High-resolution mass measurements of single budding yeast reveal linear growth segments

Abstract

The regulation of cell growth has fundamental physiological, biotechnological and medical implications. However, methods that can continuously monitor individual cells at sufficient mass and time resolution hardly exist. Particularly, detecting the mass of individual microbial cells, which are much smaller than mammalian cells, remains challenging. Here, we modify a previously described cell balance ('picobalance') to monitor the proliferation of single cells of the budding yeast, Saccharomyces cerevisiae, under culture conditions in real time. Combined with optical microscopy to monitor the yeast morphology and cell cycle phase, the picobalance approaches a total mass resolution of 0.45 pg. Our results show that single budding yeast cells (S/G2/M phase) increase total mass in multiple linear segments sequentially, switching their growth rates. The growth rates weakly correlate with the cell mass of the growth segments, and the duration of each growth segment correlates negatively with cell mass. We envision that our technology will be useful for direct, accurate monitoring of the growth of single cells throughout their cycle.

Fig. 1. Experimental set-up for the simultaneous acquisition of mass and morphology of single yeast cells.

a Block diagram and main components of the picobalance. The total cell mass is detected by a cantilever that acts as microresonator and cell substrate. A blue laser photoactuates the cantilever at its natural resonance frequency f_N and the cantilever movement is detected by an infrared (IR) laser reflected onto a photodiode. The blue laser power is reduced by a neutral density (ND) filter. Attaching a cell to the cantilever shifts f_N, which is tracked over time by keeping the cantilever phase at 90° by a lock-in amplifier, phase-locked loop, function generator and laser controller. A temperature (T)-controlled environmental chamber provides cell culture conditions (30 °C). The setup is placed on an inverted optical microscope as add-on module. Measurements are acquired in yeast culture medium. b The sensitivity of the cell mass detection is a function of the cantilever mass (see Eqs. (1) and (2), “Methods”). c Scanning electron microscopy image of a silicon nitride cantilever micromachined for yeast cell mass measurements. d Mass measurements (n = 3 from 3 independent experiments) using the continuous mode (only one measurement shown, all shown in Supplementary Fig. 1c). Top, to attain a high time and mass resolution, f_N of the cantilever is continuously tracked by the phase-locked loop (see a). Bottom, a typical background measurement. On average, the noise is 2.3 ± 0.6 pg (mean ± SD) while smoothing (100 s moving window) reduces the noise to 0.45 ± 0.13 pg. e Mass measurements (n = 3 from three independent experiments) using the sweep mode (only one measurement shown, all shown in Supplementary Fig. 1d). Top, to minimize possible impact of the blue laser on cell viability, the cantilever is oscillated across frequency sweeps while recording the cantilever amplitude and phase. In between the frequency sweeps the blue laser is switched off. The sweep mode provides lower mass and time resolution. Bottom, a typical background measurement. On average, the noise is 11.0 ± 1.4 pg while smoothing (350 s moving window) reduces the noise to 4.6 ± 0.5 pg.

Fig. 2. Mass and morphology of single S. cerevisiae cells budding daughter cells.

a, b While continuously measuring the total mass of a single yeast cell, differential interference contrast (DIC) images (taken every 5 min) show the budding of the cell attached to the cantilever of the picobalance (Supplementary Movie 1). White arrows indicate budding daughter cells. The raw data (black curve) shows the total mass of a growing yeast cell acquired every 10 ms using the continuous mode. The red curve shows the average raw data (100 s moving window, “Methods”). Cyan bars indicate where cells are in the S/G2/M phase when bud growth is observed. Yeast cells were attached to ConA-coated microcantilevers and the measurements were recorded in yeast culture medium at 30 °C (“Methods”). c Growth curves (n = 19) of single yeast cells in the S/G2/M phase (bud growth) as measured by the picobalance using the continuous mode (moving average of 100 s) in (n = 19) independent experiments. The overall growth rates between starting and end mass range between 0.2 and 1.6 pg min^–1, with an average of 0.6 ± 0.4 pg min^–1 (mean ± SD). The duration of the S/G2/M phase ranges from 50 to 140 min, with an average of 94 ± 23 min (Supplementary Fig. 4). DIC images in a, b were contrast enhanced using a custom flat field correction (“Methods”). Scale bars (white lines), 10 µm.

