3-MB-PP1 inhibits analog-sensitive Ssn3

To investigate the direct targets of the C. albicans Ssn3 kinase, we sought to develop a strain in which Ssn3 kinase activity could be rapidly inhibited using approaches that have been successfully applied in S. cerevisiae [17]. Based on the alignment of the C. albicans and S. cerevisiae Ssn3 orthologs, we predicted a phenylalanine to glycine substitution at position 257 would yield a variant that retained the functions of the wild-type kinase, while being able to be specifically inhibited by the ATP analog 3-MB-PP1 [20].The Ssn3^F257G was constructed and is henceforth referred to as the analog-sensitive variant, Ssn3^AS (see S1 Table for summary of strains).

We first assessed the inhibition of Ssn3^AS by 3-MB-PP1 with an in vitro kinase assay using purified Mediator complex containing the Cdk8 module. Ssn3^WT or Ssn3^AS were expressed in a background that contained a His-FLAG-tagged derivative of Ssn8 for purification. The activity of Ssn3^WT or Ssn3^AS was assessed using ³²P in vitro kinase assays [21] in which phosphorylation of recombinantly produced C-terminal domain (CTD) of RNA Pol II, an Ssn3 substrate, was monitored. The Ssn3^WT and Ssn3^AS kinases had equal CTD phosphorylation activity in the absence of inhibitor (Fig 1). The Ssn3^AS kinase activity was inhibited by 3-MB-PP1 in a dose-responsive fashion, while addition of this compound had no effect on the Ssn3^WT kinase activity (Fig 1). A concentration of 2.7 μM 3-MB-PP1 inhibited ~85% of the Ssn3^AS activity and 24 μM virtually eliminated the activity.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. 3-MB-PP1 inhibits the activity of analog-sensitive Ssn3^AS, but not Ssn3^WTin vitro.

In vitro kinase reactions contained purified Mediator from a strain with Ssn3^WT or Ssn3^AS, ³²P-ATP, purified GST-tagged RNA Pol II C-terminal domain (CTD) and the indicated concentrations of 3-MB-PP1 inhibitor. Reactions were analyzed by SDS-PAGE and visualized by phosphorimaging.

https://doi.org/10.1371/journal.pgen.1009622.g001

To analyze Ssn3 kinase activity in live fungal cells, we engineered a derivative of C. albicans SC5314 in which both alleles of SSN3 had been replaced by ssn3^AS. For comparison, we also constructed a strain in which each SSN3 allele was replaced with ssn3-D325A, which encodes a non-functional, or kinase dead, Ssn3 variant [22] due to mutation of a key aspartate in the catalytic site. This kinase-dead variant is referred to here as ssn3^KD. We have previously shown that, under certain growth conditions, ssn3Δ/Δ mutants form hyperwrinkled colonies compared to the SC5314 wild type (WT) [13], a phenotype associated with increased hyphal growth. Thus, we asked if the Ssn3^AS-expressing strain had a hyperfilamentous phenotype specifically in the presence of 3-MB-PP1, but not in its absence. To quantify hypha formation differences in WT, ssn3^AS, and ssn3^KD strains, we grew cells in YNB containing 11 mM glucose and amino acids (YNBAG) and N-acetyl-glucosamine (GlcNAc). Both amino acids and GlcNAc are inducers of hyphal growth, but cultures were incubated at 30°C, a temperature lower than that generally used to induce robust hyphal growth. All cultures received either 3-MB-PP1 or the vehicle DMSO. The morphology of the WT was almost entirely yeast and pseudohyphae, and the relative fractions of these morphologies were unaffected by 5 μM 3-MB-PP1 (Fig 2A and 2B).While the morphological profiles for ssn3^AS cells were similar to the WT in DMSO control cultures, the addition of 3-MB-PP1 tossn3^AS cultures caused a significant increase in the number of hyphae (Fig 2A and 2B).The increase in the fraction of cells in the hyphal morphology in 3-MB-PP1 treated ssn3^AS was concomitant with a significant decrease in the fraction of cells present as yeast. The ssn3^KD strain formed significantly more hyphae than the WT and ssn3^AS strains in control conditions, and the fraction of cells as hyphae was unaffected by the addition of 3-MB-PP1 (Fig 2A and 2B). Together, these data indicated that 5 μM 3-MB-PP1 inhibits Ssn3^AS in vivo and that the ATP analog has no discernible effects on the morphology of WT or ssn3^KD strains.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. 3-MB-PP1 stimulates hyphal growth in a strain bearing analog-sensitive alleles of SSN3.

