A highly synchronous mating assay allows quantifying the MAPK^Spk1 activity during the fission yeast mating process

To monitor fission yeast pheromone signalling throughout the mating process (Fig 1A and 1B), we established a protocol to induce highly synchronous mating, employing cells where the endogenous MAPK^Spk1 is tagged with GFP-2xFLAG and conducted quantitative Western blotting of phosphorylated MAPK^Spk1 (S1 Fig). Under this condition, homothallic h⁹⁰ cells started to mate 6 hours after induction of mating (Fig 1C, grey line). The phosphorylated (active) MAPK^Spk1 (ppMAPK^Spk1) signal was first detected three hours after induction of mating and reached its peak at about five to seven hours when cell fusion was also initially observed (Fig 1C, green line. Original membrane images and quantitations presented in S2 and S3 Figs). The ppMAPK^Spk1 then gradually decreased to a non-zero level as meiosis continued towards sporulation. It was also noted that the total MAPK^Spk1-GFP was essentially not expressed during the vegetative cycle, but it was promptly induced by nitrogen starvation (S2 and S3 Figs). The mapk^spk1 gene is a known target of the transcription factor Ste11 [41], which itself is activated by MAPK^Spk1 [42]. This positive feedback loop likely facilitates a swift increase of MAPK^Spk1 expression upon nitrogen starvation.

We found that MAPK^Spk1-GFP localised to the cytosol, cell cortex and the nucleus, with some nuclear accumulation, before gradually disappearing as the mating process came to an end (Figs 1D, S4 and S5). The quantification of the MAPK^Spk1-GFP in wildtype cells undergoing the sexual differentiation stages demonstrated both the nuclear and total GFP signal per cell/zygote peaked at the time of mating (for haploid cells) and karyogamy (for diploid zygotes) (S4 Fig), consistent with the Western blotting result. Transient foci of GFP signals were also found at the cell cortex, especially at the shmoo tips, as has been reported for MAPKK^Byr1, the activator of MAPK^Spk1 [36] (Figs 1D, orange arrow, S4 and S5).

As there is a strong correlation between the timings of peak MAPK^Spk1 activation and mating/cell fusion, we examined whether cell fusion is prerequisite for MAPK^Spk1 downregulation, utilising cells deficient in fus1 gene, which encodes a yeast formin homologue that plays an essential role in the cell fusion event during the mating process by regulating the cytoskeletal and membrane organisation [43–48]. In the fus1Δ mutant cells, the increased ppMAPK^Spk1 level stayed high 12 hours after induction of sexual differentiation, and the cells accumulated the MAPK^Spk1-GFP particularly in the nucleus and the shmoo tips (S6 Fig). The result demonstrated that successful mating is required to downregulate the MAPK^Spk1.

Constitutively active MAPKK^Byr1.DD causes constitutive activation of MAPK^Spk1

Activation of MAPKK family kinases is mediated by dual phosphorylation of conserved Ser/Thr residues [49], and when the corresponding Ser214/Thr218 were substituted to aspartic acid, the resultant fission yeast MAPKK^Byr1.DD was expected to act as a constitutively active MAPKK [50]. Indeed, the level of ppMAPK^Spk1 in the MAPKK^byr1.DD mutant remained high after reaching its highest intensity at around 13 hours after induction (Figs 1E, yellow line, S2 and S3), although the initial increase of the ppMAPK^Spk1 signal was slower compared to the wildtype strain. The ppMAPK^Spk1 level remained high even 48 hours after the induction of mating (S7 Fig). The result highlights that the endogenously expressed MAPKK^Byr1.DD is proficient in retaining the high ppMAPK^Spk1 level, and the counteracting de-phosphorylation by phosphatases Pmp1 and Pyp1 [51] is not efficient in downregulating the ppMAPK^Spk1 signal in the presence of MAPKK^Byr1.DD. Consistent with the observed slower increase in ppMAPK^Spk1, the nuclear localisation of MAPK^Spk1 was also delayed compared to the wildtype cells (Fig 1F). A strong nuclear MAPK^Spk1-GFP signal was then observed in the “paired” fusion-deficient cells, an intriguing phenotype of the MAPKK^byr1.DD cells (Fig 1F) [36, 50]. Interestingly, the projection tips of the paring cells often show increased MAPK^Spk1-GFP signal (Figs 1F, yellow arrows and S8).

The ras1.G17V mutation causes immediate but transient MAPK^Spk1 activation

The fission yeast equivalent of human oncogenic RAS.G12V is ras1.G17V, which induces an excessively elongated shmoo, and the cells fail to recognize a partner and become sterile [22,24]. The “elongated shmoo” phenotype was interpreted as an excess activation of the Ras1 downstream pathway(s), leading to a prediction that the ras1.G17V causes over-activation of MAPK^Spk1 [19,20]. However, the ppMAPK^Spk1 signal intensity declined comparably to that in wildtype cells, indicating that down-regulation of ppMAPK^Spk1 is effective in the ras1.G17V mutant, unlike in the MAPKK^byr1.DD mutant (Figs 1G, red line, S2 and S3). Correspondingly, the nuclear, cytoplasm and cell cortex MAPK^Spk1-GFP signals also declined 16–24 hours after induction of mating (Figs 1H and S8). Interestingly, an immediate increase of the ppMAPK^Spk1 signal upon induction of mating was observed (Fig 1G, red line).

The elongated ras1.G17V shmoos develop with a detectable level of MAPK^Spk1: neither amplitude nor duration of ppMAPK^Spk1 signal influences the ras1.G17V morphological phenotype

Having observed that MAPKK^byr1.DD and ras1.G17V show different MAPK^Spk1 activation profiles (sustained vs transient) and different morphological phenotypes (“paired” vs elongated), we examined a correlation between the MAPK^Spk1 activation profiles and the cell morphology. The ras1.G17V MAPKK^byr1.DD double mutant showed constitutive activation of MAPK^Spk1 (Figs 2A, light blue line, S2 and S3). The nuclear ppMAPK^Spk1-GFP signal in the ras1.G17V MAPKK^byr1.DD double mutant was present 24 hours after induction of mating, unlike the ras1.G17V single mutant cells (Fig 2B), confirming that the ras1.G17V MAPKK^byr1.DD double mutant cells retained a high ppMAPK^Spk1 level. Thus, MAPKK^byr1.DD is overall epistatic to ras1.G17V in terms of the MAPK^Spk1 activation status.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. In the ras1.G17V MAPKK^byr1.DD double mutant, the MAPK^Spk1 phosphorylation profile follows MAPKK^byr1.DD single mutant phenotype whilst cell morphology mimics the ras1.G17V single mutant phenotype.

(A) MAPK^Spk1 phosphorylation status in the ras1.G17V MAPKK^byr1.DD double mutant cells (KT3439). Cells were induced for mating by the plate mating assay system as described in the materials and methods. Quantitated ppMAPK^Spk1 signal (arbitrary unit) from western blots is presented. Original membrane images are presented in S2 Fig. Three biological replicates were used for quantitation (error bars are ± SD). α-tubulin was used as a loading control, and quantitation was carried out using the Image Studio ver2.1 software (Licor Odyssey CLx Scanner). The wildtype ppMAPK^Spk1 result (Fig 1C) is also presented as a reference. (B) The terminal mating phenotype of ras1.G17V MAPKK^byr1.DD double mutant is a phenocopy of ras1.G17V single mutant which shows the “elongated” morphology. Images were taken of ras1.G17V MAPKK^byr1.DD double mutant (KT3439) in the same way as in Fig 1. Time after induction of mating in hours is indicated on the left. (C) There is no morphological change in the absence of MAPK^Spk1 signalling. Cell images of MAPKK^byr1Δ (KT4300) and ras1.G17V MAPKK^byr1Δ (KT5215) strains are shown. Images were taken in the same way as in Fig 1. Time after induction of mating in hours is indicated on the left of each series. The scale bars represent 10µm.

https://doi.org/10.1371/journal.pgen.1012117.g002

However, for cell morphology, the ras1.G17V MAPKK^byr1.DD double mutant cells showed the “elongated” ras1.G17V phenotype (Fig 2B), demonstrating that ras1.G17V was epistatic to MAPKK^byr1.DD regarding cell morphology.

Interestingly, the morphological change in the ras1.G17V MAPKK^byr1.DD double mutant was first noticed 12 hours after induction of mating, much later than the ras1.G17V single mutant, but as a comparable timing as the “pair” formation of the MAPKK^byr1.DD single mutant (Fig 2B). Given the slower increase of the ppMAPK^Spk1 signal in the MAPKK^byr1.DD mutant, we predicted that the “elongated” shmoo formation still requires a certain level of MAPK^Spk1 activity. Indeed, when MAPKK^byr1 was deleted in the ras1.G17V mutant, not only was the MAPK^Spk1 activation abolished (S1G Fig), but also shmoo formation was abolished (Fig 2C). Based on these observations, we concluded that the Ras1 status, but not the MAPK^Spk1 activation profile, determines the cell morphology. Yet, the ras1.G17V “elongated shmoo” phenotype still requires a certain level of MAPK^Spk1 activity, which defines the timing of the shmoo formation.

