Fig 1.
Indecisive and committed phases of polarity behavior in mating yeast.
(A) Pheromone (α or a-factor, yellow) binds to pheromone receptors (Ste2 or Ste3, green), activating receptor-associated Gα and Gβγ. (Left) Gβγ recruits Ste5, which activates a MAPK cascade leading to cell cycle arrest in G1, transcription of mating-related genes, and polarization. (Right) Gβγ recruits Far1, which is bound to the GEF Cdc24, leading to local Cdc42 activation. (B) Representative images of cells with fluorescent polarity probes in a mating mixture (a, Bem1–tdTomato, magenta; α, Bem1–GFP, green). At 0 min (top), cells are budding; those that will go on to mate are circled. The same mating pairs are indicated at 84 min (bottom). By this time, two pairs have fused, forming zygotes with mixed magenta/green fluorescence (outlined in red, blue), and two pairs have polarized toward one another but not yet fused (outlined in white, orange). (C) Localization of Bem1–GFP in a representative mating cell. Top: Inverted maximum z-projection images of Bem1–GFP at selected times (cytokinesis = 0 min). A weak Bem1 cluster appears 4 min (blue box). The cluster moves and fluctuates in intensity during an “indecisive phase” until 38 min (orange box), when it strengthens and remains stationary during a “committed phase” until fusion occurs at 54 min. Bottom: quantification of Bem1 clustering (CP) in the same cell (see Materials and Methods). (D) Bem1 CP in 10 representative mating cells, as in (C). The timeline extends back to the time of cell birth from the time of fusion (0 min). Color switches from blue to orange at Tp. (E) Bem1 CP at Tic and Tp (n = 44, error bars = SD, *t test, p < 0.05). (F) Localization of Bem1–GFP and Spa2–mCherry in a mating cell from birth (−82 min) to fusion (0 min). Top: Inverted maximum z-projection images of the indicated probes. Bottom: quantification of Bem1 and Spa2 CP in the same cell. (G) Spa2 CP in nine representative cells, displayed as in (D). (H) The cumulative distribution (n = 246) of the interval between birth and Tic in mating cells. (I) The cumulative distribution (n = 246) of the duration of the indecisive (blue) and committed phase (orange). Dashed lines indicate median. Scale bar, 3 μm. Strains: DLY12943, DLY7593 (B–E, H, I), DLY9070 (E), DLY21379 (F, E). Bem1, bud emergence 1; Cdc, cell division control; CP, clustering parameter; Far1, factor arrest 1; GEF, guanine nucleotide exchange factor; GFP, green fluorescent protein; MAPK, mitogen-activated protein kinase; Spa2, spindle pole antigen 2; Ste, sterile; tdTomato, tandem dimer tomato; Tic, time of initial clustering; Tp, time of polarization.
Fig 2.
Timing of commitment by mating partners.
(A) Two possibilities for polarization timing in mating partners that “meet” when they are at different stages of the cell cycle. Top: The first-born cell (blue) locates the partner (orange) while the latter is still completing the cell cycle. The first-born cell polarizes and waits during an extended committed phase for its partner to catch up. Bottom: The first-born cell cannot locate its partner until the partner enters G1 phase. It remains in an extended indecisive phase until the partner enters G1, after which both cells polarize. Dashed line: median. (B) Cumulative distribution of the interval between stable polarization and fusion (committed phase) in first-born (blue, n = 93) and second-born cells (orange, n = 153). (C) Cumulative distribution of the interval between initial clustering and commitment (indecisive phase) in first-born (blue, n = 93) and second-born cells (orange, n = 153) (*two-sample KS test, p < 0.05). Dashed lines: median. (D) Cumulative distribution of the duration of the indecisive phase (dashed) and committed phase (solid) in second-born cells of MATa (blue, n = 87) and α (orange, n = 66). (E) Pheromone synthesis is high in G1 and decreases in S/G2/M. Cells (MATα) expressing sfGFP from the MFα1 promoter were imaged for 150 min. Reporter fluorescence was normalized to the value at the time of first bud emergence (0 min: black dashed line). Curves were colored orange from birth to bud emergence (G1 phase) and blue from bud emergence to cytokinesis (S, G2, and M phases). (F) A reporter driven by the constitutive TEF1 promoter was analyzed as in E. (G) Modeling MFα1 promoter activity as switching between high in G1 and low in S/G2/M (inset, blue) can predict the expected fluctuation in fluorescence intensity (inset, orange). Main graph: fractional change in output fluorescence intensity signal (y axis) predicted as a function of fold-change in the input MFα1 promoter activity across the cell cycle (x axis). The observed 20% change (0.2: black dashed line) would require >5-fold change in pheromone synthesis rate (inset). Code for Fig 2G is available at https://github.com/DebrajGhose/Ratiometric-GPCR-signaling-enables-directional-sensing-in-yeast. Strains: DLY12943, DLY7593 (B-D), DLY22883 (E), DLY22928 (F). GFP, green fluorescent protein; KS, Kolmogorov–Smirnov; MAT, mating type; MFα1, major α-factor gene 1; sf, superfolder; TEF1, translation elongation factor 1.
