Screening of the expanded GRACE library identifies functionally enriched gene clusters with temperature-dependent importance for fitness

Despite the clinical importance of C. albicans, our understanding of this fungal pathogen remains incomplete due to a lack of comprehensive mutant collections to interrogate gene function. Thus, we first sought to expand the GRACE collection to improve genome coverage. Specifically, C. albicans heterozygous deletion mutants marked with strain-specific molecular barcodes [23] were transformed with a DNA construct consisting of a nourseothricin (NAT)-resistance marker, a DOX-repressible promoter, and homologous sequences to drive homologous recombination immediately upstream of the start codon of the target gene of the remaining allele (Fig 1A). Through this approach, we generated 1,240 novel GRACE strains (GRACEv3) representing 1,228 genes absent from the previous two iterations, expanding genome coverage of this collection from ~48% to ~71% (Fig 1B). For those strains present in the GRACE collection, genes were grouped into broad categories based on manual curation of their annotation status from the Candida Genome Database (CGD) in the gene ontology category of “biological process” [28]. Specifically, genes were determined to have been annotated (i) directly with experimental characterization in C. albicans, (ii) indirectly with conservation in sequence and/or protein domains to other fungal species, or (iii) unannotated based on evidence codes denoting the type of experiments or prediction data presented to support functional annotations. Construction of GRACEv1 and v2 libraries was guided by orthologous gene functions covering all major biological processes in Saccharomyces cerevisiae. Thus, unannotated genes only make up 11.6% and 15.4% of genes in these two iterations [23,25] (Figs 1B and S1A). In contrast, the construction of GRACEv3 library was not guided by gene functions in S. cerevisiae, resulting in the inclusion of 375 genes (30.6%) lacking functional annotations (Figs 1B and S1A). The remaining ~50% of unannotated genes in the C. albicans genome will benefit from continued efforts of library expansion.

Download:

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

Fig 1. Screening of expanded GRACE mutant collection at six temperatures reveals distinct functional enrichment in genes clustered by fitness profile.

A) Schematic of GRACE mutant construction. B) Summary of GRACE mutant collection expansion based on manual curation of biological process gene ontology annotation on Candida Genome Database (CGD). Pie chart with four concentric rings visualizes the relative proportion of genes from various versions of the GRACE collection with biological process descriptions completely unannotated or annotated based on: experimentally characterization, S. cerevisiae ortholog, S. pombe ortholog, orthologs in other fungal species, conserved protein domains, or unannotated; see Materials and methods for full details. C) Screening results of GRACE v1, v2, and v3 at 30 and 39 °C. Colors in stacked bar graph represent fitness scores of mutants, with a score of 4 (red) representing no growth and 0 (yellow) representing growth comparable to wild type. Representative images of mutants corresponding to scores are shown. Dotted line denotes cutoff for genes with fitness score > 1 selected for further screening. D) Heatmap of genes ordered by hierarchical clustering of essentiality scores across screening temperatures. A score of 4 (red) represents no growth and 0 (yellow) represents growth comparable to wild type. E) Dot plot representing gene ontology (GO) enrichment analysis of genes in clusters from panel D generated by hierarchical clustering. Dot size represents the number of genes identified for the given GO term and the P-value adjusted by false discovery rate is shown on a blue to red scale with red being the most significant. The data underlying this Figure can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003409.g001

With our expanded C. albicans GRACEv3 collection, we next wanted to identify genes that were important for fitness at 30 °C, as previously performed for GRACEv1 and v2 mutants [25]. To do so, the arrayed GRACEv3 library was grown in liquid culture with a high concentration (100 μg/mL) of DOX before transferring onto solid medium containing 100 μg/mL DOX in technical duplicate to ensure maximum transcriptional repression. Plates were incubated at 30 °C for 48 h. Colony images were scored independently by two researchers from 0 to 4, with 4 representing no growth and 0 representing growth comparable to wild type [25] (Figs 1C and S1B). Given the importance of thermotolerance for the success of C. albicans as a human fungal pathogen, we also screened the entire GRACE collection (v1, v2, and v3) for growth at elevated temperature (39 °C) using an analogous approach. Screening results for the GRACEv3 collection at 30 °C were compiled alongside previous results from the v1 and v2 collections [25] and compared to those obtained at 39 °C. This revealed 793 and 887 genes that were important for C. albicans fitness at 30 and 39 °C, respectively, with 734 genes that were important at both temperatures (essentiality score >1, Figs 1C and S1C).