A. Morphology of SC5314 (WT), ssn3^AS and ssn3^KD strains was assessed after growth in YNBAG supplemented with either 5 μM 3-MB-PP1 or vehicle (DMSO) for 3h at 30°C. Quantification of yeast, pseudohyphae and hyphae in cultures was performed by microscopic analysis of blinded samples. Results from ANOVA with multiple comparisons for the percent of cells as hyphae is shown; ****, p<0.001, ns, not significant. Lower case letters indicate statistically significant differences in comparison to a) WT with 3MB-PP1 at p<0.001 or b) to WT with DMSO at p<0.001. B. Representative images of cell populations from cultures analyzed in panel A are shown.

https://doi.org/10.1371/journal.pgen.1009622.g002

Transcriptomic analyses reveal that inhibition of Ssn3 by 3-MB-PP1 leads to the induction of hypha-specific genes

To further explore the effects of Ssn3 inhibition on the regulation of hypha formation [13] and on transcription more broadly, we examined the transcriptomes of ssn3^ASand WT strains grown with either 3-MB-PP1 or DMSO vehicle. The WT strain was included in this experiment to assess off-target effects of 3-MB-PP1. Cells from stationary phase cultures were inoculated into the same medium used for morphology assessment, YNBAG, with either 5 μM 3-MB-PP1 or an equivalent volume of DMSO. Three replicate cultures for each strain-treatment combination were incubated for 60 min at 30°C prior to RNA harvest and sequencing as described in the Materials and Methods section. In the ssn3^AS strain, we found that 249 genes were significantly up regulated upon treatment with 3-MB-PP1 by 2- to 49-fold (p<0.05 FDR); transcripts for 33 genes were significantly lower by more than 2-fold with treatment (S2 Table). Fewer genes were affected by 3-MB-PP1 in the WT with 157 and 4 genes significantly increased and decreased, respectively; the magnitudes of the changes were also smaller (2- to 6-fold) (S2 Table). The greater number of transcripts at higher levels upon inhibition of Ssn3 is consistent with the Cdk8 module being a predominantly negative regulator of transcription factors [3], transcriptional re-initiation, or “pausing” of RNA polymerase II (reviewed in [23]). We found seventy-three transcripts that exhibited a significant fold increase (>2-fold) in both the comparison of the ssn3^ASstrain treated with 3-MB-PP1 to its treatment with DMSO, and the comparison of 3-MB-PP1-treated ssn3^AS strain to 3-MB-PP1-treated WT (Fig 3) with gene expression shown in log₂-transformed counts per million). Consistent with phenotypic analysis of the effects of Ssn3 inhibition on morphology (Fig 2), transcripts encoding ECE1 and HWP1 were the two genes most highly induced by Ssn3 inhibition in the AS strain (46- and 22-fold higher, respectively) and other transcripts associated with growth in the hyphal morphology, such as ALS1, ALS3, IHD1, RBT1, HYR1, and HGC1, were also differentially-increased upon Ssn3 inhibition [15,24]. Consistent with the previous finding that Ssn3 represses induction of Mrr1 controlled genes, MDR1 was significantly induced with Ssn3 inhibition [12]. Other transcriptional responses to the inhibition of Ssn3 activity are discussed below.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Heat map for genes induced upon inhibition of Ssn3^AS by 3-MB-PP1.

Transcripts for seventy-three genes were 2-fold higher (p<0.05, FDR corrected) in both the comparison of Ssn3^AS with 3-MB-PP1 compared to WT with 3-MB-PP1 and of Ssn3^AS with 3-MB-PP1 compared to with the DMSO control. Genes are listed in order with the greatest magnitude difference between Ssn3^AS with and without 3-MB-PP1 at the top. The heat map shows three replicates per sample of Log2-transformed counts per million for expressed transcripts.

https://doi.org/10.1371/journal.pgen.1009622.g003

Specific inhibition of Ssn3 affects the phosphoproteome during the induction of hyphal growth

To identify direct phosphorylation targets of Ssn3 that led to changes in phenotype and the transcriptome, we used mass spectrometry to quantitatively analyze the phosphoproteomes of the ssn3^AS strain compared to the WT after 3-MB-PP1 treatment. We grew ssn3^AS and WT cells to stationary phase, then incubated culture aliquots with 5 μM 3-MB-PP1 for five minutes to allow for drug entry into the cell. We then added concentrated, fresh, pre-warmed YNBAG medium to reproduce the conditions that induce hyphal growth in strains with low Ssn3 activity but not in the WT. Cultures were incubated for an additional 15 min at 30°C with shaking followed by rapid harvest to minimize secondary effects resulting from Ssn3 inhibition. Cells were lysed under liquid nitrogen by grinding.

To reveal the direct targets of Ssn3, we focused on phosphosites that became less abundant in the presence of 3-MB-PP1 in the ssn3^AS strain, but not the WT (Tables 1 and S3 for complete dataset). We found that 977 phosphopeptides from 552 proteins were significantly depleted in ssn3^AS, but not WT, by more than 2-fold (P<0.05) (Table 1). Within this group of depleted phosphosites, 82.7% were serines, 15.8% were threonines, and 1.5% were tyrosines. These proportions are very similar to those we observed with a quantification of the whole C. albicans phosphoproteome [21], with a slight increase in the number of phosphoserines and a concomitant decrease in the number of phosphothreonines. Although promiscuous, Cdk kinases have been described as proline-directed kinases, and consistent with this, 293 phosphopeptides had phosphosites within a proline motif containing a proline in the position adjacent to C-terminal side of the phosphoresidue (Table 1) [25]. A much smaller number of total phosphopeptides and phosphosites in proline motifs increased upon Ssn3 inhibition (Table 1), which likely reflect secondary effects of Ssn3 inhibition or off-target effects on 3-MB-PP1 in the ssn3Δ/Δ background.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Summary of phosphopeptides and phosphoproteins detected by motif.