Fig 3.
Commitment coincides with an increase in MAPK activity.
(A) Localization of MAPK activity sensor varies through the cell cycle. Inverted maximum z-projection images of the sensor Ste71-33–NLS–NLS–mCherry in representative vegetatively growing cells. (B) The sensor is exported from the nucleus in response to MAPK activation. MATa cells harboring Ste71-33–NLS–NLS–mCherry were mixed with MATα cells and imaged as in (A). Representative mating cells are illustrated from birth to fusion (0 min = fusion). (C) MAPK activity calculated from sensor distribution (see Materials and Methods) in the 60 min prior to fusion for 10 representative cells. The transition from the indecisive (blue) to committed phase (orange) was determined from a Spa2-GFP probe in the same cells. (D) Left: MAPK activity (blue, as in C) and Spa2 CP (orange, as in Fig 1F) in a representative mating cell. Right: six other cells. (E) Average MAPK activity (blue) and Spa2 CP (orange) in the 60 min prior to fusion (n = 41 cells). Shading: SD. (F) Cross-covariance of MAPK activity and Spa2 CP during the indecisive phase (window from 60 min to 20 min before fusion) in mating cells (n = 41 cells). Lag represents the time by which the Spa2 CP was shifted forward in time relative to the MAPK activity. 1 = perfect cross-covariance. (G) Cells harboring Pgal1–Ste5–CTM allow MAPK induction by β-estradiol without pheromone treatment. Cells with the MAPK sensor and Spa2-GFP were imaged following β-estradiol treatment. Inverted maximum z-projection images of selected time points show Spa2 neck localization during cytokinesis, indecisive behavior upon intermediate MAPK activation, and committed behavior following high MAPK activation in representative cells. Scale bar, 3 μm. Strains: DLY22259 (A-F), DLY22764 (G). CP, clustering parameter; CTM, carboxy-terminal transmembrane domain; GFP, green fluorescent protein; MAPK, mitogen-activated protein kinase; MATα, mating type α; NLS, nuclear localization sequence; Pgal1, galactose metabolism 1 promoter; Spa2, spindle pole antigen 2; Ste, sterile.
Fig 4.
Nonrandom initial clustering of polarity factors.
