Live imaging of yeast cells grown in periodically fluctuating environments.

(a) Overview of the hyperosmotic response in yeast. Both glucose deprivation and osmotic stress lead to cell cycle arrest—through different molecular mechanisms. (b) Sketch of the microfluidic setup used to generate a time-varying environment and achieve timelapse imaging of yeast cells. Bright field and fluorescence images are captured every 6 minutes at 25 positions for 12-24 hours depending on the experiment. Nuclei expressing HTB2-mCherry fusion protein are segmented and tracked over time to compute the cell division rate as a function of time. (c-f) The four periodically varying environments used in this study. (c) Periodic osmotic stress: Cells are periodically exposed to hyperosmotic stress (1 M sorbitol) in a constant glucose environment (2% or 0.1%). (d) Periodic glucose deprivation: environment alternates between presence and absence of glucose. (e) In-phase stresses (IPS): periodic exposure to glucose in absence of hyperosmotic stress followed by glucose depletion with hyperosmotic stress (1 M sorbitol). (f) Alternating stress (AS): periodic exposure to glucose with hyperosmotic stress (1 M sorbitol), followed by glucose depletion without hyperosmotic stress. (c-f). Hatching represents stress, blue indicates presence of sorbitol; orange, presence of glucose.

The frequencies of osmostress and glucose availability impact cell division rates in opposite ways.

(a,b) Impact of the frequency of periodic osmotic stress (a) and glucose deprivation (b) on the average division rate. Each dot shows the mean division rate measured in different growth chambers of the microfluidic chip. Error bars are 95% confidence intervals of the mean. Colored dotted lines are Loess regressions obtained using a smoothing parameter of 0.66. Colored areas represent 95% confidence intervals of the regression estimates. Gray dotted lines show the average division rate in the absence of sorbitol (no osmotic stress) in 2% glucose (top line), or half this average division rate (bottom line). (c) Overlay of the Loess regressions shown in (a) and (b) at 2% glucose. The frequency and division rate at which the two regression curves intersect are highlighted by vertical and horizontal black dotted lines.

Cell division rate depends on the phasing of hyperosmotic stress and glucose availability.

(a,b) Division rates measured in four fluctuating conditions with a period of 24 minutes and a glucose concentration of 2% (2 g/l). The four conditions are periodic glucose deprivation, periodic osmostress, in-phase stresses (IPS) and alternating stresses (AS). Bars represent mean division rates among different growth chambers. Error bars are 95% confidence intervals of the mean. Red symbols show the average division rate in each growth chamber, with different symbols representing experiments performed on different days with different microfluidic chips. Mean division rates were compared between IPS and AS conditions using t-tests (*** P < 0.001). (c, d) Temporal dynamics of division rate during a period of 96 minutes in (c) IPS and (d) AS conditions for wild-type cells. Each dot shows the division rate during a 6-minute window centered on that dot and averaged for all fields of view sharing the same condition and all periods in the experiment. Gray areas are 95% confidence intervals of the mean division rate. Horizontal dotted lines show the mean division rate for all data collected in each half period. The colored bars represent the periodic fluctuations in glucose (orange) and/or sorbitol (blue); hatching represents stress. (a-d) Cells were grown under fluctuations of 2% glucose and 1 M sorbitol.

HOG pathway mutants grow faster under in-phase stresses than under alternating stresses.

(a) Division rates measured during growth in IPS and AS conditions with 2% glucose and a fluctuation period of 24 minutes. Bars show the mean division rate measured in different growth chambers of the microfluidic chip. Error bars are 95% confidence intervals of the mean. Red symbols show the average division rate in each growth chamber. Results of t-tests comparing the wild-type and mutant strains under the same condition are indicated above each bar; results comparing the same strain under different conditions are shown above each pair of bars (ns: P > 0.05; * 0.01 < P < 0.05; ** 0.001 < P < 0.01; *** P < 0.001). (b, c) Temporal dynamics of division rate during a period of 96 minutes in (b) IPS and (c) AS conditions for wild-type (black) and pbs2Δ mutant (green) cells. Each dot shows the division rate during a 6-minute window centered on that dot and averaged for all fields of view sharing the same condition and all periods in the experiment. Gray and green areas are 95% confidence intervals of the mean division rate. Horizontal dotted lines show the mean division rate for all data collected in each half period. The colored bars represent the periodic fluctuations in glucose (orange) and/or sorbitol (blue); hatching represents stress.