Analysis limited to those peptides to which P-values of <0.05 were assigned in the comparison of ssn3^AS to the wild type (WT). Down indicates phosphopeptides lower in ssn3^AS treated with 3-MB-PP1 compared to WT treated with 3-MB-PP1; up indicates phosphopeptides that were more abundant in the ssn3^ASstrain treated with 3-MB-PP1 compared to WT treated with 3-MB-PP1. The total refers to the number of peptides or proteins that are significantly different regardless of fold-difference. “#” indicates the detected phosphorylation on a serine or threonine, S/T, in a -1 position relative to a proline, S/T#P motif.

https://doi.org/10.1371/journal.pgen.1009622.t001

Of the 552 proteins with phosphosites that were reduced by Ssn3 inhibition, 218 proteins contained two or more depleted phosphopeptides) in the ssn3^AS strain compared to WT (a consideration taken to increase stringency). Forty one of these 218 proteins were annotated as having known or predicted nuclear localization in UniProt (Fig 4) [26]. Among these proteins with phosphopeptides depleted upon Ssn3^AS inhibition was Med4, with phosphorylated S21 and S33; these Med4 sites have previously been identified as target sites for the C. albicans Ssn3 kinase [21]. We also observed depletion of specific phosphosites (T589 and S620) in Flo8, a regulator of filamentous growth in both C. albicans and S. cerevisiae [18,19,27], and both T589 and S620 were found within proline motifs (Table 1). The differences in Flo8 phosphopeptides were not accompanied by differences in FLO8 mRNA levels upon3-MB-PP1 addition to the ssn3^AS strain (S2 Table).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Overview of proteins for which two phosphopeptides were lower by >2 fold upon inhibition of Ssn3.

Focusing on peptides that were significantly lower in Ssn3^AS treated with 3-MB-PP1 relative to control cultures (p<0.05 FDR-corrected), we found 977 peptides that were 2-fold lower upon 3-MB-PP1 inhibition of ssn3^AS relative to changes in the wild type. Two hundred and eighteen proteins had two more peptides that met these criteria, forty one of which were predicted to have nuclear localization. Med4 is a validated C. albicans Ssn3 target. If an alias is available for an unnamed gene, it is shown in parentheses.

https://doi.org/10.1371/journal.pgen.1009622.g004

Flo8 is required for Ssn3-dependent hyperfilamentation and hypha-specific gene expression

To determine whether there was a genetic interaction between SSN3 and FLO8 that accompanied the phosphoproteomic data above, we investigated the phenotype of single and double null mutants of these two genes. We previously reported that an ssn3 null mutant forms wrinkled colonies in the presence of a metabolic inhibitor, pyocyanin, while the WT does not [13]. While both the ssn3 null strain and the WT formed smooth colonies on YNBA agar at 30°C (which is the same as YNBAG used above but without GlcNAc), only the ssn3Δ/Δ strain formed wrinkled colonies at 37°C. Similarly, the ssn3Δ/Δ strain, but not the WT, formed wrinkled colonies on solid YNBA medium with 110 mM added glucose (Fig 5A). This wrinkled colony phenotype of the ssn3Δ/Δ mutant under the above conditions was abolished upon deletion of FLO8 (ssn3Δ/Δflo8Δ/Δ) (Fig 5A). The flo8Δ/Δ mutant formed smooth colonies, like the WT, under all conditions. In liquid medium, we observed a similar epistatic relationship between FLO8 and SSN3. Only the ssn3Δ/Δ strain, and not the WT, formed hyphae and the hyperfilamentation phenotype in ssn3Δ/Δ was dependent on FLO8 (Fig 5B). Expression levels of hypha-specific transcripts that were induced upon inhibition of Ssn3^AS (Fig 3) were significantly higher in the ssn3Δ/Δ background compared to the WT, but not in the ssn3Δ/Δflo8Δ/Δ background (Fig 5C). In Fig 6, we demonstrate the ability to complement the filamentation defects of the flo8Δ/Δ and ssn3Δ/Δflo8Δ/Δ with the native FLO8 allele, and this result is described in more detail below.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Ssn3 repression of filamentation is FLO8 dependent.

A. Colony morphology of SC5314 wild type (WT), ssn3Δ/Δ, flo8Δ/Δ, and ssn3Δ/Δflo8Δ/Δ strains grown on YNBA agar medium alone or supplemented with 110 mM glucose at 30°C or on YNBA at 37°C. B. Cell morphology of strains tested in (A) after growth in YNBN_2.5AG (11 mM glucose) at 30°C for 3 hours. C. NanoString analysis of indicated hypha-associated genes in WT, ssn3Δ/Δ, and ssn3Δ/Δflo8Δ/Δ. RNA was extracted from cells grown as indicated for (B) but for 75 minutes. Expression levels was represented by mean and standard deviation after normalization to TEF1 and ACT1 reference transcripts. Expression of each gene was normalized to levels for the WT. Significant differences from WT are indicated as **, p<0.01 and ***,p<0.001. D. Embedded colonies of the same set of strains shown in (A) grown in YPS agar at 25°C.

https://doi.org/10.1371/journal.pgen.1009622.g005

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Alteration of Flo8 phosphosite residues influences Flo8 function.