To further interrogate the role of temperature in C. albicans fitness, we arrayed 947 strains that showed moderate to severe growth defects (average score > 1 in the presence of DOX) in our initial screen (Fig 1C). To avoid preselection for genes important for fitness during adaptation to DOX suppression, mutants were grown overnight in liquid culture in the absence or presence of 0.5 µg/mL DOX at 30 °C. Mutants were then spotted onto agar plates in the presence and absence of 100 µg/mL DOX and incubated at six different temperatures (16, 22, 30, 37, 39, and 40.5 °C). At the respective timepoints, plates were imaged, and growth of mutants was scored independently by three individuals and averaged to obtain a final growth score for each mutant (S1D Fig and S1 Data). Compared to the 887 strains that were annotated with an essentiality score >1 at 39 °C in the initial screen, 648 strains scored >1 at 39 °C in the secondary screen, suggesting lower DOX pretreatment enabled a more fine-tuned assessment of genes important for fitness at different temperatures (S1 Data). Mutants were then organized by hierarchical clustering of growth scores across screened temperatures to identify clusters of genes with minimal to moderate importance for growth (cluster 4), as well as temperature-independent (cluster 1), heat-sensitive (clusters 2 and 6), cold-sensitive (cluster 3), and extreme temperature-sensitive (cluster 5) phenotypes (Fig 1D). To identify functional enrichment within each cluster, gene ontology (GO) analysis was performed using the 947 genes screened as the background (Fig 1E). As expected, mutants displaying temperature-independent essentiality (cluster 1) were enriched for core biological processes like ‘translation’ and ‘protein metabolic process’, which are indispensable for survival. Cold tolerance genes (cluster 3) showed significant enrichment of various RNA metabolism and regulatory-related processes. Additionally, the GO term “filamentation in response to pH” was enriched in cluster 5, which consisted of genes essential for survival at both low and high temperatures. Finally, genes grouped together based on essentiality at extreme high temperatures (cluster 6) were enriched for functions related to cell cycle, filamentation, and stress response, biological processes previously shown to be important for thermotolerance [13,29]. Overall, this highlights that genes involved in similar biological processes often cluster together based on growth phenotypes at different temperatures.

Gar1 modulates translation activity enabling fungal survival

The findings from our functional genomic screen revealed a distinct set of genes required for growth, particularly at elevated temperature, many of which were uncharacterized or poorly characterized based on S. cerevisiae orthologs. One of these genes from cluster 6 was the putative H/ACA snoRNP pseudouridylation complex protein, Gar1. Pseudouridylation is the most abundant RNA modification in noncoding RNAs in all living organisms and plays essential roles in transfer RNA and ribosomal RNA (rRNA) functions [30]. To verify the importance of GAR1 for C. albicans growth at elevated temperature, we performed spot dilution assays with the tetO-GAR1/Δ strain in the absence and presence of DOX and observed a severe growth defect at all temperatures tested upon DOX-mediated transcriptional repression, particularly at 22 and 40.5 °C (Figs 2A and S2A). To confirm the importance of GAR1 for C. albicans fitness, a homozygous deletion mutant was constructed and assessed for growth at different temperatures. Similar to the GRACE strain, spot dilution assay confirmed the importance of this gene for C. albicans fitness, particularly at 40.5 °C where no growth was observed for the deletion mutant. Complementation of the deletion mutant with two wild-type alleles (GAR1A/A) resulted in growth phenotypes similar to wild type (Fig 2B).

Download:

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

Fig 2. Candida albicans GAR1 is important for rRNA maturation, translation, and virulence.