Cdc42 is required for the shmoo formation and full activation of the MAPK^Spk1

During the vegetative cycle, ras1∆ cells show spherical cell morphology [22,24] where polarised localisation of active Cdc42 is compromised [52,53] (S9 Fig) which can be detected by CRIB-GFP, a specific binder of the active GTP-bound form of Cdc42 [54]. The current understanding is that Ras1 activates Cdc42, which leads to the activation of downstream Ste20-like kinase, Pak1/Shk1 [55–58], resulting in actin reorganisation and shmoo formation under mating conditions [35,59].

The elongated shmoo phenotype was lost when the scd1, encoding a Cdc42-GEF, was deleted in the ras1.G17V mutant. Instead, the cells showed a mating-deficient phenotype similar to the cdc42-GEF^scd1∆ single mutant, supporting the model that Cdc42 acts downstream of Ras1 to cause morphological changes (Fig 3A).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Ras1 activates both MAPK^Spk1 and Cdc42 pathways during pheromone signalling.

(A) The deletion of scd1 causes round cell morphology and sterility. Images of WT (KT3082), scd1Δ (KT4061) and scd1∆ ras1.G17V double mutant (KT4056), expressing the MAPK^Spk1-GFP, were taken in the same way as in Fig 1. Numbers on the left represents hours after induction of mating. (B) The ppMAPK^Spk1 levels during the sexual differentiation in scd1Δ (KT4061) cells. Results of three biological replicates (error bars are ± SD) are presented. Original membrane images are presented in S2 Fig. The wildtype ppMAPK^Spk1 result presented in Fig 1C is also shown in green as a reference. (C) Cell images of scd1Δ MAPKK^byr1.DD double mutant (KT4047). The images were taken in the same way as in Fig 1. Numbers on the left represent hours after induction of mating. (D) The ppMAPK^Spk1 levels in scd1Δ (KT4061), scd1Δ MAPKK^byr1.DD (KT4047) and scd1Δ ras1.G17V (KT4056) cell extracts. Original Western blotting data is presented in S10A Fig and the total MAPK^Spk1 levels in these samples are presented in S10D Fig. (E) The ppMAPK^Spk1 levels in ras1Δ (KT4323), MAPKK^byr1.DD (KT3435) and ras1Δ MAPKK^byr1.DD (KT4359) cell extracts. Original Western blotting data is presented in S10B Fig and the total MAPK^Spk1 levels in these samples are presented in S10E Fig. (F) The ppMAPK^Spk1 levels in mapkkk^byr2Δ (KT3763), MAPKK^byr1.DD (KT3435) and mapkkk^byr2Δ MAPKK^byr1.DD (KT4010) cell extracts. Original Western blotting data is presented in S10C Fig and the total MAPK^Spk1 levels in these samples are presented in S10F Fig. For (D), (E) and (F), quantification was carried out using the Image Studio ver2.1 (Li-cor). (G) Cell images of the strains mentioned in (E) and (F) were taken in the same way as in Fig 1. Numbers on the left represent hours after induction of mating. For all the images presented in (A), (C) and (G), the scale bars represent 10µm.

https://doi.org/10.1371/journal.pgen.1012117.g003

Interestingly, the ppMAPK^Spk1 level was substantially reduced in the cdc42-GEF^scd1∆ mutant compared to the wildtype (Figs 3B, purple line, S2 and S3). The result agreed with a previous study that predicted active Cdc42 to contribute to the activation of MAPKKK^Byr2 [33]. Even so, some MAPK^Spk1 activation occurred in the cdc42-GEF^scd1∆ mutant, and a nuclear MAPK^Spk1-GFP signal was low, but detectable (Fig 3A).

A reduced but substantial level of MAPK^Spk1 activation in the cdc42-GEF^scd1∆ mutant means that the mating deficiency of this mutant is unlikely to result from the lack of MAPK^Spk1 activation. Indeed, the introduction of MAPKK^byr1.DD to the cdc42-GEF^scd1∆ mutant did not restore the mating deficient phenotype, even though nuclear MAPK^Spk1-GFP highly accumulated and the ppMAPK^Spk1-GFP level was substantially increased (Figs 3C, 3D, and S10). The result shows that Cdc42 activity is required for the mating process regardless of the MAPK^Spk1 activation status. Taken together with the essential role of MAPK^Spk1, we concluded that the mating pheromone signalling feeds into two pathways, MAPK^Spk1 and Cdc42.

Ras1 activates two effector pathways, MAPK^Spk1 and Cdc42

In order to further clarify the role of Ras1, we examined the MAPK^Spk1 activation status and cell morphology in the following four strains: ras1∆ mutant, ras1∆ MAPKK^byr1.DD double mutant, MAPKKK^byr2∆ mutant and MAPKKK^byr2∆ MAPKK^byr1.DD double mutant. As mentioned earlier, the vegetatively growing ras1∆ cells show a round morphology with reduced cortical signal of CRIB-GFP, demonstrating that Cdc42 activation is compromised (S9 Fig). The deletion of ras1 also causes substantial reduction, but not complete elimination, of the ppMAPK^Spk1 (Figs 3E, blue line, and S10B); thus, Ras1 plays an important role in activating both Cdc42 and MAPK^Spk1 pathways. Introduction of the MAPKK^byr1.DD mutation into the ras1∆ mutant cells induces the constitutive ppMAPK^Spk1 (Figs 3E, grey line and S10B) but does not affect the round cell morphology, and cells remain sterile (Fig 3G, the 2^nd left panel).

In striking contrast, the sterile phenotype of the MAPKKK^byr2∆, associated with a complete lack of shmoo formation (Fig 3G, the 2^nd right panel), was converted to the “paired” phenotype when combined with the MAPKK^byr1.DD mutation (Fig 3G, the far right panel). As expected, the MAPKKK^byr2∆ MAPKK^byr1.DD double mutant shows MAPK^Spk1 constitutive activation (Figs 3F and S10C). Thus, unlike the cases of scd1∆ or ras1∆, the lack of MAPKKK^byr2 can be bypassed by constitutive activation of MAPK^Spk1, indicating that the sole role of MAPKKK^Byr2 is to activate the MAPK^Spk1, unlike its upstream activator, Ras1, which also activates Cdc42 pathway.

It was also noted that the loss of MAPKKK^Byr2 completely blocked the MAPK^Spk1 production, whereas deletion of scd1 did not cause detectable changes to the total MAPK^Spk1 level and the ras1Δ cells still produced a detectable level of MAPK^Spk1 (S10D, S10E and S10F Fig).

Ras1.G17V causes accumulation of Cdc42-GTP at the cell cortex

Having observed a relatively mild influence of Ras1.G17V towards the MAPK^Spk1 activation profile, we next examined whether the Cdc42 pathway was affected by the ras1.G17V mutation. As previously observed, dynamic foci of CRIB-GFP appeared on the cell cortex upon induction of mating [59] (Fig 4A and 4B). In our experimental condition, more than 80% of the wildtype cells showed the cortical CRIB-GFP signal at 4.5 hours after induction of mating (Fig 4B). The cortical CRIB-GFP foci became concentrated at the mating site and quickly disappeared once cells fused successfully to form zygotes (Fig 4A and 4B). In striking contrast, in the ras1.G17V mutant cells, the cortical CRIB-GFP signal persisted, often at the elongated tip end of the cells, even 12.5 hours after induction of mating (Fig 4A and 4B). The signal could still be seen in about 40% of the cells 22.5 hours after induction of mating (Fig 4A and 4B). The result shows that the Cdc42 pathway is excessively activated in the ras1.G17V mutant, and the tip localisation of Cdc42^GTP indicates that the signature “elongated” ras1.G17V morphological phenotype is caused by hyperactivation of the Cdc42 pathway.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Ras1.G17V induces cortical Cdc42^GTP accumulation.

(A) Cell morphology and localisation of Cdc42^GTP, indicated by CRIB-GFP signal, during the sexual differentiation process. Wildtype (KT5077) and ras1.G17V (KT5082) mutant cells were induced for sexual differentiation by the plate mating assay condition (Materials and Methods), and live cell images were taken at the indicated time after the induction. Representative CRIB-GFP signal images are presented. Orange stars indicate cells with cortical CRIB-GFP foci. Rapidly disappearing CRIB-GFP signals at the fusion site of wildtype mating cells are indicated by green arrows at the time 8.5h image. The scale bar represents 10 µm. (B) Quantification of the results presented in (A). At each time point (4.5h, 6.5h, 10.5h, 12.5h and 22.5h after induction of sexual differentiation), 150 cells were examined to determine whether they have cortical CRIB-GFP foci. % cells with cortical CRIB-GFP foci are presented. The experiment was repeated three times, and the mean values and SDs were plotted in the graph.

https://doi.org/10.1371/journal.pgen.1012117.g004

The effect of ras1.G17V mutation towards Cdc42 activation was also observed during vegetative growth, producing the strongest cortical CRIB-GFP signal at the cell tip in the ras1.G17V mutant (S9 Fig), as previously reported [20]. The effect of ras1.G17V on cell morphology was best recognised when the rga4 that encodes a GTPase activation protein negatively regulating Cdc42 [52,54,60] was deleted, as the resultant rga4∆ ras1.G17V double mutant cells were larger and round with an increased CRIB-GFP signal (S9 Fig). These results fit the hypothesis that Ras1.G17V has an enhanced capability to activate Cdc42.