A.Wild type (WT), ssn3Δ/Δ, flo8Δ/Δ and ssn3Δ/Δflo8Δ/Δ strains were imaged after growth as colonies on YNBA or YNBA+110 mM glucose at 37°C. In addition, ssn3Δ/Δflo8Δ/Δ or flo8Δ/Δ were complemented with constructs that encoded Flo8^WT, Flo8^T589A/S620A or Flo8^T589E/S620E with C-terminal 3XHA tags. B. Cell morphology of WT, flo8Δ/Δ, and flo8Δ/Δ expressing 3XHA-tagged Flo8^WT, Flo8^T589A,S620A or Flo8^T589E,S620E after growth in YNBN_2.5AG(11 mM glucose) at 30°C for 3h. C. Expression of hypha-associated transcripts in flo8Δ/Δ expressing either 3XHA-tagged Flo8^WT or Flo8^T589A/S620A after growth in YNBNAG for ~2h^. Data are normalized levels of the corresponding transcript in flo8Δ/Δ. Data show the mean and standard deviation from measurement on triplicate RNA samples. P-values, determined by t-test, indicated by *, p<0.05; **, p<0.01; ***, p<0.001; or the specific value was indicated. D. Immunoblotting analysis on 10% SDS-PAGE with an α-HA antibody shows levels and mobility of 3X-HA derivatives of Flo8, Flo8^T589A/S620A or Flo8^T589E/S620E in either WT or ssn3Δ/Δ. CBS indicates levels of Coomassie Blue stained total protein in a different region of the blot and is used as a loading control. E. Immunoblot of Flo8-HA levels and mobility on 10% SDS-PAGE in strains with full Ssn3 activity, WT, ssn3^AS, ssn3Δ/Δ (Δ/Δ) or ssn3^KD (which expresses the kinase dead variant; kd/kd). CBS indicates levels of Coomassie Blue stained total protein in the region of the target protein, and is used as a loading control.

https://doi.org/10.1371/journal.pgen.1009622.g006

In addition to the wrinkled colony and liquid media hyperfilamentation phenotypes, the ssn3Δ/Δ strain was also hyperfilamentous, compared to WT, in embedded conditions. As observed with the other hyperfilamentation phenotypes, deletion of FLO8in the ssn3Δ/Δ background was able to suppress this phenomenon (Fig 5D). Consistent with the results of Cao et al. [18], we observed that FLO8 was necessary for embedded filamentation (Fig 5D). To determine whether FLO8 is also required for embedded hyphal growth in efg1Δ/Δ, another strain that is hyperfilamentous under embedded conditions [28,29], we generated a flo8 and efg1 null double mutant. We found that FLO8 disruption suppressed embedded filamentation in the efg1Δ/Δ background (S1A Fig). While Efg1 is a negative regulator of embedded hyphal growth, it is well-known for its positive regulation of hypha formation under other conditions. The efg1Δ/Δ mutant and ssn3Δ/Δefg1Δ/Δ mutant strains were not filamentous in colonies grown on agar medium under filament-inducing conditions at 37°C, and the hyperfilamentous phenotype of the ssn3Δ/Δ mutant upon incubation under mild hypha-inducing conditions was not observed in the ssn3Δ/Δefg1Δ/Δ mutant (S1B Fig). These data indicate that both Flo8 and Efg1 contribute to the hyperfilamentation of the ssn3Δ/Δ mutant. While we focus on Flo8 in further studies below, it is worth noting that we identified phosphosites in Efg1 (serine 531) and Ndt80 (threonines 495, 497, and 499), another hyphal growth associated transcription factor, among those that were at lower abundance upon Ssn3 inhibition (S3 Table).

Loss of putative Ssn3-phosphosites in Flo8 leads to increased filamentation

To study the effects of Ssn3 activity on Flo8 protein, we complemented the ssn3Δ/Δflo8Δ/Δ mutant with an allele that encodes a 3XHA-Flo8. This allele restored the hyperfilamentous ssn3Δ/Δ phenotype to the ssn3Δ/Δflo8Δ strain (Fig 6A). To further explore the effects of Flo8 phosphorylation by Ssn3 we sought to determine whether Flo8 function was altered upon mutation of candidate Ssn3 target sites identified in the phosphoproteomics analysis of the Ssn3^AS bearing strain. Thus, we generated an allele in which both T589 and S620 were each mutated to alanine (a phospho-incompetent residue) or glutamic acid (which also abolishes phosphorylation, but sometimes can serve as a phosphomimetic). Neither the WT or flo8Δ/Δ strain gave a wrinkled colony morphology on either medium (Fig 6A) and, as expected, complementation of the flo8Δ/Δ with epitope tagged Flo8 did not change colony morphology. However, complementing the flo8Δ/Δ strain with either flo8^{T589A, S620A} or flo8^{T589E, S620E} led to a wrinkled colony phenotype in the presence 110 mM glucose, suggesting that the phosphorylation of Flo8, potentially via Ssn3, represses Flo8 activity and that the absence of these sites releases that repression (Fig 6A). Consistent with the increased wrinkled colony phenotype in the flo8Δ/Δstrains with flo8^T589A,S630A or flo8^{T589E, S620E} relative to the strain with unmutated FLO8, both Flo8 variants also led to increased hyphal development in YNBAG (Fig 6B). Because both flo8^T589A,S620A and flo8^T589E,S620E alleles conferred increased filamentation phenotypes consistent with high Flo8 activity in flo8Δ/Δ and ssn3Δ/Δflo8Δ/Δ strains we concluded that the Flo8^T589E,S620Evariant was not acting as a phosphomimetic, but rather both had phenotypes consistent with decreased negative regulation. To quantitatively assess the differences in activity of the morphology program attributed to the loss of putative Ssn3 phosphorylation sites, levels of hypha-associated transcripts were measured in the flo8Δ/Δ strains with either HA-tagged FLO8 or flo8^{T589A, S620A} (Fig 6C). We found significantly increased expression levels of several core filamentation response genes in the strain with flo8^{T589A, S620A} versus that with FLO8.

