Severe osmotic compression triggers a slowdown of intracellular signaling, which can be explained by molecular crowding

Abstract

Regulation of the cellular volume is fundamental for cell survival and function. Deviations from equilibrium trigger dedicated signaling and transcriptional responses that mediate water homeostasis and volume recovery. Cells are densely packed with proteins, and molecular crowding may play an important role in cellular processes. Indeed, increasing molecular crowding has been shown to modify the kinetics of biochemical reactions in vitro; however, the effects of molecular crowding in living cells are mostly unexplored. Here, we report that, in yeast, a sudden reduction in cellular volume, induced by severe osmotic stress, slows down the dynamics of several signaling cascades, including the stress-response pathways required for osmotic adaptation. We show that increasing osmotic compression decreases protein mobility and can eventually lead to a dramatic stalling of several unrelated signaling and cellular processes. The rate of these cellular processes decreased exponentially with protein density when approaching stalling osmotic compression. This suggests that, under compression, the cytoplasm behaves as a soft colloid undergoing a glass transition. Our results shed light on the physical mechanisms that force cells to cope with volume fluctuations to maintain an optimal protein density compatible with cellular functions.

Fig. 1.

Dynamics of the HOG cascade. (A) HOG cascade. (B) Illustration of the sealed gasket chambers used to experimentally regulate the cells’ environment. (C) Osmotic stimulation (1 M sorbitol) induced immediate cell shrinkage (blue curve) followed by nuclear accumulation of Hog1p (red curve). Data are represented as the mean of several cells ± 1 SD (n > 20). Cells progressively recovered their size and switched off the HOG cascade through negative-feedback loops. (D) Nuclear accumulation of Hog1p was rapid and robust after a 1 M sorbitol osmotic stress. (E) Phosphorylation levels of Hog1p after a 1 M sorbitol osmotic stress followed the same pattern with a rapid induction and then a slow decay. Typically, within 30 min, nuclear Hog1p was no longer observed, and Hog1p returned to basal phosphorylation levels. Phosphoglycerate kinase (PGK) was used as a positive control.

Fig. 2.

Severe osmotic stress reduces the kinetics of HOG activation and nuclear translocation. (A) Time-lapse imaging of the nuclear localization of Hog1p in cells exposed to 1 M (red) and 1.75 M (blue) sorbitol. (B) The rate of Hog1p nuclear translocation was reduced by severe osmotic shock (red, 1 M sorbitol; blue, 1.75 M sorbitol). Maximal nuclear accumulation was observed ∼45–60 min after 1.75 M sorbitol treatment (E). (C) Cell volume decreased with increasing concentrations of sorbitol, down to a plateau at around ∼40% of the initial volume. (D) Western blots showing Hog1p phosphorylation over time and under increasing osmotic stress. Phosphorylation of Hog1p after 2 min at 1 M sorbitol is used as a positive control (C+). (E) Time at which the Hog1p nuclear concentration and phosphorylation levels reached a maximum increased with the strength of the osmotic shock. Above 2.2 M sorbitol (blue dashed), no cells displayed nuclear accumulation of Hog1p (Fig. S2). (F) Same data as in E, shown as a function of relative cell volume compression. The blue line is a fit consistent with what is expected for a soft colloid near a glassy transition, , where T₀ = 0.065 and α = 3.6.

Fig. 3.

Diffusion of Hog1p-GFP in the cytoplasm is decreased by osmotic compression. (A) FRAP experiments on the HOG1-GFP pbs2Δ strain in SC medium (red curve) and after 2 M sorbitol stress (blue curve). (B) FRAP images of HOG1-GFP pbs2Δ cells in SC medium before bleaching (Left), immediately after bleaching (Center), and 5 s later (Right). The bleached area is indicated by the black arrow, and its recovery is indicated by the orange arrow. (C) FRAP images of a HOG1-GFP pbs2Δ cell in a severe osmotic environment (2 M sorbitol). In contrast to B, the bleached area did not recover after 5 s. (D) FLIP time series for a HOG1-GFP pbs2Δ cell in SC medium. The same spot was continuously bleached (black arrow), and total cell fluorescence rapidly decreased compared with the unbleached neighboring cell. (E) After severe osmotic compression, the cell maintained some cytoplasmic fluorescence over a long time (orange arrow), indicating that the diffusion of fluorescent protein was very limited. Pictures are color-coded to improve visualization (blue for low intensity and red for high intensity).

Fig. 4.

Several signaling cascades are delayed when the cell volume is suddenly decreased by osmotic compression. (A) Nuclear localization of Msn2p in response to gentle osmotic stress (0.45 M sorbitol, red) and severe osmotic stress (1.75 M sorbitol, blue). (B) Nuclear localization of Crz1p in response to calcium shock (200 mM, green), gentle osmotic stress (0.45 M sorbitol, red), severe osmotic stress (1.75 M sorbitol, blue), and combined calcium shock and severe volume reduction by osmotic stress (200 mM Ca²⁺ plus 1.4 M sorbitol, black). (C) Nuclear localization of Yap1p in response to oxidative stress (300 µM H₂O₂, green), oxidative stress and gentle osmotic compression (300 µM H₂O₂ plus 0.45 M sorbitol, red), and combined oxidative stress and severe osmotic stress (H₂O₂ plus 1.4 M sorbitol, black). (D) Nuclear localization of Mig1p in response to glucose (20 g/L, green), combined glucose and gentle osmotic stress (glucose plus 0.45 M sorbitol, red), and combined glucose and severe osmotic stress (glucose plus 1.4 M sorbitol, black). Cells were precultured in galactose without glucose. For A–D, data are represented as the mean of many cells (n > 20) ± 1 SD. (E–H) Time-lapse images of the nuclear localization of Msn2p (E), Yap1p (F), Crz1p (G), and Mig1p (H). “S” indicates sorbitol, and the concentrations used and color code are the same as for A–D. (I and J) Measurement of the time to reach maximum nuclear localization of Msn2p (sorbitol), Yap1p (H₂O₂), Crz1p (Ca²⁺), Mig1p (glucose), and Hog1p (sorbitol) as a function of either the additional sorbitol concentration used (I) or the relative cell volume (J). The curves fit the function that models the behavior of a soft colloid. For Msn2p, T₀ = 0.035 and α = 3.9; for Crz1p, T₀ = 0.096 and α = 4.4; for Yap1p, T₀ = 0.077 and α = 4.3; for Mig1p, T₀ = 0.043 and α = 4.8.

Fig. 5.

Signaling-cascade dynamics are immediately restored upon recovery of the cell volume. (A) Principle of signaling activation by volume recovery. Cells are stressed and, at the same time, osmotically compressed by severe osmotic stress (3 M sorbitol). The osmotic compression was then relaxed by a return to a lower osmotic stress conditions, and we investigated whether the cells could detect and respond to the chemical stress that was still present in the cells’ environment. (B–F) Time courses of the nuclear localization of Hog1p (B), Msn2p (C), Crz1p (D), Mig1p (E), and Yap1p (F) under these experimental conditions. As soon as the volume of the cell was restored, the studied signaling cascades could function normally, as indicated by nuclear localization of their respective transcription factors.