Two Ras effectors, MAPKKK^Byr2 and Cdc42-GEF^Scd1, compete with each other for Ras1

We next examined whether the two Ras effectors, MAPKKK^Byr2 and Cdc42-GEF^Scd1, compete with each other for Ras1. First, we overexpressed the Ras binding domains (RBDs) of MAPKKK^Byr2 and Cdc42-GEF^Scd1 in the ras1.G17V cells and examined the activation status of MAPK^Spk1 and Cdc42. For MAPKKK^Byr2, the region spanning the residues 65–180 (Byr2-RBD) was used following previous structural studies [61], and for Cdc42-GEF^Scd1, the region spanning the residues 760–872 (Scd1-PB1) was used based on its amino acid sequence similarity with the Ras Associating (RA) domain of mammalian RalGDS.

Overexpression of Byr2-RBD in the ras1.G17V cells substantially reduced the ppMAPK^Spk1 level, whereas the total MAPK^Spk1 level was only marginally affected (Figs 5A and S11), indicating that the Byr2-RBD competed against the endogenous full-length MAPKKK^Byr2 in activating the MAPK^Spk1. Strikingly, overexpression of Scd1-PB1 also inhibited the MAPK^Spk1 activation, exhibiting its capability to interfere with the MAPK^Spk1 pathway (Figs 5A and S11).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Two Ras effectors, MAPKKK^Byr2 and Cdc42-GEF^Scd1, compete each other for Ras1.

(A) Cells harboring ras1.G17V and MAPK^Spk1-GFP-2xFLAG (KT5940) were transformed with either pRep1 empty vector, pRep1-scd1(760-872)-2xFLAG or pRep1-byr2(65-180)-2xFLAG and cultured in MM + N without thiamine for 24 hours. Cells were induced for sexual differentiation by the plate mating assay system, and the levels of ppMAPK^Spk1-GFP, total MAPK^Spk1 and α-tubulin (loading control) were examined by Western blotting. Three biological replicates, shown in S13 Fig, were quantified using the Image Studio ver2.1 software (Licor Odyssey CLx Scanner). The data was analyzed by generalized linear mixed models and their least-squares means (± 1SE) are shown. Different letters (a, b and c) indicate significant differences (α = 0.05) among the three samples, pRep1-empty, pRep1-scd1-PB1 and pRep1-byr2-RBD, within each hour after induction of mating. When all three samples were deemed distinct, they were denoted differently: a, b and c. When all the samples were deemed indistinguishable, they were denoted with the same category: a. (B) Cell morphology and localisation of Cdc42^GTP, indicated by the CRIB-GFP signal, during the sexual differentiation process were compromised by the expression of Scd1-PB1 or Byr2-RBD. Cells harbouring ras1.G17V and CRIB-GFP (KT5938) were induced for mating/sexual differentiation by the plate mating assay condition (Materials and Methods), and live cell images of the CRIB-GFP signals were taken 16 hours after induction of the sexual differentiation. Navy arrows indicate cortical CRIB-GFP foci. Cells with the cortical CRIB-GFP foci (marked by an orange star) are counted as “cortical CRIB-GFP positive” cells. The scale bar is 10 µm. (C) Quantitation of the results presented in (B). 300 cells were scored for cell deformity (left panel) and the cortical CRIB-GFP signal (right panel). % cells is presented. The experiment was repeated three times, and the mean and SD values of the % cells are plotted in the graphs. Ordinary one-way ANOVA followed by post hoc Dunnett’s multiple comparisons test showed that the expression of scd1-PB1 or byr2-RBD reduces the % deformed cells and % cells with cortical CRIB-GFP. (D) GTP-loaded Ras1.G17V (1-172) directly binds to Byr2 (65-180) and Scd1 (760-872). In vitro GST pull-down assays of bacterially expressed Ras1.G17V (1-172), GST-Byr2 (65-180) and GST-Scd1 (760-872) were conducted as described in materials and methods. GTP-loaded Ras1.G17V (1-172) bound to both GST-Byr2 (65-180) and GST-Scd1 (760-872). (E) Two Ras1 effectors, Byr2 and Scd1, compete for GTP-loaded Ras1.G17V (1-172). In vitro GST pull-down assays of bacterially expressed Ras1.G17V (1-172) and GST-Byr2 (65-180) were conducted as in (D). The addition of the Scd1 (760-872) fragment interfered with Ras1-Byr2 binding (the 4^th lane). Quantitated signal intensities of the Ras1.G17V (1-172) band in the gel are shown in the right panel.

https://doi.org/10.1371/journal.pgen.1012117.g005

When Cdc42 activation was examined in the ras1.G17V cells, overexpression of Scd1-PB1 reduced the cortical CRIB-GFP signal and abolished deformed cells (Fig 5B and 5C), indicating that the endogenous Scd1 function was interfered with the over-expressed Scd1-PB1. Byr2-RBD overexpression also resulted in a comparable phenotype (Fig 5B and 5C). Collectively, two Ras1 effectors, MAPKKK^Byr2 and Cdc42-GEF^Scd1, are capable of competing with each other for Ras1 in vivo when overexpressed. The effect of overexpression of Scd1-PB1 and Byr2-RBD on cell morphology was also observed in vegetatively growing cells, consistent with the Ras1 function of maintaining the cell morphology of vegetative cells (S12 Fig).

Next, we conducted in vitro binding competition assays using bacterially expressed GST-tagged fragments of Byr2-RBD and Scd1-PB1, both of which bound Ras1.G17V (1–172) loaded with GTP (Fig 5D). The binding of Ras1.G17V^GTP to GST-Byr2-RBD was substantially decreased when the GST-Scd1-PB1 was added (Fig 5E). The result indicates the biochemical competitive nature of MAPKKK^Byr2 and Cdc42-GEF^Scd1 for active Ras1.

We also probed the in vivo status of MAPKKK^Byr2 and Cdc42-GEF^Scd1 during the mating process by tagging these proteins with a 1xGFP tag, which may produce minimal artefact compared to a more sensitive 3xGFP. Scd1 was tagged at its C-terminal end, and MAPKKK^Byr2 was tagged at its N-terminal end, as C-terminally tagged MAPKKK^Byr2 cells became sterile (S13A Fig). In vegetatively growing cells, the Scd1-GFP signal was detected at the growing cell tips as previously described [62], whilst the GFP-Byr2 signal was undetectable (S13B Fig). In cells undergoing sexual differentiation, we could detect both Scd1-GFP and GFP-Byr2 signals at the shmoo tip of mating cells (S13C Fig). As both signal intensities were low, rigorous quantification could not be performed. However, both signals were within a similar intensity range, and it is unlikely that one is in excess of the other; hence, a competition between MAPKKK^Byr2 and Cdc42-GEF^Scd1 might exist, provided that the number of available Ras1 molecules is limited.

Ras1 and an adaptor protein Ste4 are both necessary to fully activate MAPKKK^Byr2

Although Ras1 plays a significant role in activating MAPK^Spk1, a low but detectable level of ppMAPK^Spk1 was still induced in the ras1∆ mutant (Figs 3E and S10B), indicating that there is a Ras1-independent mechanism to activate MAPK^Spk1. Previous studies proposed that the adaptor protein Ste4 is involved in the activation of MAPKKK^Byr2 [33,39,40,63]. We therefore examined whether Ste4 is required for MAPK^Spk1 activation.

As was the case for the byr2Δ (S10F Fig), the loss of ste4 abolished the expression of MAPK^Spk1 (S14C Fig), and we detected virtually no MAPK^Spk1 phosphorylation in the ste4∆ mutant (Figs 6A, yellow line and S14A), revealing that the adaptor^Ste4 is a prerequisite for the MAPK^Spk1 activation. The introduction of the ras1.G17V mutation neither restored the MAPK^Spk1 expression nor activation, nor mating (Figs 6A, 6B, and S14C). Thus, activation of Ras1 cannot take over Ste4 function. In a striking contrast, the ste4∆ MAPKK^byr1.DD double mutant induces the expression of MAPK^Spk1, constitutive MAPK^Spk1 activation and the “paired” phenotype as the MAPKK^byr1.DD single mutant cells (Figs 6A, 6B, and S14C). Collectively, a full MAPKKK^Byr2 activation requires both Ras1 and Ste4 functions.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Distinct contributions of Ste4 and Ste6 to MAPK^Spk1 activation.