In multiple instances, phosphorylation by Ssn3 leads to decreased levels of transcription factors [6,8,16]. Thus, we tested the hypothesis that Flo8-HA^T589A,S620A and Flo8-HA^T589E,S620E were present at higher levels than Flo8-HA. Immunoblotting with an α-HA antibody indeed showed that the two variants were present at higher relative levels compared to Flo8-HA in an SSN3 wild-type background (Fig 6D). Furthermore, native Flo8-HA levels were higher in the ssn3Δ/Δ strain than in the SSN3/SSN3 (+/+) strain, while there were no significant differences in levels for the variant Flo8 proteins. We observed a modest increase in mobility (suggestive of decreased phosphorylation [30,31]) when Ssn3 activity was lacking (Fig 6D and 6E); however, because even Flo8-HA^T589A,S620A showed a difference in migration between WT and ssn3Δ/Δ (Fig 6D), we speculate that there are other post-translational modifications (e.g. additional phosphorylations) that influence protein mobility and that are controlled by Ssn3. It is worth noting that one other Flo8 phosphosite was identified in the phosphoproteomics analysis, but it did not reach the significance cutoff.

Ssn3 metabolic hyperalkalinization phenotype depends on Flo8

Previous work has shown that the C. albicans ssn3Δ/Δ mutant had increases in several metabolic pathways including glycolysis and the utilization of amino acids relative to the WT [13], and that these contribute to a strong hyperalkalinization phenotype for ssn3Δ/Δ. Alkalinization of cultures is due to greater accumulation of products that raise the pH (e.g. ammonia) than products that decrease culture pH, such as the fermentation product acetic acid [32]. We constructed stp2Δ/Δ and ssn3Δ/Δstp2Δ/Δ strains which lack the transcription factor Stp2, a downstream component of the SPS system that enhances ammonia release by up regulating amino acid permeases and amino acid catabolic enzymes [33]. The stp2Δ/Δ mutant had a defect in alkalinization at 5.5 mM glucose compared to WT, and showed medium acidification, rather than alkalinization, at both 27.5 mM and 110 mM glucose, while the WT only acidified the medium when glucose was at the highest concentration tested (Fig 7A). The ssn3Δ/Δ mutant showed the previously reported hyperalkalinization phenotype, and this phenotype was suppressed in the ssn3Δ/Δstp2Δ/Δ strain. We found that the ssn3Δ/Δstp2Δ/Δ strain still formed wrinkled colonies, like the ssn3Δ/Δ strain, suggesting that Stp2 is not upstream of Flo8 and that alkalinization is not required for hyperfilamentation (S2 Fig). Because Flo8 was required for hyperfilamentation in the ssn3Δ/Δ strain, we assessed the phenotype of theflo8Δ/Δ and ssn3Δ/Δflo8Δ/Δ strains in the alkalinization assay (Fig 7B). While the flo8Δ/Δstrain did not differ from the WT in its effects on medium pH, the ssn3Δ/Δflo8Δ/Δ cultures had lower medium pH values than the ssn3Δ/Δ strain, and the hyperalkalinization phenotype was restored by complementation of FLO8 (Fig 7A). Unlike for filamentation, complementation of the flo8Δ/Δ mutant with flo8^T589A,S620A did not increase medium alkalinization relative to complementation with the native allele suggesting that Ssn3-Flo8 effects on metabolism depends on additional phosphorylation sites of Flo8 (Fig 6D) or that Flo8 effects on metabolism may require other factors directly or indirectly regulated by Ssn3.The flo8^T589A,S620A allele also was indistinguishable from the native FLO8 allele in the ssn3Δ/Δflo8Δ/Δ background. In light of our data and work by other groups that suggest complex interactions between Flo8 and Efg1 [(S1 Fig) and [18,34,35]] and Efg1 control of metabolism [28], we sought to determine if, like Flo8, Efg1 contributed to the hyperalkalinzation phenotype of the ssn3Δ/Δ mutant (Fig 7C). We found that efg1Δ/Δ and efg1Δ/Δflo8Δ/Δ strains were similar to the WT in terms of medium alkalinization, but that the ssn3Δ/Δefg1Δ/Δ and ssn3Δ/Δefg1Δ/Δflo8Δ/Δ strains did not show the alkalinization phenotype of the ssn3Δ/Δ despite robust growth. Furthermore, the phenotype of the ssn3Δ/Δefg1Δ/Δflo8Δ/Δ triple mutant was comparable to the ssn3Δ/Δefg1Δ/Δ and ssn3Δ/Δflo8Δ/Δdouble mutants suggesting that the effects of the efg1Δ/Δ and flo8Δ/Δ mutations were not additive. Together, these data lead us to propose that the increased alkalinization upon loss of Ssn3 activity requires both Flo8, Efg1 and Stp2 and future work will determine the specific relationships between Ssn3 and these regulators (see Fig 8 for model). Analysis of transcripts that were differentially expressed upon inhibition of Ssn3 found a statistically significant enrichment of genes in KEGG pathways involved in amino acid metabolism, phospholipid metabolism and glycolysis (S4 Table) and both glycolysis and amino acid catabolism appear to be upregulated by decreased Ssn3 activity [13]. Thus, we speculate that the effects of Ssn3 on metabolism are likely multifactorial.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. Medium alkalinization by Ssn3 is dependent on Stp2, Flo8, and Efg1.