A) Strains were grown overnight in YPD with or without 0.05 µg/mL DOX and spotted in 10-fold dilutions starting from an OD₆₀₀ of 0.8 onto YPD with or without 100 µg/mL DOX. Plates were incubated at indicated temperatures and imaged after 5 days. B) Spotting assay was conducted as described in A. GAR1A/A represents the gar1Δ/Δ strain which was complemented with both A alleles. C) Cells were sub-cultured and grown to log phase at 30 °C before RNA extraction and cDNA synthesis. Total levels for the internal transcribed spacers 1 (ITS1) and 2 (ITS2) were measured by reverse transcriptase quantitative PCR (RT-qPCR) and normalized to ACT1. Error bars represent the mean ± SD for biological triplicates (* p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, One-way ANOVA Bonferroni’s correction). D) Overnight cultures were sub-cultured to an OD₆₀₀ of 0.1 in SD medium and grown at 30 °C for 4 h. Wild-type cells treated with 100 µg/mL anisomycin and 4 µg/mL fluconazole for 10 min were used as translation inhibitor and non-translation inhibitor controls, respectively. A l-homopropargylglycine (HPG) alkyne methionine analog was added, and cells were fixed. An azide fluorophore was added, and cells were imaged on the GFP channel to detect translation activity. Wild-type cells without HPG were used as no-signal control. Scale bar represents 10 µm. Fluorescence intensity was analyzed by flow cytometry using the fluorescein isothiocyanate (FITC) channel and plotted as a histogram. The median FITC signal intensity for each sample is listed in the table. E) Y-MAP plot visualizing chromosome copy number and loss of heterozygosity events from WGS analysis. Haplotypes relative to the reference genome SC5314 are color-coded (see legend). F) Spotting assay was conducted as described in A, except the plates were imaged after 3 days. G) Strains were inoculated at 1 × 10⁶ cells per larva with 20 larvae per group. A mock infection using PBS served as a control. Larvae were incubated at indicated temperatures and survival was monitored. Log rank (Mantel-Cox) test was performed to test significance between wild type and gar1Δ/Δ groups. The data underlying Fig 2C, 2D and 2G can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003409.g002

In S. cerevisiae, the box H/ACA ribonucleoprotein complex converts uridine into pseudouridine in rRNAs and consists of four core proteins (Cbf5, Nop10, Nhp2, and Gar1), one RNA-binding protein (Naf1), and one chaperone (Shq1) [31–33]. Spot dilution assays confirmed the importance of the C. albicans core genes for growth at all temperatures. In contrast, the chaperone SHQ1 had the least impact on fitness across all temperatures compared to other genes of the box H/ACA ribonucleoprotein complex (S2A and S3 Figs). These observations suggest Gar1 may be partially dispensable for complex function under optimal conditions, but becomes absolutely required at extreme temperatures.

The box H/ACA ribonucleoprotein complex is important for rRNA pseudouridylation modifications and maturation in S. cerevisiae [33,34]. To investigate the role of Gar1 in C. albicans, we quantified the presence of precursor internal transcribed spacers (ITS1 and ITS2) by qRT-PCR. During rRNA maturation, ITS1 and ITS2 are cleaved from precursor rRNAs to form mature 18S, 5.8S, and 25S rRNAs [35]. Deletion of GAR1 resulted in significant increases in ITS1 and ITS2 relative to wild-type and complemented strains (Fig 2C), similar to what was observed upon depletion of the core H/ACA ribonucleoprotein CBF5 (S2A and S2B Fig), supporting a role of Gar1 in rRNA maturation (Fig 2C). Defects in rRNA maturation ultimately result in impaired translation [36]. Hence, we performed a fluorescence-based assay to monitor translation activity in the presence or absence of GAR1 [37]. An alkynylated methionine analog, l-homopropargylglycine (HPG), was used to label newly synthesized proteins, and a click reaction of the alkyne with a fluorescent azide provided a fluorescent readout of the translation activity. Treatment of wild-type cells with the translation inhibitor anisomycin or deletion of GAR1 reduced the fluorescent signal compared to wild type or the complemented control. Wild-type cells treated with the non-translation inhibitor fluconazole did not show reduced fluorescent signal, confirming the reduction in fluorescence intensity was not simply due to reduced cell growth (Fig 2D).