(A) Orientation with respect to partner. Top: cumulative distribution of initial Bem1 cluster location (inset: 0° = cluster oriented towards mating partner) relative to the mating partner (n = 246). Black line: hypothetical random distribution (*KS test, p < 0.05). Bottom: polar histogram display of the same data. (B) Top: cumulative distribution of initial cluster location relative to the mating partner, plotted separately for first-born (blue, n = 93) and second-born (orange, n = 153) cells (*KS test, p < 0.05). Bottom: polar histogram display of the same data. (C) Orientation with respect to neck. Cumulative distribution of initial cluster location relative to the site of cytokinesis plotted as in (A) (n = 246, KS test, NS). Bottom: polar histogram display of the same data. (D) (Left) Single-plane inverted fluorescent images of representative Ste27XR-GPAAD–sfGFP (top) and Snc2-GFP (bottom) cells that were bleached to assess fluorescence recovery due to diffusion. (Right) Fluorescence recovery of the bleached region in the depicted cells, with exponential fits. (E) Estimated diffusion constants; each dot is one cell, and horizontal lines mark averages (n = 17 and 11 for Ste2–sfGFP and Snc2–GFP cells, respectively). Code for Fig 4D and 4E is available at https://github.com/DebrajGhose/Ratiometric-GPCR-signaling-enables-directional-sensing-in-yeast. Strains: DLY12943, DLY7593 (A-C), DLY21705, DLY17966 (D, E). Bem1, bud emergence 1; GFP, green fluorescent protein; GPAAD, GPFAD to GPAAD mutation; KS, Kolmogorov–Smirnov; NS, not significant; sf, superfolder; Snc2, suppressor of the null allele of CAP 2; Ste, sterile; 7XR, 7 lysine-to-arginine mutations.
Fig 5.
Pheromone receptor density variation around the cell membrane.
(A) Single-plane inverted image of vegetatively growing cells expressing Ste2–sfGFP. Membrane signal (blue arrow) is Ste2–sfGFP, but vacuole signal (orange arrow) is probably sfGFP cleaved from Ste2-sfGFP after internalization. (B) Ste2 distribution through the cell cycle in representative cells. (C) G1 cells display Ste2 distributions ranging from almost uniform (top) to very asymmetric (bottom). (D) Quantification of Ste2–sfGFP membrane distribution in G1 cells. Individual linescans (examples in color) were normalized to have the same total fluorescence and centered on the peak of a smoothed spline fit. Black line, average (n = 71). (E–G) Particle-based simulations of receptor–G-protein interactions at the cell membrane. Receptors and G proteins were simulated as diffusing particles on a spherical surface. G proteins were activated when they encountered an active receptor, and active G proteins were spontaneously inactivated with first-order kinetics. (E) Receptors were distributed unevenly: receptor density is indicated by the thickness of the black line (inset) and reflects a 3-fold gradient, similar to the Ste2 distribution. (F) A 1.5-fold pheromone gradient was simulated along the x axis by varying the % of active receptors from 40% to 60% across the cell diameter. (G) Simulations were conducted with receptor activity and density gradients oriented as in the illustrations. The locations of all of the active G proteins were used to calculate a G-protein vector, whose angle to the direction of the pheromone gradient is plotted (y axis) against time (x axis) (left). 0° indicates perfect orientation: active G-protein vector in the same direction as the applied receptor activity gradient. The approximate range of G-protein vectors (blue wedge) is shown in the cartoon on the right, along with the pheromone gradient (green shading) and receptor density (as in E). Code and key data for Fig 5G are available at https://github.com/mikepab/ratiometric-gpcr-particle-sims. Scale bar, 3 μm. Strains: DLY20713 (A–D). A.U., arbitrary unit; GFP, green fluorescent protein; GPCR, G-protein–coupled receptor; sf, superfolder; Ste, sterile.
Fig 6.
Ratiometric sensing allows cells to orient towards partners despite uneven receptor density.