Medium pH was assessed on YNBA with increasing concentrations of glucose; bromocresol purple was added as a pH indicator. All cultures were incubated at 37°C. A. Comparison of medium pH upon growth of SC5314 (WT), ssn3Δ/Δ, stp2Δ/Δand ssn3Δ/Δstp2Δ/Δstrains. B. WT, ssn3Δ/Δ, flo8Δ/Δ and ssn3Δ/Δflo8Δ/Δ alone or complemented with C-terminally 3XHA-tagged Flo8^WT or Flo8^T589A/S620A were grown as described above. C. WT, ssn3Δ/Δ, flo8Δ/Δ and ssn3Δ/Δflo8Δ/Δ were compared to efg1Δ/Δ, ssn3Δ/Δefg1Δ/Δ, flo8Δ/Δefg1Δ/Δ, and ssn3Δ/Δflo8Δ/Δefg1Δ/Δ strains.

https://doi.org/10.1371/journal.pgen.1009622.g007

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 8. Summary of SSN3-centered genetic interactions revealed by this study.

Hyperfilamentation of ssn3Δ/Δ depends on Flo8 under all conditions tested including in liquid, on agar, and in embedded conditions (orange arrow). Flo8 activity in filamentation assays is tuned at T589 and S620, which were identified as sites phosphorylated by Ssn3. Efg1 plays positive roles in the hyperfilamentation of the ssn3Δ/Δmutant (blue arrow) but plays a negative role in filamentation in embedded conditions (blue block). Deletion ofSSN3also leads to hyperalkalinization of amino acid-rich media, and alkalinization is not required for filamentation. Alkalinization is reduced or no longer observed in ssn3 mutants lacking FLO8, EFG1 or STP2. Future studies will determine the mechanisms by which Ssn3 impacts metabolism through Flo8 and Efg1, if it directly modulates Efg1 levels (grey, dotted block arrow) and if Ssn3, Flo8 or Efg1 affect Stp2-dependent metabolism (grey dotted arrow).

https://doi.org/10.1371/journal.pgen.1009622.g008

Media and growth conditions

Strains were maintained on YPD plates, and overnight cultures for morphology and RNA seq were grown in YNB/1% glucose at 30°C in culture tubes on a roller drum. YPD plates were routinely streaked from glycerol stocks, and experiments were only conducted with overnight cultures from plates not more than 5 days old. YNB medium with 2% (w/v) casamino acids and 11 mM or 110 mM glucose (YNBAG₁₁ or YNBAG₁₁₀) with or without 5 mM N-acetyl glucosamine (GlcNAc) was adjusted to pH 5.1 (morphology and RNA seq experiments) or pH 6.0 (phosphoproteomics) with concentrated hydrochloric acid, and filter sterilized. When indicated, the pH indicator bromocresol purple was included in the medium as described in [13].Other details regarding temperature and incubation time can be found in the figure legends and in the methods sections below.

Wrinkled colony and alkalinization assays

Overnight cultures (YPD) of each strain were washed once with water and diluted to OD₆₀₀ 4. 8 μL of cell suspension was spotted onto the indicated medium. Images were typically taken after 2 days growth at 37°C or 3 days growth at 30°C. For alkalinization assays, 2X YNB based media was adjusted to pH 5.0 by HCl, filter sterilized and mixed with autoclaved 4% agar solution. Bromocresol purple (BCP) was added to 0.01% from a 0.1% aqueous stock into the media before 2 mL was aliquoted into each well of a 24-well plate. pH references were generated using YNB media buffered by phosphate buffer (20 mM) with known pH. Glucose concentrations 0–110 mM) were as indicated in the text and/or figure legends.