To further elucidate the molecular mechanism underlying Gar1’s role in temperature-dependent essentiality, we devised a selection strategy to identify mutants with a restored ability to grow at elevated temperature. To this end, we plated 10⁸ gar1Δ/Δ cells on agar plates, incubated the cells at 42 °C for 14 days, and successfully isolated one bypass mutant (ByM1) that was able to grow at 40.5 °C (Figs 2B and S4A) and partially restored ITS1 and ITS2 to wild-type-like levels (Fig 2C). Whole genome sequencing analysis of the evolved mutant showed no aneuploidy or copy number variation in comparison to the reference SC5314 genome or compared to the gar1 mutant from which it was derived (Fig 2E). Variant calling identified a mutation in KRR1 in the LOH region (cyan) on Chr1 at position 109, resulting in an amino acid substitution at position 37 from glutamic acid (E) to lysine (K). KRR1 encodes a nucleolar protein functioning in the 90S pre-ribosome facilitating pre-rRNA cleavage during rRNA maturation [38]. To confirm this substitution was sufficient for the restored growth observed at elevated temperature, we replaced both KRR1 alleles in the gar1 mutant with those identified in ByM1 (Krr1^E37K) and confirmed this strain was able to grow at high temperature (Fig 2F). To confirm this substitution was necessary for the growth at elevated temperatures, we replaced both KRR1 alleles in ByM1 with those from the parental gar1 mutant (Krr1^K37E), which resulted in a loss of growth at 40.5 °C (Fig 2F).

Finally, we evaluated whether the temperature-dependent essentiality of GAR1 plays a role in virulence using the wax moth Galleria mellonella model at both 28 and 39 °C [27]. This invertebrate model has successfully been used to study C. albicans pathogenesis and highlight those genes important for C. albicans virulence in a temperature-dependent manner [27,39]. We infected larvae by proleg injection with wild-type, gar1Δ/Δ, or complemented strains and incubated the G. mellonella at 28 or 39 °C. Deletion of GAR1 significantly attenuated virulence at both temperatures compared to wild-type and the complemented strain with a greater significance at 39 °C, confirming the importance of Gar1 in C. albicans pathogenesis in vivo, especially at human febrile temperature (Fig 2G). Overall, these experiments support the role of Gar1 in thermal tolerance and highlight the increased dependency on translation and rRNA processing at elevated temperature for C. albicans.

C1_11680C encodes Ysf3 and modulates splicing activity

While the ability of C. albicans to survive within the human host requires it to thrive at physiological and febrile temperatures, this organism also causes a significant number of chronic mucocutaneous infections on the skin, nails, and other mucous membranes, and can survive on different environmental surfaces [40,41]. This requires C. albicans to grow at temperatures below 37 °C. Thus, to investigate which genes are specifically important for growth at lower temperature (cluster 3), we focused on the uncharacterized gene C1_11680C lacking functional annotation on CGD [28]. Using spot dilution assays on YPD agar plates with and without DOX, we confirmed transcriptional repression of C1_11680C impaired fitness at all temperatures, with a more severe growth defect observed at temperatures below 37 °C (Figs 3A and S2A). While we were unable to identify a potential homolog based on DNA sequence comparisons with S. cerevisiae, BLASTp searches against S. cerevisiae and Schizosaccharomyces pombe genomes predicted C1_11680c as a splicing factor (Ysf3) belonging to the SF3b subcomplex of the U2 snRNP [42–45]. Template modeling (TM) of the predicted C. albicans C1_11680c structure with S. cerevisiae Ysf3 and S. pombe Sap10 by AlphaFold [46] yielded TM-align [47] scores above 0.7, supporting C1_11680c as Ysf3 in C. albicans (Fig 3B).

Download:

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

Fig 3. C1_11680C encodes Ysf3 in Candida albicans and modulates splicing.

