Alternative sulphur metabolism in the fungal pathogen Candida parapsilosis

Abstract

Candida parapsilosis is an opportunistic fungal pathogen commonly isolated from the environment and associated with nosocomial infection outbreaks worldwide. We describe here the construction of a large collection of gene disruptions, greatly increasing the molecular tools available for probing gene function in C. parapsilosis. We use these to identify transcription factors associated with multiple metabolic pathways, and in particular to dissect the network regulating the assimilation of sulphur. We find that, unlike in other yeasts and filamentous fungi, the transcription factor Met4 is not the main regulator of methionine synthesis. In C. parapsilosis, assimilation of inorganic sulphur (sulphate) and synthesis of cysteine and methionine is regulated by Met28, a paralog of Met4, whereas Met4 regulates expression of a wide array of transporters and enzymes involved in the assimilation of organosulfur compounds. Analysis of transcription factor binding sites suggests that Met4 is recruited by the DNA-binding protein Met32, and Met28 is recruited by Cbf1. Despite having different target genes, Met4 and Met28 have partial functional overlap, possibly because Met4 can contribute to assimilation of inorganic sulphur in the absence of Met28.

Fig. 1. Response of C. parapsilosis to oxidative and osmotic stresses, cell wall disturbing agents, antifungal drugs, and metal toxicity.

A The pie chart represents the final collection of mutants (351 strains), with the blue segment representing the new strains generated in this study. TF: transcription factors; PKs: protein kinases; mORFs: miscellaneous ORFs; UFs: unknown function. Ed = edited strain; Δ = deleted strain. B Growth of the mutant collection was determined in conditions designed to identify different stress phenotypes (Supplementary Data 3). The growth of each strain was compared between the control plate (YPD) and each experimental plate and strains with a Z-score > 2 or < − 2 (Ln Ratio values that were greater than two standard deviations below (−) or above (+) the mean for each plate) were considered to have a growth defect (or advantage). Strains with Z-scores between − 3 and − 6 are indicated in yellow; strains with Z-score < − 6 are indicated in orange; complete absence of growth is indicated in black. Only mutant strains that showed a growth defect in at least one condition are included in the heatmap. The concentrations of the stressing agents are indicated in Supplementary Data 3. Caspo = caspofungin; fluco = fluconazole; keto = ketoconazole (C) The involvement of the protein kinase Prk1 in response to copper toxicity was confirmed in a spot assay. The growth of two independent lineages of the edited and deleted prk1 mutants (Ed prk1 and Δprk1, respectively), and of the complemented strain cPRK1 was tested on YPD supplemented with CuCl₂ (10 and 13 mM) or CuSO₄ (13 and 15 mM). The cup2 deletion mutant was included as a control. Plates are shown after 2 days of incubation at 30 °C. Source data (uncropped photographs) are provided as a Source Data file.

Fig. 2. Phylogeny of the Met4 and Met28 paralogs in budding yeasts.

A Protein sequences retrieved from YGOB, CGOB or the indicated accession numbers were aligned using ClustalO implemented in Seaview. Trees were inferred using PhyML restricted to conserved regions selected using Gblocks. WGD = Whole Genome Duplication; KLE = Kluyveromyces/Lachancea/Eremothecium. B Met4 and Met28 sequences of S. cerevisiae (Sc), C. parapsilosis (Cp) and O. parapolymorpha (Op) are drawn to scale. The A. nidulans duplicated protein pair MetR/MetZ is also included. Met4-like sequences in yeast species from the CUG-Ser1 clade, the Pichiaceae, and clades within the Saccharomycetaceae share a common domain organisation. Starting from the N-terminal end of the protein: Activation Domains (AD) encompassing Ubiquitin Interaction Motifs (UIM, which protects ubiquitinated Met4 from degradation), a conserved Lysine (K) that is the target of poly-ubiquitination, Inhibitory Region (IR, required for Met30-mediated inhibition of activity in the presence of a high concentration of methionine), Auxiliary domain (AUX, required to fully relieve IR-mediated repression), Interaction domain (INT, required for binding of Met31/Met32), and the bZIP DNA binding domain. The degenerated basic region of the bZIP domain in S. cerevisiae Met4 is indicated by the pink vertical stripes. Met28-like sequences are shorter and lack the N-terminal domain organisation observed in Met4, but they have a bZIP binding domain at the C-terminal end.

Fig. 3. Either Met4 or Met28 can sustain C. parapsilosis growth in the absence of methionine.

