrial size versus number of mitochondria from images of single induced pluripotent stem cells taken from the Allen Cell Atlas. (d) Average mitochondrial intracellular size CV versus average mitochondrial size.