A) Spotting assay performed as described in Fig 2A. B) Protein structures of AlphaFold-generated C. albicans C1_11680c (UniProt: A0A1D8PF89), S. cerevisiae Ysf3 (UniProt: P0C074), and S. pombe Sap10 (UniProt: Q9P7R6). Dark blue to yellow to red indicates a reducing confidence score based on the predicted local distance difference test. TM score was calculated by TM-align. A score between 0.5 and 1 implies significant similarity in structure. C) Cells were sub-cultured and grown to log phase in the absence or presence of 20 µg/mL DOX at 22 °C or 37 °C before RNA extraction and cDNA synthesis. Primers flanking the ACT1 intron were used to amplify the spliced and unspliced products from cDNA. D) Principal component analysis of gene expression between conditions. The data underlying Fig 3D can be found in S2 Data. E) Scheme illustrating intron retention (IR) fraction quantification from RNA-Seq data and intron architecture. The cumulative distribution shows the IR fraction of 389 introns passing the read count cutoff in all samples. The data underlying Fig 3E can be found in S3 Data. F) Correlation between Δ IR fraction (+/- DOX) at 37 and 22 °C in tetO-YSF3/Δ. The data underlying Fig 3F can be found in S4 Data. G) IR difference (± DOX) of wild type (WT) and tetO-YSF3/Δ strain at 22 °C of introns categorized by their length, BP motif, splice site motif and BP-to-3′-splice-site-distance, respectively. Top numbers indicate introns per group. BP-3′SS distance groups have equal bin size. Significances obtained using Wilcoxon rank-sum test (intron lengths and motifs) and paired Wilcoxon rank-sum test with Benjamini–Hochberg correction (BP-3′SS distance), ** p ≤ 0.01, **** p ≤ 0.0001. The data underlying Fig 3G can be found in S4 Data. H) Pie chart of unspliced or spliced intron-containing gene expression in tetO-YSF3/Δ without and with DOX at 22 °C. The data underlying Fig 3H can be found in S5 Data. I) Cumulative distribution of log2-fold change in gene expression between tetO-YSF3/Δ with and without DOX for the indicated gene groups at 22 °C. Significances are relative to the ‘all’ genes (Kolmogorov–Smirnoff test). The data underlying Fig 3I can be found in S5 Data. J) Significantly enriched ‘Biological process’ GO terms among upregulated genes relative to the genome. GO: 0070682: proteasome regulatory particle assembly; GO: 0010499: proteasomal ubiquitin-independent protein catabolic process; GO: 0061136: regulation of proteasomal protein catabolic process; GO: 0043161: proteasome-mediated ubiquitin-dependent protein catabolic process. The data underlying Fig 3J can be found in S6 Data. Numerical data and source gel images underlying this Figure can be found in S1 Data and S1 File Raw_Images.

https://doi.org/10.1371/journal.pbio.3003409.g003

In S. cerevisiae, the SF3b complex consists of six subunits (Hsh155, Cus1, Rse1, Hsh49, Ysf3, and Rsd3) that play a role in recognizing branch point sequences on pre-mRNA during the early stages of spliceosome assembly [48]. Spot dilution assays confirmed that across all temperatures depletion of RDS3 and RSE1 had little to no impact on fitness, whereas depletion of HSH49 and HSH155 caused severe growth defects, with HSH155 being essential (S2A and S5A Figs). Interestingly, depletion of CUS1 resulted in similar growth phenotype across all temperatures tested to depletion of YSF3 (S5A Fig). This highlights the differential dependency of various splicing genes in fitness across different temperatures.

To test the hypothesis that C1_11680C encodes Ysf3 and is important for C. albicans splicing, we examined the splicing efficiency of the ACT1 transcript upon depletion of YSF3. We also included the tetO-PRP19/Δ strain, as PRP19 is known to be important for splicing of ACT1 in C. albicans [49]. To avoid potential off-target effects by DOX, 20 µg/mL was used, which was sufficient to deplete YSF3 transcript levels in the tetO-YSF3/Δ GRACE strain (S5C Fig) [50]. ACT1 was efficiently spliced in the wild-type control in the absence or presence of DOX, suggesting DOX has no impact on splicing efficiency of the ACT1 transcript (Fig 3C). In the tetO-YSF3/Δ and tetO-PRP19/Δ strains, we observed an accumulation of unspliced ACT1 in the presence of DOX at both 22 and 37 °C, supporting that Ysf3 is important for splicing in C. albicans (Fig 3C).