(A) Ste4 is essential for MAPK^Spk1 activation. MAPK^Spk1 phosphorylation status in ste4Δ (KT4376), ste4Δ ras1.G17V (KT5143) and ste4Δ MAPKK^byr1.DD (KT5136) at times-points 0, 8, 16 and 24 hours after the induction of sexual differentiation were examined. Quantified results of three biological replicates (error bars are ± SEM) are presented. The ppMAPK^Spk1 Western blotting result of Biological Replicate 1 is presented as a representative result where the ppMAPK^Spk1 levels are close to 0 in the ste4Δ (KT4376) and ste4Δ ras1.G17V (KT5143) strains. All the original Western blotting membranes are presented in S12A Fig. (B) The MAPKK^byr1.DD but not ras1.G17V mutation suppresses the incapability of ste4Δ to cause pheromone-induced morphological change. Cell images of ste4Δ (KT4376), ste4Δ ras1.G17V (KT5143) and ste4Δ MAPKK^byr1.DD (KT5136) strains were taken 24 hours after induction of mating. A yellow arrow indicates a typical “paired” phenotype example. Blue arrow indicates accumulated vacuoles at the conjugation tips as seen in the MAPKK^byr1.DD mutant (Fig 1F). The scale bar represents 10 µm. (C) Lack of Ste6 does not result in the complete loss of MAPK^Spk1 phosphorylation. MAPK^Spk1 phosphorylation status in ste6Δ (KT4333), ste6Δ ras1.G17V (KT4998) and ste6Δ MAPKK^byr1.DD (KT5139) at times-points 0, 8, 16 and 24 hours after the induction of sexual differentiation were examined. Quantified results of three biological replicates (error bars are ± SEM) are presented. The ppMAPK^Spk1 Western blotting result of Biological Replicate 1 is presented as a representative result where the ppMAPK^Spk1 level of the ste6Δ (KT4333) is clearly detectable. All the original Western blotting membranes are presented in S12B Fig. (D) The MAPKK^byr1.DD but not ras1.G17V mutation suppresses the incapability of ste6Δ to cause pheromone-induced morphological change. Cell images of ste6Δ (KT4333), ste6Δ ras1.G17V (KT4998) and ste6Δ MAPKK^byr1.DD (KT5139) strains were taken 24 hours after induction of mating. The scale bar represents 10 µm.

https://doi.org/10.1371/journal.pgen.1012117.g006

Interestingly, Ste4’s role is limited to the MAPKKK^Byr2 activation process as MAPKKK^Byr1.DD can suppress the ste4∆ mutant phenotype, whereas MAPKKK^Byr1.DD could not bypass the Ras1 function (Fig 3G).

Ste6, a Ras1 GTP-GDP exchange factor, contributes to both the MAPK^Spk1 and the Cdc42 pathway activation

The activation of Ras1 is mediated by two GDP-GTP exchange factors (GEFs), Ste6 and Efc25 [64,65]. Ste6 is essential for mating but is dispensable during vegetative growth, whereas Efc25 is dispensable for mating but is required to maintain cell morphology during vegetative growth [64,65]. There has been an interesting proposition that Ste6 may specifically help Ras1 to activate the MAPK^Spk1 pathway but not the Cdc42 pathway, whilst Efc25 specifically facilitates Ras1 to activate the Cdc42 pathway [66]. We examined this hypothesis by monitoring the MAPK^Spk1 activation status and conducting genetic epistasis analysis of ras1.G17V and MAPKK^byr1.DD in the ste6∆ mutant.

In the ste6∆ cells, MAPK^Spk1 phosphorylation occurred at a detectable level (Figs 6C yellow line and S14B). When the MAPKK^byr1.DD mutation was introduced, the MAPK^Spk1 phosphorylation level was maximised (Figs 6C and S14B) and the nucleus MAPK^Spk1-GFP signal accumulated (Fig 6D). However, the sterile phenotype remained (Fig 6D). In contrast, the ras1.G17V mutation rescued the “pheromone-insensitive sterile” morphology of ste6∆, as previously reported [64], exhibiting the “elongated” phenotype (Fig 6D) even though the increase of the ppMAPK^SPK1 accumulation was not as high as the case of the MAPKK^byr1.DD mutation (Figs 6C and S14B). The result indicates that, unlike the ste4∆ mutant, the mating deficiency of ste6∆ is not caused by a mere lack of MAPK^Spk1 activation but by a lack of Ras1 activation, which mediates both the MAPK^Spk1 and Cdc42 pathway activation in response to the pheromone signalling.

Activation of Gpa1 represents the pheromone signalling

Gpa1, which plays the primary role in pheromone signalling [21], is expected to act upstream of Ste4 and Ste6. In agreement, in the gpa1Δ mutant, we detected no MAPK^Spk1 activation nor morphological change (Figs 7A, yellow line, 7B, left panel, and S15A). The ras1.G17V did not rescue the lack of MAPK^Spk1 activation nor did it induce a shmoo-like morphological change (Figs 7A, grey line, 7B, the second right panel, and S15A), supporting our earlier observation that a Ras1-independent mechanism, involving Ste4, is essential for MAPK^Spk1 activation. Meanwhile, the MAPKK^byr1.DD caused the constitutive activation of MAPK^Spk1 (Figs 7A, blue line and S15A), but cells showed no morphological change (Fig 7B, the second left panel). When both ras1.G17V and MAPKK^byr1.DD mutations were introduced into the gpa1∆ strain, MAPK^Spk1 was activated, and a shmoo-like morphological change occurred (Figs 7A, red line, 7B, the right panel and S15A). Therefore, the activation of both the MAPK^Spk1 pathway and the Ras1 pathway is sufficient to take over the Gpa1 function.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Gpa1 plays the central role of the pheromone signal transduction by activating both MAPK^Spk1 and Ras1 pathways.

(A) MAPK^Spk1 phosphorylation status in homothallic gpa1Δ (KT4335), gpa1Δ ras1.G17V (KT5023), gpa1Δ MAPKK^byr1.DD (KT4353) and gpa1Δ ras1.val17 MAPKK^byr1.DD (KT5035) at times-points 0, 8, 12, 16 and 24 hours after the induction of sexual differentiation were examined. Results of three biological replicates (error bars are ± SEM) are presented. Original Western blotting membranes are presented in S15A Fig. (B) Cell images of the above-mentioned strains at 16 hours after the induction of sexual differentiation. All the cell images were taken and processed as in Fig 1. The scale bar represents 10 µm. (C) MAPK^Spk1 phosphorylation status in h^- WT (KT4190), h^- gpa1.QL (KT5059), h^- ras1.G17V (KT4233), h^- gpa1.QL ras1Δ (KT5070) and h^- MAPKK^byr1.DD (KT4194) at times-points 0, 8, 12 and 24 after the induction of sexual differentiation were examined. Results of three biological replicates (error bars are ± SEM) are presented. Original Western blotting membranes are presented in S15B Fig. (D) Cell images of the above strains at 12 h after the induction of sexual differentiation. All the cell images were taken and processed as in Fig 1. The scale bar represents 10 µm.

https://doi.org/10.1371/journal.pgen.1012117.g007

The signalling hierarchy of Gpa1, Ras1 and the MAPK^Spk1 cascade was further validated in a heterothallic h^- cell population. This population serves as a reference for measuring responses to nitrogen starvation without mating factor signalling as it lacks a mating partner and, consequently, the mating pheromone [41]. Interestingly, upon nitrogen starvation, the h^- wildtype cells displayed a basal level of ppMAPK^Spk1 activation, although no morphological changes were observed (Figs 7C, yellow line, 7D, the left panel, and S15B). In contrast, h^- cells with the constitutively active gpa1.QL mutation exhibited a “shmoo-like” morphological change as previously reported [21] and showed strong MAPK^Spk1 activation (Figs 7C, red line, 7D, the second left panel and S15B). Conversely, the h^- ras1.G17V mutant showed no apparent morphological alternation and exhibited only a basal level of ppMAPK^Spk1, comparable to that observed in the h^- wildtype strain (Figs 7C, grey line, 7D, the middle panel, and S15B). Meanwhile, the h^- MAPKK^byr1.DD mutant induced strong constitutive MAPK^Spk1 activation, confirming that the MAPKK^Byr1.DD can activate MAPK^Spk1 regardless of the pheromone signal input (Figs 7C, blue line and S15B). Yet, the cell morphology remained unchanged (Fig 7D, the second right panel). Collectively, these results support the model in which Gpa1 serves as the central transducer of the pheromone signalling. It was noted that the Gpa1^QL-induced phenotype was largely dependent on Ras1 function as the h^- gpa1.QL ras1∆ double mutant exhibited only a basal level of ppMAPK^Spk1, and round cell morphology (Figs 7C, purple line, 7D, the right panel and S15B).