A C. parapsilosis CLIB214 and derivative strains disrupted in MET4, MET28, or both were tested for their ability to grow in the absence of sulphur-containing amino acids. Cell dilutions were spotted on SC media supplemented with all amino acids (SC + AA), or lacking cysteine/methionine (SC – C/M), cysteine (SC – C) or methionine (SC – M) and photographed after 48 h at 30 °C. Only disrupting both MET4 and MET28, either by deletion or by insertion of a premature stop codon (met4Δ/met28Δ and Ed met4/met28), abolished methionine synthesis. Restoration of the wildtype sequence of either MET4 (cMET4) or MET28 (cMET28) recovered methionine prototrophy. B The same result was obtained when the mutations were introduced in three additional C. parapsilosis strains from Clades 5 (02−203), 3 (81-042), and 4 (CDC179). On the contrary, disruption of MET28 in the related species C. tropicalis (CAS08-102) was enough to abolish methionine synthesis. Source data (uncropped photographs) are provided as a Source Data file.

Fig. 4. Loss of function of Met4 or Met28 affects a different subset of sulphur-responsive genes in C. parapsilosis.

The transcriptional response to cysteine/methionine (C/M) starvation was analysed by RNAseq in C. parapsilosis CLIB214 and strains in which MET4, MET28, or both were deleted. The results are presented in volcano plots (see also Supplementary Data 4, Supplementary Table 2, Supplementary Note 3 and Supplementary Fig. 6). For each gene, the log₂ fold change (log₂FC) and the -log₁₀ of the adjusted P-value (-log₁₀Padj) are indicated. Values of FC > 2 and adjusted p < 0.05 were selected to define differentially expressed genes (DEGs) between the two conditions (absence of C/M versus presence of C/M). Three dotted grey lines delineate the threshold of significance and fold change. Unaffected genes are coloured light brown, while DEGs are blue. Genes that are implicated in sulphur metabolism are highlighted in red (Supplementary Table 2). Deleting MET4 or MET28 individually affects the expression of two separate subsets of genes, circled in two clusters in the figure (categories as in Supplementary Table 2 and Supplementary Fig. 6), and the absence of both regulators completely shuts down the induction of sulphur-responsive genes: the expression of all the sulphur-related genes collapses below the threshold of significance/fold change in the met4Δ/met28Δ mutant.

Fig. 5. Met4 and Met28 regulate different branches of sulphur metabolism in C. parapsilosis.

Sulphur-responsive genes based on the transcriptomic analysis, and their involvement in the import and assimilation of sulphur sources. All the genes written in black are upregulated in response to cysteine/methionine starvation in C. parapsilosis CLIB214. The expression of the genes in grey is not significantly affected by the absence of cysteine and methionine. The DEGs are boxed in coloured rectangles to incorporate information on how deleting MET4, MET28, or both affected the network that is controlled by these regulators (Supplementary Data 4 and Supplementary Table 2). The genes are boxed in blue if their expression is induced by Met28, and in pink if they are controlled by Met4 (Supplementary Data 4 and Supplementary Table 2). A split rectangle indicates that both regulators participate (see legend below the pathway). We could not determine if MET4 regulates itself because our data was obtained in a met4Δ strain: for this reason, half of the rectangle was left white. The pathway of inorganic sulphur assimilation into methionine (methyl cycle) and cysteine (transsulfuration pathway) through sequential reduction of sulphate into sulphite and then sulphide is conserved between C. parapsilosis and S. cerevisiae, but in C. parapsilosis the genes involved are mainly controlled by Met28. Met28 also controls the O-Acetyl-Serine (OAS) pathway (in green), which leads to the synthesis of cysteine from serine and sulphide and is absent in S. cerevisiae. On the contrary, Met4 is responsible for the expression of genes with a predicted role in the import and assimilation of organosulfur compounds (e.g., sulphate esters or sulfonates) (highlighted by the dotted-lined pink rectangle). The numbers in parenthesis indicate the presence of additional members of a gene family in the genome of C. parapsilosis, which are not overexpressed in the absence of Cys/Met in our dataset. The boxes representing the transporters are colour-coded based on their putative specificity for sulphate (teal), organosulfur compounds (green and yellow), or amino acids (purple). AdenoBisphos = adenosin 3’,5’-bisphosphate; THF = tetrahydrofolate.

Fig. 6. C. parapsilosis sulphur responsive genes contain Cbf1 and Met32 binding sites.

The promoter regions (1-kb regions upstream of the starting codon) of the genes upregulated in the absence of Met/Cys were analysed using MEME-suite. The promoters are depicted as lines (0 = −1000bp from ATG; 1000 = − 1bp from ATG). The genes are listed on the left, colour-coded based on their dependency on Met4 (pink) or Met28 (light blue). Two motifs were found, corresponding to sites that are known to mediate the DNA binding of Met32 and Cbf1 in S. cerevisiae (illustrated above the graph). The distribution of the motifs is indicated by red (Met32) or blue (Cbf1) dots, located above (+ strand) or below (− strand) the promoter lines. The precise coordinates are provided in Supplementary Table 3. The promoters contain Met32 only, Cbf1 only, or both binding sites (indicated by the motif checkboard on the left-hand side; grey box: no motifs found with the selected stringency on MEME).