While it has been previously reported that high temperatures reduce splicing efficiency in C. albicans leading to enhanced intron retention (IR) [49], the link between low temperature (temperatures below 30 °C) and splicing efficiency has yet to be investigated. To explore the impact of low temperature on global IR and assess the importance of Ysf3 for global pre-mRNA splicing, we employed 150 bp paired-end RNA-Seq on polyadenylated RNA to map splicing of all intron-containing genes using the tetO-YSF3/Δ GRACE strain and the wild-type control in the presence and absence of 20 µg/mL DOX at both 22 and 37 °C. Principal component analysis of gene expression profiles showed that biological replicates behaved similarly and clustered together (Fig 3D). Furthermore, a significant reduction of YSF3 RNA expression upon DOX treatment led to altered gene expression profiles and the formation of distinct clusters at the two tested temperatures (Fig 3D and S2 Data). Confirming the RT-PCR results (Fig 3C), we detected increased intronic reads for ACT1 when YSF3 was depleted by DOX (S6A Fig). To assess the degree of pre-mRNA splicing globally, we performed a junction-read based analysis to estimate the fraction of IR in all 8 conditions (Fig 3E). This revealed that most introns are well spliced in wild type and the tetO-YSF3/Δ strain in the absence of DOX (median IR fraction for WT: 0.02 (22 °C)/0.04 (37 °C); for tetO-YSF3/Δ without DOX: 0.02 (22 °C)/0.05 (37 °C)) (S3 Data). A modest, but significant increase in IR was seen at 37 °C in wild type and the tetO-YSF3/Δ GRACE strain without DOX, as reported previously (Wilcoxon rank-sum test, p < 0.0001****, Figs 3E and S6B) [49]. However, depletion of YSF3 at both temperatures increased IR substantially to median IR fractions of 0.43 and 0.39 at 22 °C and 37 °C, respectively (Fig 3E). IR differences upon depletion of YSF3 were similar and well correlated between all intron-containing genes at both temperatures; however, slightly more pronounced at 22 °C than at 37 °C (Fig 3F and S4 Data).

According to the function of S. cerevisiae Ysf3 within the SF3b complex, we hypothesized that C. albicans Ysf3 could be involved in recognizing branchpoint sequence (BPS) and influence splicing of introns with diverse cis-regulatory features as the U2 snRNP [45,48]. To assess this, we analyzed the impact of YSF3 depletion on the accumulation of introns with diverse features, such as intron length, BPS to 3′ splice site (SS) distance, SS consensus, polypyrimidine tract length, and composition and intronic GC-content (S6C and S6D Fig and S4 Data). All features could be extracted from the annotated intronic sequence (Methods). To analyze C. albicans BPSs, we scored intronic 7-nt (nucleotide) motifs according to their similarity to the 7-nt BPS found in S. cerevisiae and also reported for C. albicans [51,52]. For most introns, a single motif close to the intron end (3′ SS) was detected, consistent with the evolutionarily conserved distance of BPS and 3′ SS [52]. For introns with multiple high-scoring 7-nt motifs, the motif was chosen that had the closest match to the typically observed distance of 15−74 nts (median 24 nts) to the 3′ SS and the higher score. As reported previously, the consensus BPS is 5′-UACUAAC-3′ and can be found in >80% of introns (S3 Data) [51,52]. Consistent with our hypothesis, IR and IR differences were increased for introns with non-consensus BPSs and for BPSs that are more distal or proximal to the 3′ SS than the typical branchpoint at both temperatures (Figs 3G and S6D). Interestingly, variation in other intronic features, such as intron length, and 5′ SS sequence also correlated with variation in IR difference (Figs 3G and S6C and S6D). This analysis suggests C. albicans Ysf3 is an important splicing factor, including for those introns that are short and/or exhibit nonconsensus cis-regulatory elements.