Delayed negative feedback explains the transient increase of the total and phosphorylated MAPK^Spk1 in ordinary differential equations-based mathematical modelling

Our results, along with previous findings, can be summarised in a diagram illustrating that the pheromone signalling activates Ras1, leading to the activation of two effectors, MAPKKK^Byr2 and Cdc42-GEF^Scd1 (Fig 8A). In addition, the MAPKKK^Byr2 activation also requires Ste4 function, and a reciprocal activation occurs between MAPK^Spk1 and Cdc42 for their full activation (Fig 8A).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 8. Mathematical modelling of the fission yeast pheromone signalling dynamics.

(A) Schematic diagram of the fission yeast pheromone signalling pathway. The diagram highlights the bipartite mechanism to activate MAPKKK^Byr [2], the branch for two Ras1 effectors, MAPKKK^Byr2 and Cdc42GEF^Scd1, and the interplay between the MAPK^Spk1 and Cdc42 pathways. Successful mating requires both the MAPK^Spk1 and Cdc42 pathway activation. (B) Model B components and framework. For the detailed implementation of the mutants, see S16 Fig and Materials and Methods. The measured components, total MAPK^Spk1, ppMAPK^Spk1 and Cdc42^GTP, are shown in purple, green and red, respectively. (C) Model B fittings for the ppMAPK^Spk1 levels in wildtype, ras1.G17V, MAPKK^byr1.DD, ras1.G17V MAPKK^byr1.DD and Cdc42GEF^scd1Δ mutants. Mean values (open circles) and SD values (error bars) of the experimental results are shown in pale colours, and the model-fitted values are shown as filled circles in darker colours. Model B fittings for the total MAPK^Spk1 are shown in S18 Fig. (D) Model B fitting of the active Cdc42 levels in wildtype and ras1.G17V mutants. Mean values (open circles) and SD values (error bars) of the experimental results are shown in pale colours, and the model-fitted values are shown in filled circles in darker colours.

https://doi.org/10.1371/journal.pgen.1012117.g008

We next aimed to build a network structure that explains the experimental data we obtained. We are aware that whilst the wildtype data represents the entire sexual differentiation process, including G1 arrest, pheromone sensing, mating, followed by DNA recombination and sporulation, all the sterile mutants are blocked at the stage before successfully completing mating. Therefore, we need to be cautious when interpreting differences in the later timepoint data. However, we still decided to include all time points in our attempt to “describe” our experimental results using a single signalling framework. Towards this goal, we sought to determine whether the experimentally observed temporal profile of total MAPK^Spk1 ([tSpk1]), phosphorylated MAPK^Spk1 ([ppSpk1]) and active Cdc42 ([aCdc42]) (Figs 1C, E, G, 2A, 3B, 4 and S3)–in particular, the transient increase of ppSpk1 followed by attenuation–can be explained by specific network structures composed of the regulatory interactions identified and characterized in this study (see Material and Methods section and S16 and S17 Figs for details).

To address this question, we postulated up to 29 biochemical processes, each represented by a rate parameter k, to capture the documented relationships among the signalling components (S1 Table and Figs 8B, S16, S17A and S19A; see Material and Methods for details). Because many kinetic parameters of the pheromone signaling pathway have not been experimentally determined in fission yeast, model parameters were treated as effective quantities representing coarse-grained regulatory influences rather than elementary biochemical reaction rates, and the reported values therefore represent best-fit examples rather than unique or physiologically exact intracellular parameters. Initial parameter values were obtained using the MATLAB Optimization Toolbox (Mathworks, Natick, MA, USA) and were subsequently refined using a Markov chain Monte Carlo (MCMC) approach to improve agreement with the experimental data [67]. Importantly, the purpose of the modelling in this study is not to infer biologically exact kinetic parameter values, but to evaluate alternative candidate network structures based on their ability to account for this characteristic transient MAPK^Spk1 response, as well as the experimentally measured temporal dynamics of total Spk1 and active Cdc42.

Our initial model, Model A, did not yield a set of parameters that aligned with the experimental results (S17 Fig); [tSpk1] and [ppSpk1] did not decrease following an initial increase. To replicate the transient increase of [tSpk1] and [ppSpk1], a delayed negative feedback regulation was deemed necessary [68]. Given that the MAPKK^byr1.DD mutant entirely lacks downregulation (Fig 1E), we considered downregulations occurring downstream of MAPKK^byr1 (such as Pyp1 and Pmp1) to be physiologically insignificant. On the other hand, Sxa2 (a serine carboxypeptidase against a mating pheromone P-factor) and Rgs1 (a regulator of Gpa1), both of which are induced upon successful pheromone signalling [41,69–71], receptor internalization [72] and regulation of the mapk^spk1 transcript or other components by antisense RNA [73] fit well to the criteria for the delayed negative feedback. We consolidated all these elements collectively as a single circuit and provided it with the rate constant k₆ (Model B, Fig 8B). Incorporating this negative feedback allowed us to obtain a set of parameters that effectively reproduced the transient increase of [tSpk1] and [ppSpk1] in wildtype, ras1.G17V and scd1Δ condition (Figs 8C and S18). Furthermore, the fit for the active Cdc42 was also significantly improved (Fig 8D).

Additional regulations to further improve the model fit to the experimental results

Model B phenocopied the experimental results of tSpk1, ppSpk1 and active Cdc42 in wildtype and scd1Δ strains and qualitatively reproduced most data points in other strains. Notably, whilst Model A failed to reproduce the transient increase and subsequent attenuation of MAPK^Spk1 for any parameter set tested, Model B successfully reproduced these dynamics, indicating that the difference between the models is structural rather than parametric. However, Model B did not predict the early increase of the ppSpk1 levels in the ras1.G17V and ras1.G17V MAPKK^byr1.DD strains (Fig 8C).

We revised Model B to generate Model C, where we introduced two aspects (S19 Fig). First, we added ras1.G17V strain-specific negative regulatory paths from ppSpk1 to modulate k₉ and k₁₄, the rate constants of two Ras1 downstream reactions. Second, to distinguish the outcomes between MAPKK^byr1.DD and MAPKK^byr1.DD ras1.G17V mutants, we added a positive regulation of ppSpk1 by Cdc42 (observed in the Cdc42-GEF^scd1Δ mutant, Fig 3B) with a rate constant k₂₇. As a result, model C reproduced both (i) the earlier ppSpk1 peak with the height comparable to the WT in the ras1.G17V, and (ii) the difference between MAPKK^byr1.DD and MAPKK^byr1.DD ras1.G17V strain (S19C Fig). The modelling analyses suggested that the constitutively active form of Ras1, Ras1.G17V, may have a qualitatively different biochemical activity compared to the wildtype Ras1.

Prediction ability of our quantitative model

To examine the validity of Models B and C, we tested whether the models can predict the ppSpk1 levels of other experimental results, presented in Figs 3E, 3F, 6A, 6C, 7A and 7C, involving 20 strains of different genotypes (S20 Fig). Both Models B and C faithfully simulated Figs 3F and 6A, involving MAPKK^byr1.DD, byr2Δ, ste4Δ and ras1.G17V. Model B also produced a simulated outcome highly comparable to the results presented in Figs 3E and 7A, involving MAPKK^byr1.DD, ras1Δ, gpa1Δ, and ras1.G17V, whereas Model C’s predictions for these experiments were less quantitative. On the other hand, Model C predicted Fig 7C results, involving heterothallic strains of wild type, MAPKK^byr1.DD, gpa1.QL, ras1.G17V and ras1Δ, better than Model B.

It was noted that both Model B and C predictions for Fig 6C experiment involving the strains with the ste6Δ mutation were the least accurate among all the tested experiments, although Model B correctly predicted high ppSpk1 for ste6Δ MAPKKK^byr1.DD and low ppSPK1 for ste6Δ and ste6Δ ras1.G17V at a later stage of about 15 hours after induction of mating and Model C predicted a low ppSpk1 level in ste6Δ, and a high ppSpk1 level in ste6Δ MAPKKK^byr1.DD.

In summary, both models predicted the dynamics of ppSpk1 to a good extent in the 20 strains. However, Model C did not show a significantly improved prediction capability compared to Model B, despite two additional modelling parameters, k₂₈ and k₂₉. Therefore, the simpler Model B would represent the core framework of fission yeast pheromone signalling.

The predicted negative feedback was experimentally validated in the sxa2Δ mutant cells

Our model predicts that a negative feedback circuit, impacting the pheromone production and/or sensing, plays a critical role in shaping the MAPK activation profile. One possible candidate to participate in such negative feedback is Sxa2, a carboxypeptidase that modulates the local concentration of one of the mating factors, P-factor [69,74]. We examined the effect of sxa2 deletion in the ras1.G17V and byr1.DD mutant cells.