Strain construction

All strains used in this study are listed in S1 Table. Primers and plasmid sequences are available upon request. Strains expressing Ssn3 that were analog-sensitive (AS) or kinase-dead (KD) variants were generated by transformation of SC5314 with a DNA fragment containing ssn3^F257G (ssn3^AS) or ssn3^D325A (ssn3^KD) adjacent to the SAT-FLP cassette directed to the SSN3 locus. These constructs were transformed alongside the Candida-optimized CRISPR/Cas9 machinery [51] and a guide sequence targeting the nuclease to the SSN3 open reading frame. Transformants were selected on YPD with 200 μg neourseothricin (GoldBio) and resistant colonies were regrown on the same medium, then maintained on YPD. Amplification using primers that spanned the SSN3 locus from DNA isolated from these heterozygotes revealed the SAT1 marker was excised during outgrowth on YPD. Heterozygous strains underwent a second transformation with the same construct using the methods described above to generate mutant homozygotes. Transformants after the second round of transformation were confirmed by PCR amplification of a ~970 base pair internal region of SSN3 covering the region encoding residues 257 and 325 using primers Ssn3 Internal FWD and Ssn3 Internal Rev, and Sanger sequencing with the reverse primer. As in the first round, a number of these transformants had excised the resistance marker as observed by PCR. Generation of other knockout mutants was carried out using a previously described transient CRISPR-Cas9 system using a SAT-flipper selection marker [52]. To generate double mutants, the SAT1 cassette was recycled by inducing FLP recombinase expression in YP maltose (1% yeast extract, 2% peptone, 2% maltose) for 24 hours. Generation of Flo8-HA-tagged strains was similarly accomplished through a repair construct in which the desired allele was fused to the SAT-FLP marker. For stp2Δ/Δ, in light of a known aneuploidy, there were three copies of STP2 in the genome and all three were deleted [53,54].

Mediator purification and in vitro kinase assays

Ssn8-tagged Mediator containing various SSN3alleles was purified and used for in vitro kinase assays with a GST-CTD substrate as previously described [21,55]. Amounts of the kinase were normalized by the signal on the FLAG tag.

Analysis of C. albicans morphology

Morphological assessment was conducted in YNBNAG (11 mM glucose) at 30°C. For the analysis of the effects of 3-MB-PP1 (the ATP analog used to inhibit analog-sensitive kinases), cells grown overnight at 30°C in YNB with 1% (w/v) glucose were pelleted by centrifugation and resuspended in 5 mL YNBNAG pH 5.1 containing 5 μM 3-MB-PP1 (EMD MILLIPORE) or DMSO as a vehicle control. The cells were then incubated at 30°C for 3h in culture tubes on a roller drum, fixed in formaldehyde, and morphology quantified from image captured using differential interference contrast microscopy. Cells were considered to be true hyphae if germ tubes had parallel sides and no invagination at the junction of the filament and the mother blastospore. A minimum of 175 cells per replicate were counted, and data presented represent three independent biological replicates conducted on three different days. Cultures were inoculated to an initial density of 1x10⁷ cells per mL.

Analysis of the C. albicans transcriptome upon Ssn3 inhibition using RNA seq

Cells were grown as described above for morphological assessment, but the incubation time was reduced to decrease indirect effects of Ssn3 inhibition. Specifically, the time following drug exposure was reduced from three hours to one hour. The 5 mL cultures were collected one hour after 3-MB-PP1 addition, pelleted by centrifugation, snap frozen in liquid nitrogen. RNA was isolated from frozen cell pellets with the MasterPure Yeast RNA Purification Kit (Epicentre MPY03100) as described in [56]. For RNA sequencing, 500 ng of total RNA was input into the Kapa mRNA HyperPrep kit (Kapa Biosystems, Wilmington, MA) and processed according to the manufacturer’s instructions. All 24 samples were multiplexed together into a single High Output 2x75bp run on a NextSeq500 instrument (Illumina, San Diego, CA). Raw reads were mapped to the C. albicans genome SC5314 (version A21-s02-m09-r04, candidagenome.org), and normalized using EdgeR. KEGG enrichment analysis was carried out using KOBAS 2.0 [57]. Data were presented after normalization by geometric mean of positive controls and geometric mean of TEF1 and ACT1 reads. Gene expression in a wild-type strain or a flo8 mutant was set to ‘1’ as mentioned in the figure legends. The accession number for the data is GSE171859.

Ssn3 phosphoproteomic experimental design

300 mL YNBNAG (11 mM glucose, pH 6.0) cultures of SC5314 and ssn3^AS were inoculated at a density of OD₆₀₀ of 0.01 and grown to OD₆₀₀ of 4.5 in flasks with shaking at 30°C, then incubated for a subsequent 4.5 hours to ensure cells had entered stationary phase. Stationary phase cells were incubated in the presence of 5 μM 3-MB-PP1 for five minutes to enable drug entry into the cell. The cells were then concentrated and added to fresh, pre-warmed medium containing either drug or vehicle and incubated for 15 minutes at 30°C with shaking. Either DMSO vehicle or 5 μM 3-MB-PP1 was then added, after which cultures were incubated with shaking for five minutes at 30°C. Then, 1.2L of prewarmed fresh YNBNAG (pH 6.0) medium was added as a 1.25X concentrate, and cells were incubated for an additional 15 minutes. The cells were then harvested by centrifugation and cell lysis by grinding under liquid nitrogen. All growth and incubation steps were conducted at 30°C, and the data represent the average of three independent replicates conducted on three separate days.

