Timing of gene expression in a cell-fate decision system

Abstract

During development, morphogens provide extracellular cues allowing cells to select a specific fate by inducing complex transcriptional programs. The mating pathway in budding yeast offers simplified settings to understand this process. Pheromone secreted by the mating partner triggers the activity of a MAPK pathway, which results in the expression of hundreds of genes. Using a dynamic expression reporter, we quantified the kinetics of gene expression in single cells upon exogenous pheromone stimulation and in the physiological context of mating. In both conditions, we observed striking differences in the timing of induction of mating-responsive promoters. Biochemical analyses and generation of synthetic promoter variants demonstrated how the interplay between transcription factor binding and nucleosomes contributes to determine the kinetics of transcription in a simplified cell-fate decision system.

Keywords: MAPK pathway; gene expression; single‐cell measurements; yeast mating.

Figure 1. Interplay between kinase activity and promoter induction in the mating pathway

1. A, B

   Microscopy images of cells stimulated with a saturating pheromone concentration (1 μM) at time 0 min. The cells bear a histone tagged with CFP, a yellow SKARS reporting on Fus3p and Kss1p activities, and a red dPSTR reporting on pFIG1 (A) or pAGA1 (B) induction. For all experiments, unless stated otherwise, the stimulation was performed by addition of 1 μM α‐factor at time 0 min.

2. C, D

   Quantifications of the kinase activity (green, left axis), measured by the ratio of cytoplasmic to nuclear YFP, and of the pFIG1 (C) and pAGA1 (D) expressions, measured by the difference between nuclear and cytoplasmic fluorescence of the dPSTR (right axis). For all similar graphs, the solid line is the median response and the shaded area represents the 25^th–75^th percentiles of the population.

3. E

   Microscopy images of a strain carrying pFIG1‐dPSTR^R and pAGA1‐dPSTR^Y.

4. F

   Quantification of the response time of pFIG1 and pAGA1 reporters (see Materials and Methods). The inset is the difference response time between the pAGA1‐dPSTR^Y and the pFIG1‐dPSTR^R, for all cells expressing both promoters. The red shaded area represents cells expressing pAGA1 before pFIG1 (87%).

5. G

   Correlation of normalized dPSTR nuclear enrichments from all single cells of a representative experiment at different time points after stimulation.

6. H

   Northern blot detection of mRNAs from AGA1 and FIG1 after stimulation of the cells with mating pheromone. See also Appendix Fig S15.

Data information: All scale bars on microscopy images represent 2.5 μm.

Figure EV1. Dynamics and expression level of mating‐dependent promoters

1. A–N

   Population median (solid line) of the nuclear enrichment of the red dPSTR^R (left axis) or the pAGA1‐dPSTR^Y (right axis, yellow curves) for the 14 promoters of the study. Panels (A–F): early promoters, panels (G–K): intermediate promoters, panels (L–N): late promoters. Note that the scale of the dPSTR^Y is identical for all graphs, whereas dPSTR^R scales are different. The basal level and induced level can vary according to the measured promoter. For instance, pFAR1 has a high basal level, due to its cell cycle‐dependent induction.

2. O–Q

   Similar graphs for strains carrying different combinations of dPSTRs.

Data information: In all graphs, the solid line is the median response and the shaded areas represent the 25^th–75^th percentiles of the single‐cell responses. The curves are one representative experiment from at least three replicates.

Figure 2. Dynamics of induction of mating promoters after pheromone stimulation

1. A

   Response time versus mean expression output for the 14 mating‐dependent promoters. Dots represent the median response times of the cell population, and lines represent the 25^th and 75^th percentiles. All promoters were measured with the dPSTR^R. The strains also bear the pAGA1‐dPSTR^Y for direct comparison of the dynamics of promoter induction. The dashed line represents the detection sensitivity of the dPSTR^R reporter.

2. B

   Distributions of the differences in the response times between the pAGA1‐dPSTR^Y and the dPSTR^R in the same cell for pFUS1, pFUS2, and pFIG1.