Fig. 7. Met4 and Met32 are required for the assimilation of organosulfur compounds in C. parapsilosis.

The growth of C. parapsilosis CLIB214 and derivative strains disrupted in MET4, MET28, or MET32 was tested in the presence of sulphur-free media supplemented with (i) no sulphur (control, no S); (ii) inorganic sulphur (sulphate, AS), or (iii) organosulfur compounds from different structural classes as only available sulphur source (MOPS, SDS, DMSO). The OD₆₀₀ of the cultures after 8 days at 30 °C is shown. Individual observations are plotted, and the median is indicated by the horizontal bars. The box represents the range of data. For each condition, 3 biological replicates were tested (n = 3), except for L. elongisporus CBS2605 and C. parapsilosis CLIB214, met4Δ and met28Δ in the absence of sulphur, in which 6 biological replicates were tested (n = 6). Detailed information on the mean and standard deviation is available in the Source Data file. One Way ANOVA and Tukey HSD test (F = 222, degrees of freedom = 29) were used for statistical analysis. ns = not significant, ****P ≤ 0.0001. The exact P-values are provided in the Source Data file. L. elongisporus CBS2605 was included as positive control. A C. parapsilosis CLIB214 can use both inorganic and organic sulphur sources. Whereas either Met4 or Met28 can sustain growth on inorganic sulphur (AS), Met4 is required for assimilating the organosulfur compounds tested, as shown by the fact that deleting this gene prevented growth on MOPS, SDS, and DMSO. Restoring the wildtype sequence in a met4-disrupted strain recovered this ability (cMET4). B Growth in the absence of cysteine or methionine confirmed that the transcription factor Met32 is not required for the synthesis of sulphur-containing amino acids. However, this protein – like Met4 – is also required for assimilating organic sulphur, as disruption of MET32 specifically abolished growth in media containing organosulfur compounds. Source data (uncropped photographs) are provided as a Source Data file.

Fig. 8. Loss of function of Met28 abolishes biofilm formation in C. parapsilosis in the absence of methionine.

Biofilms were formed in SD media with or without Met at 37 °C for 72 h, and then the biomass was weighed. Each dot represents an independent measurement. The bar represents the mean. The exact number of biological and technical replicates and the standard deviation can be found in the Source Data file. Statistical significance was calculated using the Kruskal Wallis test followed by the post-hoc Dunn’s test. Significant comparisons (p < 0.05) are shown. Source data are provided as a Source Data file.

Fig. 9. Sulphur metabolism in S. cerevisiae and C. parapsilosis.

A The expression of the MET regulon (44 genes) in S. cerevisiae depends strictly (class 1) or more loosely (class 2) on the interaction of Met4 with Cbf1-Met28 (Cbf1-Met28 dependent genes) or Met32 only (class 3, Met32-only dependent genes)^,. The fractions indicate the number of genes in each class, based on ref. . Met4 can be recruited by Met31/32 to high-affinity Met31/Met32 binding sites (class 2 and 3 genes), or by Cbf1-Met28 to variant recruitment sites, in which a Met4 recruitment motif (RYAAT) is found 2 bp upstream of the Cbf1 E-box sequence (CACGTG), and potentially recognised by Met28 (class 1 genes). The yeast vignette shows the different branches of sulphur metabolism that are controlled by genes belonging to Cbf1-Met28 dependent (in blue) and Met32-only dependent (in red) genes, or by genes that are not differentially expressed in sulphur starvation (in grey). AD = activation domain. OSCs = organosulfur compounds. Dotted lines at Met32 binding motifs (bm) in class 1 promoters indicate low-affinity sites. B Based on our transcriptional and promoter analyses, the same three classes can be identified in the MET regulon (45 genes) in C. parapsilosis. However, while Met4 drives the expression of class 3 genes when in association with Met32, Met28 is the core regulator of class 1 genes, possibly recruited by Cbf1. Class 2 genes are targets of both transcriptional complexes and both Met4 and Met28 (Fig. 5). In the absence of MET28, Met4 may allow basal expression of class 1 genes through weak binding to low-affinity Met32 binding motifs in promoters of class 1 genes (dotted-lined Met4-Met32 complex). The distribution of Cbf1-Met28 dependent and Met32-only dependent genes in the different branches of sulphur metabolism is not completely conserved between C. parapsilosis and S. cerevisiae (yeast vignette). C Schematic phylogenetic tree showing the relation of the species mentioned in the study. Met28 and Met31/32 were recruited to the sulphur regulatory network in the common ancestor of all budding yeasts (node 1) with Cbf1, and not at a later stage only in Saccharomyces and related species (node 2), as previously thought.