Phosphoproteomic analysis

Yeast powder was lysed in ice-cold lysis buffer ((8 M urea, 25 mM Tris-HCl pH 8.6, 150 mM NaCl, phosphatase inhibitors (2.5 mM beta-glycerophosphate, 1 mM sodium fluoride, 1 mM sodium orthovanadate, 1 mM sodium molybdate) and protease inhibitors (1 mini-Complete EDTA-free tablet per 10 ml lysis buffer; Roche Life Sciences)) and sonicated three times for 15 sec each with intermittent cooling on ice. Lysates were centrifuged at 15,000 x g for 30 minutes at 4°C. Supernatants were transferred to a new tube and the protein concentration was determined using a BCA assay (Pierce-ThermoFisher Scientific). For reduction, DTT was added to the lysates to a final concentration of 5 mM and incubated for 30 min at 55°C. Afterwards, lysates were cooled to room temperate and alkylated with 15 mM iodoacetamide at room temperature for 45 min. The alkylation was then quenched by the addition of an additional 5 mM DTT. After 6-fold dilution with 25 mM Tris-HCl pH 8, the samples were digested overnight at 37°C with 1:100 (w/w) trypsin. The next day, the digest was stopped by the addition of 0.25% TFA (final v/v), centrifuged at 3500 x g for 30 minutes at room temperature to pellet precipitated lipids, and peptides were desalted on a 500 mg (sorbent weight) SPE C₁₈ cartridge (Grace-Davidson). Peptides were lyophilized and stored at -80°C until needed for future use.

Phosphopeptide enrichment

Phosphopeptide purification was performed as previously described [58]. Briefly, peptides were resuspended in 1.5 M lactic acid in 50% ACN (“binding solution”). Titanium dioxide microspheres were added and vortexed by affixing to the top of a vortex mixer on the highest speed setting at room temperature for 1 hour. Afterwards, microspheres were washed twice with binding solution and three times with 50% ACN / 0.1% TFA. Peptides were eluted twice with 50 mM KH₂PO₄ (adjusted to pH 10 with ammonium hydroxide). Peptide eluates were combined, quenched with50% ACN / 5% formic acid, dried and desalted on a μHLB OASIS C₁₈ desalting plate (Waters). Phosphopeptide enrichment was repeated once.

TMT-labeling

Phosphopeptides were resuspended in 133 mM HEPES (Sigma) pH 8.5 and 20% acetonitrile (ACN) (Burdick & Jackson). Peptides were transferred to dried, individual TMT reagent (ThermoFisher Scientific), and vortexed to mix reagent and peptides. After 1 h at room temperature, each reaction was quenched with 3 μl of 500 mM ammonium bicarbonate solution for 10 minutes, mixed, diluted 3-fold with 0.1% TFA in water, and desalted using C₁₈ solid phase extraction cartridges (ThermoFisher Scientific). The desalted multiplexes were dried by vacuum centrifugation.

Pentafluorophenyl-based Reversed Phase HPLC

Offline PFP-based reversed phase HPLC fractionation was performed as previously described [59]. Briefly, phosphopeptides were fractionated using a Waters XSelect HSS PFP 2.5 μm 2.1 × 150mm column on an Agilent 1100 liquid chromatography system, buffer A was 3% acetonitrile / 0.1% TFA, and buffer B was 95% acetonitrile / 0.1% TFA. Flow rate was 150 μl/min with a constant column temperature of 20°C. Phosphopeptides were fractioned using a 60-minute linear gradient from 8–45% acetonitrile and collected as 48 fractions between minutes 2 and 65. The 48 fractions were then combined into 24 total samples.

TMT-based quantitative data analysis

TMT-labeled samples were analyzed on a Orbitrap Fusion [60] mass spectrometer (ThermoScientific) equipped with an Easy-nLC 1000 (ThermoScientific). Peptides were resuspended in 8% methanol / 1% formic acid across a column (45 cm length, 100 μm inner diameter, ReproSil, C₁₈ AQ 1.8 μm 120 Å pore) pulled in-house across a 2 h gradient from 8% acetonitrile/0.0625% formic acid to 37% acetonitrile/0.0625% formic acid. The Orbitrap Fusion was operated in data-dependent, SPS-MS3 quantification mode [61,62] wherein an Orbitrap MS1 scan was taken (scan range = 350–1500 m/z, R = 120K, AGC target = 2.5e5, max ion injection time = 100ms), followed by ion trap MS2 scans on the most abundant precursors for 4 seconds (max speed mode, quadrupole isolation = 0.6 m/z, AGC target = 4e3, scan rate = rapid, max ion injection time = 60ms, minimum MS1 scan signal = 5e5 normalized units, charge states = 2, 3 and 4 included, CID collision energy = 33%) and Orbitrap MS3 scans for quantification (R = 15K, AGC target = 2e4, max ion injection time = 125ms, HCD collision energy = 48%, scan range = 120–140 m/z, synchronous precursors selected = 10). The raw data files were searched using COMET with a static mass of 229.162932 on peptide N-termini and lysines and 57.02146 Da on cysteines, and a variable mass of 15.99491 Da on methionines and 79.96633 Da on serines, threonines and tyrosine against the target-decoy version of the respective FASTA database (UniProt; www.uniprot.org) and filtered to a <1% FDR at the peptide level. Quantification of LC-MS/MS spectra was performed using software developed in house. Phosphopeptide intensities were adjusted based on total TMT reporter ion intensity in each channel and log₂ transformed. P-values were calculated using a two tailed Student’s t-test assuming unequal variance.