HOG pathway mutants exhibit a lower death rate and an increased population growth rate under in-phase stresses.

(a,b) Images of wild-type (a) and pbs2Δ mutant (b) cells before (t = 0 min) and after (t = 600 min) growth in IPS (left) and AS (right) conditions. Fluorescence and bright field images were merged to visualize nuclei marked with HTB2-mCherry. Scale bar represents 10 μm. White arrows indicate nuclei of representative dead cells. (c, d) Death rates (c) and population growth rates (d) of the reference strain and seven deletion mutants under IPS and AS conditions. Population growth rates were calculated as the difference between division rates (Figure 4b) and death rates (Figure 5c). Bars show mean rates measured in different growth chambers. Error bars are 95% confidence intervals of the mean. Red symbols show the average rate for each field of view. Results of t-tests comparing the wild-type and mutant strains under the same conditions are indicated above each bar; results comparing the same strain under different conditions are shown above each pair of bars (ns P > 0.05; * 0.01 < P < 0.05; ** 0.001 < P < 0.01; *** P < 0.001). (a-d) Cells were grown under fluctuations of 2% glucose and 1 M sorbitol at a period of 24 minutes.

Osmoregulation is delayed after hyperosmotic stress under in-phase stresses but not under alternating stresses.

(a-b) Timelapse images of cells expressing Htb2-mCherry and Hog1-GFP under AS (a) and IPS conditions (b) for periods of 96 minutes, showing cellular localization of Hog1p during the second osmotic shock in each experiment. Top: fluorescence and brightfield images merged to visualize cell nuclei tagged with histone HTB2-mCherry. Bottom: fluorescence images showing Hog1-GFP localization. The time since the last environmental change is indicated below each image. Scale bars represent 10 μm. (c) Temporal dynamics of the enrichment of Hog1-GFP fluorescence in cell nuclei under IPS (red curve) and AS (green curve) conditions. Colored areas indicate 95% confidence intervals. (d) Temporal dynamics of cell size (area) in IPS (green) and AS (red) conditions for the same cells as in panel (c). Each curve shows the mean area measured among cells. Colored areas indicate 95% confidence intervals of the mean. (c-d) Each curve shows the mean nuclear enrichment or mean cell size for 11 to 25 cells in one or two fields of view. The colored graphs represent the periodic fluctuations of glucose (orange) and/or sorbitol (blue); hatching represents stress; gray indicates exposure to 1 M sorbitol. (a-d) Cells were grown under fluctuations of 2% glucose and 1 M sorbitol at a period of 96 minutes.

A transcriptional target of the HOG pathway is under-expressed during in-phase stresses and over-expressed during alternating stresses.

(a) Images of cells expressing a fluorescent reporter (mCitrine) under control of the STL1 promoter known to be regulated by the HOG pathway. Rows correspond to different conditions (as indicated on the left) and columns correspond to different time points after transitioning from complete medium to each condition. Bright field and fluorescence images are overlaid. The scale bar represents 20 μm. (b-c) Temporal dynamics of PSTL1-mCitrine expression in five conditions: IPS (green curves), AS (red curve), periodic osmotic stress with constant glucose (brown curve), a single transition to constant osmotic stress with glucose for 10 hours (blue curve), and a short pulse of osmotic stress with constant glucose (purple curve). The fluctuation period is 96 minutes in (b) and 24 minutes in (c) for the first three conditions. 2% glucose was used in all conditions. Each curve shows the mean fold change of fluorescence intensity measured for 30—65 cells from four or five fields of view. Colored areas indicate 95% confidence intervals of the mean. The colored graphs represent the periodic fluctuations of glucose (orange) and/or sorbitol (blue); hatching represents stress; gray indicates exposure to 1 M sorbitol.