(A) Proposed ratiometric pheromone sensing mechanism. The G protein is activated by pheromone-bound receptor (Ste2 + α-factor) and inactivated by the RGS protein Sst2. Sst2 associates with inactive Ste2. Thus, G-protein activity reflects the ratio of bound to unbound receptors. (B) Particle-based simulations were repeated as in Fig 5, except that instead of spontaneous inactivation, G proteins were inactivated upon encountering inactive receptors. These “ratiometric” simulations (orange) were plotted as in Fig 5G. For comparison, both the nonratiometric (blue) and ratiometric (orange) results are depicted in the cartoons. (C) hsRGS4 is distributed uniformly on the membrane. Single-plane inverted image of hsRGS4–CFP. (D) Pheromone sensitivity measured via halo assay in wild-type cells (orange) and cells in which Sst2 was replaced by one copy (gray, hsRGS4, *t test, p < 0.05) or two copies (blue, hsRGS4×2, NS) of hsRGS4 (n = 9, three technical replicates at three pheromone concentrations, normalized to the average wild-type halo diameter). (E) Left: cumulative distribution of initial cluster location relative to the nearest potential mating partner in wild-type (orange, n = 222) and hsRGS4×2 cells (blue, n = 62, *two-sample KS test, p < 0.05). Right: polar histogram of the same data. (F) Left: cumulative distribution of initial cluster location relative to the site of cytokinesis in wild-type (orange, n = 222) and hsRGS4x2 cells (blue, n = 62, *two-sample KS test, p < 0.05). Right: polar histogram of the same data. (G) Bem1 initial cluster location is biased by Ste2 distribution in cells with hsRGS4×2 but not in cells with wild-type Sst2. Averaged Bem1–tdTomato distribution (shaded region = standard deviation) at the time of initial clustering, centered on the location with maximum Ste2, in wild-type (orange, n = 33) and hsRGS4×2 cells (blue, n = 33). Ste2 and Bem1 linescans were acquired from maximum projection images. (H) Example images of cells at the time of initial Bem1 clustering. Bem1 clusters (magenta) sometimes form in areas depleted of receptor (green) in wild-type cells, but in hsRGS4×2 cells, clusters tend to form where receptors are concentrated. Single-plane Ste2–sfGFP images (first column), maximum projection Bem1–tdTomato images (second column), and overlays (third column) from representative wild-type and hsRGS4×2 cells. A simplified cartoon (fourth column) depicts Ste2 distribution and the location of Bem1 initial clusters (arrow). Code and key data for Fig 6B are available at https://github.com/mikepab/ratiometric-gpcr-particle-sims. Scale bar, 3 μm. Strains: DLY22318 (C, D), DLY22321 (D), DLY22520 (D–F), 12943 (E–F), DLY22243, 22628 (G, H). Bem1, bud emergence 1; CFP, cyan fluorescent protein; GFP, green fluorescent protein; hsRGS4, Homo sapiens regulator of G-protein signaling 4; KS, Kolmogorov–Smirnov; NS, not significant; RGS, regulator of G-protein signaling; sf, superfolder; Sst2, supersensitive 2; Ste, sterile; tdTomato, tandem dimer tomato.
Fig 7.
Ratiometric sensing makes gradient detection robust to changes in receptor distribution.
(A) Single-plane inverted images of Ste2–sfGFP (top), Ste2NPF–sfGFP (middle), and Ste27XR-GPAAD–sfGFP (bottom) in representative G1 cells. Ste27XR-GPAAD–sfGFP images were scaled differently to compensate for increased abundance. (B) Average Ste2 membrane distribution, quantified as in Fig 5D, in G1 cells with Ste2–sfGFP (blue), Ste2NPF–sfGFP (orange), and Ste27XR-GPAAD–sfGFP (green). (C) Ste2–sfGFP abundance. Left: representative western blot (full uncropped blot available in S5 Fig). α-GFP antibodies label two bands—full-length Ste2–sfGFP and vacuolar sfGFP (note absence of vacuole signal for Ste27XR-GPAAD). Right: quantification of full-length Ste2 abundance (n = 3 biological replicates, normalized to the average abundance of wild-type Ste2). (D) Halo assay for global pheromone sensitivity of cells with wild-type Ste2 (blue), Ste2NPF (orange), and Ste27XR-GPAAD (green). Top: images of representative halos. Bottom: quantification of halo diameter (n = 12, 3 technical replicates, normalized to the average wild-type halo diameter; *t test, p < 0.05). (E) Halo assay as in (D) for cells with hsRGS4×2 in place of SST2, with wild-type Ste2 (blue) and Ste27XR-GPAAD (green). (F) Left: Cumulative distribution of initial Bem1 cluster orientation relative to the nearest potential mating partner for MATa cells born immediately adjacent to a MATα cell in G1. Cells with wild-type Ste2 (blue, n = 222), Ste2NPF (orange, n = 93, NS), or Ste27XR-GPAAD (green, n = 148, NS). Right: polar histograms of the same data. (G) Left: Cumulative distribution of initial Bem1 cluster orientation as in (E) for cells with hsRGS4×2 in place of SST2, with wild-type Ste2 (blue, n = 62), Ste2NPF (orange, n = 66, NS), or Ste27XR-GPAAD (green, n = 65, *two-sample KS test, p < 0.05). Note: hsRGS4×2 + Ste27XR-GPAAD (green) is not significantly different from wild type + SST2 (blue dashed line). Right: polar histograms of the same data. Strains: DLY20713, DLY20715, DLY21705 (A, B), DLY21203, DLY21206, DLY21704 (C), DLY8993, DLY21205, DLY21206 (D), DLY23623, DLY23624 (E), DLY12943, DLY22058, DLY22397 (F), DLY22520, DLY22570, DLY22606 (G). A.U., arbitrary unit; Bem1, bud emergence 1; Cdc, cell division cycle; GFP, green fluorescent protein; GPAAD, GPFAD to GPAAD mutation; hsRGS4, Homo sapiens regulator of G-protein signaling 4; KS, Kolmogorov–Smirnov; MAT, mating type; NPF, GPFAD to NPFAD mutation; NS, not significant; RGS, regulator of G-protein signaling; sf, superfolder; Sst2, supersensitive 2; Ste, sterile; 7XR, 7 lysine-to-arginine mutations.