Strikingly, the ppMAPK^Spk1 level was substantially increased in the sxa2Δ ras1.G17V double mutant compared to the ras1.G17V single mutant (S21 Fig), validating the model prediction. As the MAPK^Spk1 activation profile of the sxa2Δ ras1.G17V double mutant still exhibits a peak (at 4 h) followed by a decline, rather than becoming constitutively active, additional mechanism(s) still exist to fully compose the negative feedback circuit. Nonetheless, the result clearly demonstrates that Sxa2 plays a critical role, and modulating signal production is an effective way to shape MAPK^Spk1 signalling output even in the presence of the ras1.G17V mutation.

Intriguingly, the deletion of sxa2 in the byr1.DD mutant did not substantially alter ppMAPK^Spk1 levels, but did induced the “elongated” morphology (S22 Fig). The “elongated” phenotype, characteristic of the ras1.G17V mutant that exhibits Cdc42 hyperactivation (Fig 4), might suggest that compromised negative feedback might have resulted in constitutively active Cdc42, as predicted for the wildtype case in Model A, which lacks the negative feedback (S17D Fig). These observations collectively support the existence of the predicted negative feedback mechanism.

Yeast strains and media

Genotypes of the Schizosaccharomyces pombe strains used are listed in the S2 Table. Gene disruption and 2xFLAG-GFP-tagging of genes were performed using the direct chromosomal integration method described previously [92–94]. Fission yeast media (YE, MM-N and SPA) and basic genetic manipulations are described in [95].

Plate mating assay for synchronous mating

In order to induce synchronous mating, we established the “Plate mating assay” system as follows: Cells were grown in YE supplemented with adenine (0.2g/L) until the cell density reaches 8–10 x 10⁶ cells/ml. Cells were then washed with MM-N (1% glucose, supplemented with Leucine 40mg/L) using a filtration unit. These cells were resuspended to a cell density of 0.8-1x10⁷cells/ml in MM-N. In the timecourse experiments, this moment was considered time 0. 8x10⁷cells were re-filtered onto a PDVF membrane of 47 mm diameter (Millipore DVPP04700) in order to evenly place the cells across the defined area. The filters bearing the cells over their surface were carefully transferred onto sporulation agar plates (SPA+Leucine 40mg/L, 50ml SPA agar/ 14 cm diameter for 6 membranes), cell side up, and incubated at 30˚C. During the time course experiments, at each time-point, one membrane, which initially had 8x10⁷cells at time 0, was removed from the SPA plate and placed into 5 ml of ice-cold 20% trichloroacetic acid (TCA) in a 50 ml falcon tube. The tube was shaken vigorously to remove cells from the membrane. The tube was briefly centrifuged, and the membrane was removed from the tube. The tube was then centrifuged for 1 minute at 2000rpm to pellet cells. The supernatant was discarded, cells were resuspended in 1 ml 20% TCA and transferred to a tube capable of storage at -80˚C. The cells were pelleted again using a bench top centrifuge, supernatant discarded, and the pellet snap frozen in liquid nitrogen. Tubes were stored at -80˚C, ready for protein extraction. A cartoon representation of the outline of the plate mating assay procedure is presented in S1A Fig.

Preparation of whole cell extracts

This method is based on that described by Keogh lab Protocols (https://sites.google.com/site/mckeogh2/protocols) with a modification to use Urea buffer at the final step to increase protein solubility. All steps were performed on ice unless otherwise stated. Cell pellets (typically 8x10⁷cells) were thawed on ice and resuspended in 250 µl 20% TCA. 250 µl of acid washed glass beads (SIGMA) were added and sampled chilled on ice for 5 minutes. Cells were broken in a FastPrep24 cell beater (MP Biomedicals, speed 6.5 x 1 minute x 3 with 1 minute intervals). Cell extracts were collected by piercing the bottom of tubes with a needle and placing them into collection tubes and centrifuged at 2000 rpm for 1 minute. Beads were washed with 300 µl of 5% TCA and centrifuged at 2000 rpm for 1 minute, this was repeated twice. The contents of the collection tube were transferred to a 1.5 ml tube and centrifuged at 14K rpm for 10 minutes at 4^oC. The liquid was discarded and the cell pellet washed with 750 µl of 100% ethanol. Tubes were centrifuged briefly (~1min 14000rpm) again. Ethanol was completely removed with a pipette. The pellet was then resuspended in 100 µl of Urea buffer (8M urea, 500 mM Tris pH9.4, 1% SDS, 5mM EDTA). The sample was centrifuged and 40 ul of the supernatant was mixed with 80 µl of 3x protein loading buffer [96]. The sample was then heated for 5 minutes at 95˚C and stored at -80˚C until use.

Western blotting

Protein extracts were subject to SDS-PAGE and were transferred to Immobilon-FL PVDF membrane (Millipore). Membranes were blocked for a minimum of 1-hour in Odyssey Blocking Buffer (OBB) (Li-cor) diluted 1:1 in PBS. Primary antibody incubation was carried out overnight in OBB 1:1 in PBS with anti phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204) rabbit monoclonal antibody (#4370 Cell Signaling Technology. Used at 1:4000 dilution), anti GFP antibody ((0.4mg/ml), Roche Cat No 11814460001. Used at a 1:2000 dilution), anti FLAG M2 antibody (Sigma F1804, 1μg/μl, used at 1:2000 dilution), anti actin antibody (Life Technologies MA1–744; mouse monocolonal RRID:AB_2223496, 1:2000 dilution) and anti α-tubulin antibody, TAT1 (generous gift from Keith Gull, 1:3000). Membranes were washed 3x10 minutes with 20 ml TBST (50 mM Tris, pH 7.4 – 150 mM NaCl – 0.1% Tween 20) whilst gently shaking followed by secondary antibody incubation (IRDye 680LT goat anti-mouse antibody, Li-cor 926–32211(1.0 mg/ml), 1:16,000 dilution and IRDye 800CW goat anti-rabbit secondary antibody Li-cor 926–68020 (1.0 mg/ml), 1:16000 dilution in 0.01% SDS, 0.1% Tween 20 in OBB 1:1 PBS) for 1 hour in the dark to prevent fluorophore bleaching. This was followed by 2x 10 minute TBST washes and a 1x 10 minute TBS (50 mM Tris, pH 7.4 – 150 mM NaCl) wash before scanning on the Odyssey CLx infrared Imaging System (Li-cor). Protein quantitation was conducted using either the Image Studio V2.1 software supplied with the Odyssey CLx scanner or Fiji (Image J, [97]).

Quantitation of MAPK^Spk1 phosphorylation

To detect the activated MAPK^Spk1, whole cell extracts were prepared and Western blotting analysis was carried out using the anti-phospho ERK antibody (#4370 Cell Signaling Technology) that recognises the dual phosphorylation at the conserved TEY motif. Specificity of the antibody against the phosphorylated MAPK^Spk1 (ppMAPK^Spk1) in the fission yeast cell extracts was confirmed before proceeding with further experiments as shown in S1 Fig. As described above in the “Plate mating assay for synchronous mating”, during the time course experiments, at each time-point, one membrane, which initially had 8x10⁷cells at time 0, was used to prepare the whole cell extracts. Then an equal volume of each sample was loaded to the Western blotting. In this manner, we observed that the change of the protein level throughout the mating process per a defined starting material. Recent proteiomis studies showed that during the sexual differentiation, both actin and α-tubulin levels stay constant until sporulation starts [98]. During the sporulation process, actin stays constant whereas the level of α-tubulin decreases [98]. Therefore, we used actin as an interanal control when the Western blotting involves wildtype cells, which undergo sporulation. We also used α-tubulin as an internal control when sporulation deficient mutant strains were compared each other. The benefit of α-tubulin was that the signal intensity was stronger and therefore, quantitation can be more sensitive. For all Western blotting analysis, three biological replicates of time-course experiments were used.

Cell imaging

A Zeiss LSM 980 Airyscan 2 microscope equipped with Plan-Apochromat 63x/1.40 Oil DIC f/ELYRA objective in LSM confocal mode was used for cell imaging to capture the morphology of cells and the localisation of the GFP-tagged MAPK^Spk1 during mating time-courses. Each time point comprises 32 serial images with 0.24 µm intervals along the Z axis taken to span the full thickness of the cell. Images were cropped and combined using Fiji [97].

For a higher resolution images used to quantify the GFP-tagged MAPK^Spk1 singals (presented in S4 Fig), the Zeiss LSM 980 Airyscan 2 microscope was used in Airyscan multiplex-4Y mode. Each image comprises 47 serial images with 0.15 µm intervals along the Z axis taken to span the full thickness of the cell.

For cell images presented in Figs 4A, 5, S5 and S8, a 2D array scanning laser confocal microscope (Infinity 3, VisiTech) on a NikonTi-E microscope stand equipped with a Hamamatsu Flash 4.0V2 sCMOS camera and a Plan Apo 100x/1.45 objective was used. Each time point comprises 24 serial images with 0.25 µm intervals along the Z axis taken to span the full thickness of the cell. All of the Spk1-GFP images were deconvolved using the Huygens Essential Deconvolution software (Scientific Volume Imaging). Deconvolved images were cropped and combined using Fiji [97].