3. C

   Correlation of the population‐averaged normalized nuclear enrichment of pAGA1‐dPSTR^Y and a selected set of promoters measured with the dPSTR^R at all time points of the experiments. The curves show the evolution in course of the experiment, from the bottom left to upper right corner, of the expression levels of the two measured promoters. The dots represent the P‐value (10⁻³ > P > 10⁻⁶ for small dots and P < 10⁻⁶ for large dots) of the t‐test comparing the offset of the measured promoter relative to the x = y line with the offset of the reference promoter pAGA1.

4. D, E

   Correlation of normalized dPSTR nuclear enrichments of single cells of at different time points after stimulation in a strain with pFUS1‐dPSTR^R and pAGA1‐dPSTR^Y (D) or pFIG1‐dPSTR^R and pKAR3‐dPSTR^Y (E).

5. F

   Evolution of the correlative promoter variability (CPV) in course of time, for various pairs of promoters. The curve represents the mean of three replicates, and the error bar represents the standard deviation between replicates. A low CPV corresponds to a similar expression between two promoters in the same cell (see Materials and Methods).

Figure EV2. Characterization of promoters relative to pAGA1

1. Distribution of the difference between the response time for pAGA1‐dPSTR^Y and the specified promoter measured with dPSTR^R for one representative experiment. N_C > 100 cells (see Materials and Methods). A sign test was performed to assess distribution centered around 0 (*10⁻²⁰ < P < 10⁻⁵; **P < 10⁻²⁰).

2. Correlation of the average normalized nuclear enrichment of pAGA1‐dPSTR^Y and specified promoters measured with dPSTR^R at all time points of the experiments. The dotted line is the x = y line and indicates the time direction (from bottom left to upper right). Each curve starts at 0 at the beginning of the experiment. The dots represent the P‐value (10⁻³ < P < 10⁻⁶ for small dots and P < 10⁻⁶ for big dots) of the t‐test comparing the offset of the measured promoter from the x = y line to the offset of the reference promoter pAGA1 (red curve, upper panel).

3. Quantification of the correlative promoter variability between the indicated promoter measured by dPSTR^R and the pAGA1‐dPSTRY (see Materials and Methods). Each curve represents the average of the CPV calculated for at least three biological replicates with the standard deviation of the three experiments. The dotted line is pAGA1‐dPSTR^R curve for comparison.

Figure EV3. Dose response of pAGA1 and pFIG1 induction

1. A, B

   Mean expression output for pAGA1 (A) and pFIG1 (B) in response to different pheromone concentrations, in a WT or bar1∆ (shaded) background. The expression output is defined as the maximal dPSTR nuclear enrichment following stimulation, for all cells of the experiment. Error bars represent the standard deviation of three replicates. Note that the induction of pAGA1 gradually increases with the pheromone concentration, whereas pFIG1 displays a switch‐like response in WT. Note that in a bar1∆ background, the expression occurs at lower concentrations and with higher level for both promoters. The first dot is the non‐induced control.

2. C

   Percentage of cells expressing pAGA1 (red) or pFIG1 (blue) in a WT (upper panel) or bar1∆ (lower panel) background, at various pheromone concentrations for one representative experiment. Note that only at high concentrations, a significant proportion of the population expresses pFIG1.

3. D

   Median nuclear enrichment of the pAGA1‐dPSTR^Y (red, left axis) and of the pFIG1‐dPSTR^R (blue, right axis) in course of time, for the different pheromone concentration, in the WT background, for one representative experiment. The solid line is the median, and the shaded area represents the 25^th–75^th percentile. The reference curves at 1 μM are represented in dashed line for comparison. Note that the scale of the pFIG1‐dPSTR^R is different between the 1 μM and the other concentrations.

Figure 3. Influence of promoter architecture on expression dynamics

1. A, B

   Maps of the two promoters pAGA1 and pFIG1. The filled arrows represent the location and orientation of consensus Ste12‐binding sites (nTGAAACn). The open arrows symbolize the non‐consensus binding sites that possess mutations within the six core nucleotides of the PREs. The sequences of each binding sites are detailed above, with capital nucleotides matching the consensus sequences and small nucleotides being mutations from the consensus. The numbers between sites represent the distance in bp between them or the ATG. Blue arrows represent nucleosomes position (Brogaard et al, 2012).