Fig. 3. Mass and cell cycle measurements of single S. cerevisiae cells budding daughter cells.

a, b Single yeast cells expressing the fluorescently labeled cell cycle marker proteins (Myo1-mKate2 (3×) and Whi5-mKOκ (1×)), were imaged using differential interference contrast (DIC) and fluorescence microscopy every 2 min (upper panels). A phase and amplitude curve of the microcantilever were recorded over intervals ≈50 s to measure the cell mass using the sweep mode (Supplementary Movie 4). Between consecutive mass measurements, the infrared and blue lasers of the picobalance were switched off for ≈ 20 s to reduce bleaching of the fluorophores and to reduce potential perturbance of yeast growth. Cell mass values as derived from sets of single amplitude curves are shown as gray dots. Average raw data (350 s moving window, red line) shows the trend. Cyan bars on the time axis denote the S/G2/M phase of the yeast cell cycle, and magenta bars denote the G1 phase. The star (*) in b denotes the (partial) detachment of the daughter cell after cytokinesis, which drops the total mass. Scale bars (white), 10 µm. c Growth curves of (n = 19) single yeast cells progressing through the S/G2/M phase (bud growth) as measured by the picobalance using the sweep mode in (n = 19) independent experiments. The overall growth rates between starting and end mass range between 0.1 and 2.0 pg min^–1, with an average of 0.7 ± 0.5 pg min^–1 (mean ± SD). The duration of the S/G2/M phase ranges from 57 to 184 min, with an average of 96 ± 35 min.

Fig. 4. Single S. cerevisiae cells budding daughter cells increase mass in linear segments.

a, b Mass versus time curves of two S. cerevisiae cells segmented (green vertical dashed lines) into phases of linear growth (colored) by a segmented linear regression model (SLM) (violet lines). DIC images show budding cells attached to the cantilever of the picobalance. Shown are raw data. Scale bars (white), 10 μm. Mass measurements were recorded using the continuous mode. Cyan bars indicate where cells are in the S/G2/M phase when bud growth is observed. c, d Residuals of fitting the SLM to the raw data of growth curves. e, f Growth rate (GR) of each linear growth segment (colored) shown in (a, b). GR values given above the bars. g–j Growth analysis of (n = 19) biologically independent cells measured using the continuous mode in (n = 19) independent experiments. Data represented with violin plots show the raw data (gray dots), mean (orange star), median (pink line), mean and standard deviation (mean ± SD, medium gray vertical box), distribution as kernel density (light gray area from the 1st percentile q(0.01) to the 99th percentile q(0.99)) and the 95% confidence interval (CI, dark gray edges of hourglass hexagon). g Distribution of the number of linear growth segments measured for all cells. Mean is 5 (orange star) and median (horizontal pink line) is 4 segments per cell with 95% CI = [3.3, 4.7] (dark gray edges of hourglass hexagon). The distribution of the data is shown by the kernel density (light gray area) and the standard deviation (SD, medium gray vertical box). h Growth rates determined from linear growth segments of all cells. Median is 0.70 pg min^–1 with 95% CI = [0.6, 0.9]. i Duration (time span) of the linear growth segments of all cells. Median is 13.7 min with 95% CI = [11.4, 16.0]. j Duration of the budding (S/G2/M) phase of all cells. Median is 94 min with 95% CI = [84.1, 104.0].

Fig. 5. Model of a single yeast cell increasing mass throughout the S/G2/M phase in linear segments of constant growth rates.

The schematic of the budding yeast cell cycle above the plot ranges from the birth (G1 phase) to the budding phase (S/G2/M), the cell division, and the forthcoming G1 phase. During the S/G2/M phase, single S. cerevisiae cells increase mass ∆m in linear segments (purple lines), each lasting for a certain time ∆t. Each linear segment (here s1–s5) shows a different growth rate (GR); GR = ∆m /∆t (Fig. 4) and lasts over a different time period ∆t. At the resolution of our mass measurements, we observe single yeast cells to progress through the S/G2/M phase (cyan bar at bottom) on average in five linear growth segments.