Quantification of GFP-tagged MAPK^Spk1 signals

In order to quantify the MAPK^Spk1-GFP signals throughout the sexual differentiation stages of wildtype cells, cells carrying the MAPKSpk1-GFP and Smd2-tdTomato were imaged using the Zeiss LSM 980 Airyscan 2 microscope in Airyscan multiplex-4Y mode. These cells were induced for sexual differentiation for at least 8 hours, so that the GFP signals were clearly detectable and could be used to define the total area, comprising of both cytoplasm and nucleus. To define the area of the nucleus, tdTomato-tagged Smd2 (small nuclear ribonucleoprotein) was used. Using these two signals, 3D image segmentation was conducted using Imaris (Oxford Instruments) to generate two segments, Total and Nucleus. We then measured the MAPK^Spk1-GFP signal intensity in each segment. The process was summarised in S4B Fig.

Quantification of CRIB-GFP signal

CRIB-GFP signal was used to examine activation status of Cdc42. For images presented in Fig 4A, 25 serial Z images with step size of 0.25 µm were taken. Images were deconvolved using Huygens Essential (Scientific Volume Imaging) and maximum intensity projections are presented. Deconvolved images were Z-projected (maximum intensity), cropped and combined using Fiji [97].

For images presented in S9A Fig, for each image, 30–45 serial images with 0.2 µm interval along the Z axis were taken to span the full thickness of the cell. Using FIJI software maximum intensity Z-projections were created and used for quantitation. Intensity of GFP signal on the cell cortex was measured using FIJI, by applying line measurement with a line width of 1 μm which traced the inner cell cortex. The measurement was done along one of the cell tips that shows a stronger GFP signal as indicated in an example image in S9B Fig. 40 cells without septum were measured for each strain. Quantified CRIB-GFP intensity traces were analysed in R (R Core Team http://www.R-project.org/). First, rolling averages over +/-5 measurement points were calculated. Next, the X-coordinates of these traces were aligned by their peak intensity. Finally, the average curve from all aligned traces per strain was calculated, and displayed in S9B Fig with respective standard error of the mean curves (dashed lines).

Glutathione-S-transferase (GST) tag pull-down assays

Fission yeast cDNA fragments encoding Ras1.G17V (1–172), Byr2 (65–180) and Scd1 (760–872) were cloned in the pLEICS1 (Ras1.G17V) and pLEICS4 (Byr2 and Scd1) vectors (Protex [Protein Expression Laboratory], University of Leicester) to produce N-terminally His6-tagged Ras1.G17V (1–172), N-terminally GST-tagged Byr2 (65–180) and N-terminally GST-tagged Scd1 (760–872). Constructs were expressed in E.coli BL21 Rosetta cells in LB. All bacteria cell lysates were prepared in phosphate-buffered saline (PBS), pH 7.4.

His6-Ras1.G17V (1–172) was purified using Ni Sepharose 6 Fast Flow (GE Healthcare) and dialysed into buffer A (20 mM Tris-Cl (pH 7.5), 100 mM NaCl, 5mM MgCl2, 1mM β-mercaptoethanol). For GTP loading, buffer A of the Ras1.G17V (1–172) preparation was firstly replaced by Exchange buffer (20 mM Tris-Cl (pH 7.5), 1mM EDTA, 1mM β-mercaptoethanol) using a filtration unit (Amicon Ultra Centrifugal Filter Unit, 10K cut off, Millipore). GDP-GTP exchange reaction was carried out by adding GTP and EDTA to a final concentration of 8 mM and 12 mM respectively and incubating the sample at 37^oC for 10 min. The exchange reaction was terminated by adding MgCl2 to a final concentration of 20.5 mM. The resultant GTP-loaded Ras1.G17V (1–172) was washed with buffer A using the filtration unit and immediately used for the GST pull-down assays.

GST-Byr2 (65–180) and GST-Scd1 (760–872) were purified using glutathione sepharose 4B (GE Healthcare). The proteins bound on the glutathione beads were washed with buffer A and used for the GST pull-down assays together with the Ras1.G17V (1–172) preparation. For the competition assays shown in Fig 5E, Scd1 (760–872) fragment was prepared by cleaving it out from the N-terminal GST, which was bound to the glutathione beads, using TEV protease.

GST pull-down assays were carried out by incubating Ras1.G17V (1–172) (GDP bound or GTP bound) and GST-Byr2 (65–180) or GST-Scd1 (760–872) with or without Scd1 (760–872) fragment at 4 °C for 30 min. Supernatant of the reaction was saved and mixed with the same volume of the 3x loading buffer [96] to generate the “unbound” fraction. The remaining beads were washed twice, resuspended in buffer A of the original volume and mixed with the same volume of the 3x loading buffer to generate the “bound” fraction. Protein samples were run on a SDS-PAGE gel and proteins were visualised by InstantBlue protein stain (Expedeon). Quantitation of the intensities of Ras1.G17V (1–172) bands in Fig 5E was carried out using Fiji software.

Statistical analysis

Western blotting data presented in S2 Fig was quantified as stated in the “Western blotting” section, and individual replicate results are presented in S3A, S3B and S3C Fig. The mean (± SD) of the ppSpk1/actin triplicate data is presented in Figs 1C, 1E, 1G, 2A, and 3B. The data was analyzed by the generalized additive mixed model (GAMM) by fitting GAMMs on temporal variations in the data points (ppSpk1/actin, total Spk1/actin, and ppSpk1/total Spk1) for individual strain using the mgcv package in R. We assumed that the data were distributed according to a normal distribution (identity link). To account for the repeated measurements of data, the models included the genotype of each strain as a random factor. The smoothed curves were back-transformed from the GAMMs and drawn in the figures with twofold standard errors (S3D Fig).

Western blotting data presented in S13 Fig, which was quantified and presented in Fig 5A, was analyzed by the generalized linear mixed models (GLMMs), which involved gamma distribution (ln-link). The models included three samples, “pRep1-empty”, “pRep1-scd1-PB1”, “pRep1-byr2-RBD”, and “hours after induction of mating” as fully crossed fixed factors. Mixed models were necessary for these models to account for the repeated measurements from each Western blot membrane data.

To analyse the data presented in S4 Fig, Prism (GraphPad) was used.

Mathematical modelling

We constructed models with a system of 32 ordinary differential equations (ODEs) describing the amount of 7 molecular components (Ste11, Ste4/6, active Ras1 ([aRas1]), active Cdc42 ([aCdc42]), phosphorylated active MAPKK^Byr1 ([pByr1]), total MAPK^Spk1 ([tSpk1]) and phosphorylated ppMAPK^Spk1 ([ppSpk1])), the nutrient status ([N]) and time (t) in 5 different genetic conditions (wildtype, ras1.G17V, MAPKK^byr1.DD, ras1.G17V MAPKK^byr1.DD and scd1Δ) (S16 Fig). All the signalling components were set to be in close proximity, based on our own observation (Figs 1-4) and a series of localisation studies [35,36,38].

We postulated up to 29 biochemical processes involving the signalling components (S1 Table and Figs 8B, S17A and S19A) to reflect reported relationships among the signalling components. Each of these processes is accompanied by a rate parameter k. We first obtained an initial guess of the parameter values using an optimization toolbox of MATLAB software (Mathworks, Natick, MA, USA) and fitted the parameter values to better reproduce the experimental results using a Markov chain Monte Carlo method (MCMC) [67].

Modelling of regulatory networks for Spk1:

The initial model (Model A).

Model A (S17A Fig) consists of the following differential equations for the control (wild-type):

(1)

Equation (1) describes the temporal change in nutrient availability. The variable [N] represents the nutrient (nitrogen) status of the medium and is modeled as a smooth step-like function of time. Initially, [N] remains close to zero, corresponding to nutrient-rich conditions. After a characteristic time point, [N] rapidly increases toward one, representing nitrogen depletion. The arctangent function is used to approximate a stepwise transition while avoiding a discontinuous jump. The parameter 100 + k₂₄ determines the timing of nitrogen depletion, and k₂₅ controls the steepness of the transition. This formulation allows us to model nitrogen starvation as a gradual but rapid change in nutrient status, consistent with experimental conditions.

(2)

Equation (2) describes the dynamics of Ste11, a master transcriptional regulator of the pheromone signalling pathway. The production of Ste11 is induced by nitrogen starvation, represented by the nutrient-dependent term k₁[N], consistent with previous reports showing activation of Ste11 under nitrogen-depleted conditions [99,100]. In addition, Ste11 is upregulated by phosphorylated MAPK^Spk1 (ppSpk1) [42] through a positive feedback loop, represented by the term k₂[ppSpk1]/k₁₂. Here, k₁₂ is a normalization constant used to relate the model variable [ppSpk1] to the experimentally measured Western blot signal; a detailed explanation is provided below. Ste11 is degraded or inactivated at a constant rate proportional to its abundance (k₃[Ste11]). Through this formulation, Ste11 integrates nutrient status and MAPK signalling activity to control the expression of downstream pheromone signalling components.