2. C, D

   Quantification of molecular events at the AGA1 (C) and FIG1 (D) loci. Fold increase in Ste12‐myc and Kar4‐HA binding at the promoter quantified by chromatin‐IP (open markers). Normalized −1 histone occupancy quantified by micrococcal nuclease (MNase) digestion. Transcript levels of AGA1 (C) and FIG1 (D) quantified by Northern blot (rounds). Each data point is the mean of three biological replicates, and the error bars represent their standard deviation.

3. E

   Response time versus mean expression output for various promoters in a WT background (circles, solid lines) or kar4∆ (diamonds, dashed lines) background, as described in Fig 2A. Red is pAGA1, blue is pFIG1, green is a chimeric construct between pFIG1 and the last 150 bp of pAGA1, cyan is a construct where the free non‐consensus binding site of pFIG1 (−209) was mutated into a consensus one, and purple is a combination of the chimeric construct with the mutation of the non‐consensus binding site into a PRE.

4. F

   In vivo binding of Ste12 and Kar4 was assessed by immunoprecipitation of Kar4p‐HA and detection of Ste12‐Myc in the presence and absence of pheromone.

Figure EV4. Mutants of the Group III have differentially impaired induction of pAGA1 and pFIG1

1. A, B

   Nuclear enrichment of the pAGA1‐dPSTR^Y (A) and pFIG1‐dPSTR^R (B) after stimulation by 1 μM of pheromone in the indicated mutant. Lines represent the median of either the mutant (red) or the WT strain (black) for one representative experiment, with the solid line representing the median and the shaded area representing the 25^th–75^th percentile.

2. C

   Correlation of the expression output (maximal dPSTR nuclear enrichment following stimulation) of pAGA1‐dPSTR^Y and pFIG1‐dPSTR^R for all single cells of the experiment, for the indicated strain. Dotted lines represent the threshold of expression (defined as the 20% of the WT mean expression output for each dPSTR). The Venn diagram represents the proportion of cells expressing pAGA1 (red circle) or pFIG1 (blue circle) or none of them (black circle).

3. D

   Representative microscopy images of the indicated mutant at the specified time point of the experiment. Arrows indicate shmooing events. Scale bars represent 2.5 μm.

Figure EV5. Effect of the loss of Kar4 on the induction of various promoters

1. Nuclear enrichment of the dPSTR^R measuring the indicated promoter in a WT (same color as the axis) or a kar4∆ background (red). The solid line is the median, and the shaded area represents the 25^th–75^th percentiles of the population.

2. Distribution of the response time of each promoter in a WT (same color as in A) or a kar4∆ background (red).

3. Correlation of the expression output (maximal dPSTR nuclear enrichment following stimulation) of the promoters dPSTR^R with the pAGA1‐dPSTR^Y in a WT (same color as in A and B) or in a kar4∆ background (red) for all single cells of the experiment.

Figure 4. Dynamics of gene expression during the mating process

1. Microscopy images of a mating mixture containing the MATa strain (Hta2‐CFP, pFIG1‐dPSTR^R, and pAGA1‐dPSTR^Y) and a MATα (cytoplasmic tdiRFP) at different times after beginning of the imaging (time 0). Fusion events are marked by a white arrow. Scale bars represent 2.5 μm.

2. Quantification of the nuclear enrichment of pFIG1‐dPSTR^R (blue, left axis) and of pAGA1‐dPSTR^Y (red, right axis). Single‐cell traces were synchronized relative to their fusion time, identified by a sudden increase in tdiRFP signal into the MATa cells.

3. Distribution of the response time of pAGA1 and pFIG1 relative to the fusion time.

4. Activation dynamics of various promoters prior to fusion as measured by dPSTR^R in different mating mixtures.

5. Cumulative probability of the response time relative to fusion for nine mating‐induced promoters measured in mating conditions.

Data information: In (B and D), the solid line is the median and the shaded area represents the 25^th–75^th percentiles of the population.