Given the high prevalence of unspliced mRNAs upon YSF3 depletion, the incompletely processed mRNAs could produce incomplete and possibly toxic, truncated or misfolded proteins that could impact growth. Combining mRNA expression and IR estimates for all intron-containing genes passing our minimal read depth cutoffs for splicing and expression quantification, we estimated that the pool of unspliced, adenylated, and possibly translated cytoplasmic mRNAs increased 6-fold (from 4.1% to 23.9% at 22 °C) and 4-fold (from 5.2% to 20.3% at 37 °C) upon YSF3 depletion (Figs 3H and 6E and S5 Data). This could lead to an increase of prematurely terminated translation products that could be cleared by nonsense-mediated decay (NMD) or require increased proteasome capacity [53,54]. Although we did not see evidence for increased expression of NMD components, genes associated with proteasomal function were increased in their expression relative to all expressed genes and enriched in the group of upregulated genes (Figs 3I, 3J, and S6F and S6 Data). Among downregulated genes, genes involved in filamentous growth and cellular responses to the environment were enriched, consistent with the previously established link between splicing and filamentation [49] (S6G Fig). Interestingly, depletion of YSF3 significantly reduced filamentation in YPD supplemented with 10% newborn calf serum at 37 °C compared to wild-type controls (S5B and S5C Fig). This phenotype was similar to what was observed upon depletion of EFG1, a transcription factor known to be required for filamentation. Overall, we conclude that CaYsf3 is important for splicing of most intron-containing genes with particular importance for short introns and/or nonconsensus cis-regulatory elements.

Rapid evolution suggests C6_00110C plays a key role in genome stability in C. albicans

Uncharacterized Candida clade-specific genes essential for growth can enable discovery of new biology relevant to developing antifungal treatments. Hence, we turned to a Candida clade-specific gene C6_00110C that was identified as important for growth at elevated temperature (cluster 6). As observed by spot dilution assays, transcriptional repression of C6_00110C with DOX abrogated growth only at 40.5 °C (Fig 4A). To confirm the temperature-specific essentiality phenotype, we generated a homozygous deletion mutant of C6_00110C, as well as a complemented strain. When examining the growth phenotype of these mutants, we confirmed that deletion of C6_00110C resulted in little to no growth at temperatures at or above 37 °C (Fig 4B). Thus, we named the gene Required for Heat Tolerance 1 (RHT1). BLASTp analysis, conserved domain searches, and co-expression data [55] failed to predict a function for Rht1 and AlphaFold prediction yielded a structure with low model confidence scores for the majority of the amino acids in the protein sequence (pLDDT < 70) (Fig 4C), leaving the function of this Candida-specific protein elusive. Thus, we turned to our bypass mutant selection strategy to uncover the function of this elusive gene and plated 10⁸ rht1Δ/Δ cells on YPD agar plates and incubated at 42 °C for 5 days. Through this approach, we evolved 10 independent bypass mutants (ByMs) with a restored ability to grow at 42 °C in the absence of RHT1 (Fig 4D). Interestingly, when these ByMs were passaged at 30 °C, colonies with restored high-temperature sensitivity (ByMRs) were readily isolated (S4B Fig). ByMRs were successfully isolated for nine out of the 10 ByMs (Figs 4D and S4B).

Download:

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

Fig 4. The importance of C6_00110C for high temperature growth can be bypassed via rapid evolution.

A) Spotting assay was performed as described in Fig 2A except plates were incubated for 2 days. B) Spotting assay was performed as described in Fig 2A. C6_00110CA/A represents the C6_00110cΔ/Δ strain which was complemented with both A alleles. C) Protein structure of AlphaFold-generated Candida albicans C6_00110c (UniProt: A0A1D8PP93). The color scale from dark blue to yellow to red indicates a reducing confidence score based on the predicted local distance difference test (pLDDT). D) Spotting assay was performed as described in panel A. E) Y-MAP plot visualizing chromosome copy number and loss of heterozygosity events from WGS analysis. Haplotypes relative to the reference genome SC5314 are color-coded (see legend). F) Spotting assay was performed as described in panel A except plates were imaged after 24 h.