Fig 8.
Ratiometric sensing amplifies the gradient signal and improves accuracy even when receptors are distributed uniformly.
(A) When receptor distribution is uniform, a gradient of active receptors automatically implies an opposing gradient of inactive receptors (top). Bottom: G-protein activation and inactivation rates in ratiometric versus nonratiometric models, criterion for matching inactivation rates in the two models, and definition of β. (B) Predicted gradient profiles for active G protein in ratiometric (yellow) and nonratiometric (orange) models illustrated for receptor activity gradients (blue) of differing steepness and for different values of β. Steepness of the receptor gradient is the slope normalized to the maximum slope of a linear gradient with 0% active receptors at left and 100% active receptors at right. (C) Plot of the SR (vertical axis and color bar) for different gradient steepness and β. The SR refers to the difference in active G-protein concentration between the two ends of the gradient predicted by the ratiometric model divided by that predicted by the nonratiometric model, assuming steady state. (D) For gradients that exhibit the same signal, the shape of the gradient can affect the accuracy with which a cell would pick the right site for polarization. Three gradient shapes are illustrated for the same signal, and the shaded region indicates the part of the gradient where the active G-protein concentration is above the average. (E) Simulations with uniform receptor density. The ratiometric (orange) and nonratiometric (blue) models were simulated as in Fig 6B. (F) Left: Cumulative distribution of initial Bem1 cluster orientation relative to the nearest potential mating partner for MATa cells born immediately adjacent to a MATα cell in G1. Cells with Ste27XR-GPAAD (uniform receptors) and either wild-type Sst2 (orange, n = 148), or hsRGS4×2 (blue, n = 65). Right: polar histograms of the same data. (G) Cumulative distribution of the duration of the indecisive phase in second-born cells, plotted separately for cells in which the initial cluster formed within 60° of the mating partner (θ < 60°, blue, n = 106), and cells in which the initial cluster formed greater than 60° from the partner (θ > 60°, orange, n = 47) (*two-sample KS test, p < 0.05). Inset: diagram displaying the two groups of cells. Code for Fig 8B and 8C is available at https://github.com/DebrajGhose/Ratiometric-GPCR-signaling-enables-directional-sensing-in-yeast. Code and key data for Fig 8E are available at https://github.com/mikepab/ratiometric-gpcr-particle-sims. Strains: DLY22397, DLY22606 (F), DLY12943, DLY7593 (G). Bem1, bud emergence 1; GPCR, G-protein–coupled receptor; GPAAD, GPFAD to GPAAD mutation; hsRGS4, Homo sapiens regulator of G-protein signaling 4; KS, Kolmogorov–Smirnov; MAT, mating type; NS, not significant; RGS, regulator of G-protein signaling; SR, Signal Ratio; Sst2, supersensitive 2; Ste, sterile; 7XR, 7 lysine-to-arginine mutations.
Table 1.
Yeast strains and genotypes.