(3)

Equation (3) describes the dynamics of the pheromone sensing unit (PSU). The PSU comprises genes encoding pheromones, receptors, Gpa1, Ste4, and Ste6 that are transcriptionally regulated by Ste11 and collectively represent the expression and availability of core pheromone signalling components [41,100–103]. The production of the PSU is driven by Ste11, the master transcriptional regulator, and is modeled by the term k₄[Ste11]. The PSU is removed or inactivated at a constant rate proportional to its abundance (k₅[Ste4/6]), representing turnover and dilution of the signalling components. This formulation reflects the transcriptional control of pheromone signalling components by Ste11 and their constitutive turnover.

(4)

Equation (4) describes the activation dynamics of Ras1. Active Ras1 ([aRas1]) is produced in response to the pheromone sensing unit (Ste4/6), reflecting Ste6-dependent activation of Ras1 following pheromone signalling [64]. This activation is modeled by the term k₇[Ste4/6](1-[aRas1]), where the factor (1-[aRas1]) ensures saturation as Ras1 approaches its maximal activation level. Active Ras1 is inactivated at a constant rate proportional to its abundance (k₈[aRas1]), representing GTP hydrolysis and other inactivation processes. This formulation captures stimulus-dependent Ras1 activation with a finite maximal level and constitutive inactivation. In this model, active Ras1 serves as a shared upstream activator of both Cdc42 and Byr1 Nodes, without explicitly modelling competitive binding or exclusivity between the two effectors.

(5)

Equation (5) describes the dynamics of active Cdc42 ([aCdc42]). Activation of Cdc42 is driven by active Ras1 through the guanine nucleotide exchange factor Scd1 [32], and is modeled by the term k₉[aRas1] (1 - [aCdc42]/k₁₉). Here, k₁₉ is a normalization constant that relates the model variable [aCdc42] to the experimentally measured CRIB-GFP signal; a detailed explanation is provided below. The saturation factor ensures that Cdc42 activity does not exceed its maximal level. Active Cdc42 is negatively regulated by constitutive inactivation processes, represented by the term k₁₀[aCdc42]/k₁₉. In addition, Cdc42 is subject to enhanced inactivation during vegetative growth in the presence of nitrogen [104], reflecting higher activity of GTPase-activating proteins under nutrient-rich conditions; this regulation is captured by the nitrogen-dependent term k₁₁(1 - [N]) [aCdc42]/k₁₉. Together, this formulation models Ras1-dependent activation of Cdc42 balanced by nutrient-dependent and nutrient-independent inactivation processes.

(6)

Equation (6) describes the dynamics of the Byr1 Node, which represents the combined activity of MAPKKK^Byr2 and MAPKK^Byr1 in the MAPK signalling cascade. Activation of the Byr1 Node is driven by the pheromone sensing unit (Ste4/6), reflecting Ste4-dependent activation of MAPKKK^Byr2 [33,39,40,63], and is modeled by the term k₁₃[Ste4/6](1-[ppByr1]). In addition, the activity of the Byr1 Node is positively modulated by active Ras1 and active Cdc42 [33], represented by the terms k₁₄[aRas1] and k₁₅[aCdc42]/k₁₉, respectively. These terms summarize experimentally reported regulatory influences of Ras1 and Cdc42 on MAPK pathway activation and represent effective modulation rather than direct elementary biochemical reactions. The Byr1 Node is subject to constitutive inactivation or turnover, represented by the term k₁₆[ppByr1]. Furthermore, MAPK signalling is negatively regulated under nutrient-rich conditions [105]; this effect is incorporated as nitrogen-dependent inhibition of the Byr1 Node, represented by the term k(1-[N])[ppByr1]. Together, this equation models integration of pheromone-dependent activation, small GTPase–mediated modulation, and nutrient-dependent repression to control MAPK pathway activity.

(7)

Equation (7) describes the dynamics of total MAPK^Spk1 ([tSpk1]). The production of Spk1 is induced by Ste11, reflecting transcriptional upregulation of MAPK components during the mating response, and is modeled by the term k₁₇[Ste11]. Total Spk1 is removed at a constant rate proportional to its abundance (k₁₈[tSpk1]), representing protein turnover and dilution. This formulation captures Ste11-dependent expression and constitutive turnover of MAPK^Spk1.

(8)

Equation (8) describes the dynamics of phosphorylated MAPK^Spk1 ([ppSpk1]). Phosphorylation of Spk1 is driven by the activity of the Byr1 Node, which represents the combined MAPKKK/MAPKK activity in the MAPK cascade, and depends on the availability of unphosphorylated Spk1. This process is modeled by the term k₂₀[ppByr1]([tSpk1]-[ppSpk1]/k₁₂). Phosphorylated Spk1 is removed through dephosphorylation and other inactivation processes, represented by the term k₂₁[ppSpk1]/k₁₂. Together, this equation models MAPK activation as a Byr1-dependent phosphorylation reaction balanced by constitutive inactivation.

Note that [tSpk1] represents the amount of total Spk1, as measured by the signal intensity of the total Spk1 Western blot. Accordingly, the amounts of other model variables representing protein abundance (e.g., [Ste11]) are expressed as relative values normalized to 1 unit of [tSpk1], except for [ppSpk1] and [aCdc42]. In the case of phosphorylated Spk1, [ppSpk1] represents the signal intensity of phospho-specific Spk1 Western blot (and not that of total Spk1). The amount of phosphorylated Spk1 normalized to 1 unit of total Spk1 is therefore given by [ppSpk1]/k₁₂. In case of active Cdc42, [aCdc42] represents the experimentally measured CRIB-GFP signal (Fig 4B). The amount of active Cdc42 normalized to 1 unit of total Spk1 is given by [aCdc42]/k₁₉.

For ras1.G17V mutant, we used the same differential equations except that [aRas1] was fixed to a constant value of k₂₂. Similarly, for byr1.DD mutant, [ppByr1] was fixed to a constant value of k₂₃. For scd1-delta mutant, [aCdc42] was fixed to zero.

Model B.

In Model B (Fig 8B), a negative regulation from ppSpk1 to Ste4/6 was added to Model A. Eq. (3’) was used instead of Eq. (3) in Model A.

(3’)

In equation (3’), the additional term k₆ [ppSpk1]/k₁₂ reduces the effective level of Ste4/6 in proportion to MAPK activity. This term represents delayed negative feedback from MAPK signalling to upstream pheromone sensing components, summarizing MAPK-dependent attenuation mechanisms acting on the PSU.

Model C.

Model C (S19A Fig) extends Model B by introducing additional regulatory interactions that are specific to the ras1.G17V background. In this model, two negative regulations from ppSpk1 to downstream reactions of Ras1 are incorporated to account for experimentally observed attenuation of signalling in the presence of constitutively active Ras1. These regulations are implemented by replacing Eqs. (5) and (6) with Eqs. (5’) and (6’), respectively, only under ras1.G17V and ras1.G17V byr1.DD conditions. In these conditions, [aRas1] and [ppByr1] are fixed to constant values (k₂₂ and k₂₃, respectively), reflecting constitutive Ras1 activation and Byr1 activation in the corresponding mutants.

(5’)

Equation (5’) modifies Eq. (5) by introducing a MAPK^Spk1-dependent negative regulation on Cdc42 activation that is specific to the ras1.G17V background. The additional term k₂₈ [ppSpk1] represents attenuation of Ras1–Scd1–Cdc42 signalling by MAPK activity, thereby reducing Cdc42 activation in proportion to MAPK^Spk1 phosphorylation. This modification captures MAPK-dependent feedback acting on the Ras1–Cdc42 branch under constitutive Ras1 activation, while the remaining terms are identical to those in Eq. (5).

(6’)

Equation (6’) modifies Eq. (6) by adding a MAPK^Spk1-dependent negative regulation on the Ras1-mediated activation of the Byr1 Node. The additional term k₂₉[ppSpk1] represents attenuation of Ras1–Byr2 signalling by MAPK activity, which is applied specifically under ras1.G17V and ras1.G17V byr1.DD conditions. This modification implements MAPK-dependent suppression of Ras1-driven MAPK pathway activation under constitutive Ras1 signalling, while all other terms remain unchanged from Eq. (6).

In addition, a regulation to activate aByr1 to ppSpk1 by aCdc42 was added to expect the distinct outcome of byr1.DD and ras1.G17V byr1.DD as observed in the experiment. Eq. (8’) was used instead of Eq. (8).

(8’)

Equation (8’) modifies Eq. (8) by adding a Cdc42-dependent enhancement of MAPK^Spk1 phosphorylation, represented by the term k₂₇[aCdc42]/k₁₉. Apart from this additional term, the formulation of MAPKSpk1 phosphorylation and dephosphorylation remains unchanged from Eq. (8).