https://doi.org/10.1371/journal.pbio.3003409.g004

To decipher the genetic mechanisms governing the growth phenotypes observed for both the ByMs and ByMRs, we performed whole genome sequencing of all relevant lineages. When examining the genomic changes of the ByMs, we observed significant chromosomal copy number changes for all ByMs. In particular, we detected trisomy of Chr1, Chr2, Chr4 and Chr5 for ByMs 1, 2, 6, 7, 8, and 9; trisomy of Chr1, Chr2, and Chr4 for ByM3; segmental monosomy of Chr7 and ChrR for ByM4; and trisomy of Chr1 and Chr5 as well as segmental trisomy of Chr2 for ByM5 (Fig 4E). When analyzing chromosome copy numbers for ByMRs including 1R, 2R, 3R, 6R, 7R, 8R, and 9R, all strains shared the concurrent loss of Chr1 and Chr2 trisomies, despite differences in Chr3, Chr4, and Chr7 trisomies (Fig 4E). ByM4R and ByM5R gained segmental and full length Chr7 and ChrR trisomies, respectively (Fig 4E). While aneuploidies were detected for ByM10, variant-calling identified a heterozygous mutation in IML3. Iml3 was recently characterized as a C. albicans kinetochore component and is required for heat tolerance [27]. The identified mutation resulted in a substitution of leucine (L) to serine (S) at the amino acid position 67 (Fig 4F). Substitution of S back to L at position 67 in ByM10 abrogated growth at high temperature, whereas both heterozygous and homozygous substitution of L to S at position 67 in the rht1Δ/Δ background restored growth at 40.5 °C, confirming the mutation in Iml3 was both necessary and sufficient for bypassing high-temperature essentiality in rht1Δ/Δ (Fig 4F). Overall, the ploidy changes and mutation in a kinetochore gene observed in rht1Δ/Δ mutants grown at elevated temperature suggests Rht1 may be involved in chromosome segregation and/or genome stability.

Given a potential role for Rht1 in genome stability, we examined whether this gene was important for cell cycle progression. Interestingly, the rht1Δ/Δ strain produced both yeast and filaments under non-filament-inducing conditions with the filaments resembling wild-type cells treated with cell cycle inhibitors hydroxyurea (HU) or nocodazole (NOC), suggesting a defect in cell cycle progression in the absence of RHT1 (Fig 5A). Furthermore in a dose-response assay, deletion of RHT1 resulted in a higher sensitivity to G1/S phase inhibitors HU and methyl methanesulfonate (MMS) [56,57] and the G2/M phase inhibitor NOC [58], supporting the potential role of Rht1 in cell cycle progression (Fig 5B). Higher temperature further exacerbated sensitivity of the rht1Δ/Δ mutant to all three drugs (Fig 5B), reinforcing the greater requirement of cell cycle-related genes at high temperature. While the complemented strain fully restored tolerance to all three drugs, ByM1 and ByM10 only partially restored tolerance to HU and MMS (Fig 5B).

Download:

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

Fig 5. Rht1 functions in cell cycle progression and modulates fluconazole susceptibility.

A) Candida albicans strains were sub-cultured to an OD₆₀₀ of 0.2 in YPD and grown at 30 °C for 4 h. Cells treated with 25 mM hydroxyurea (HU) or 50 µM nocodazole (NOC) were used as controls. Scale bar represents 20 µm. B) Dose-response assays were performed with 2-fold titrations of hydroxyurea, nocodazole, and methyl methanesulfonate (MMS) at 30 and 35 °C. OD₆₀₀ was measured after 48 h, values were normalized relative to wild type without drug (see color scale bar). C) Significant interactions for Rht1-GFP were defined through SAINTexpress analysis compared with the Eno1-GFP control. (BFDR threshold = 0, average spectral counts ≥ 20, fold change relative to Eno1 ≥ 50). Protein-labeled nodes are colored based on protein function annotations. Line color ranging from dark black to light gray reflects the decreasing fold change in peptide count of Rht1 relative to Eno1. The data underlying Fig 5C can be found in S7 Data. D) G. mellonella (20 larvae per group) were inoculated with 1 × 10⁶ C. albicans cells. Larvae were incubated at 28 and 39 °C and monitored for survival. Log rank (Mantel-Cox) test assessed significant difference between wild type and rht1Δ/Δ. E) Dose-response assay was performed as in B. After 48 h, 5 μL was removed from liquid cultures and spotted onto drug-free agar to assess cidality. Plates were placed in 30 °C for 48 h. F) Wild-type (SC5314) or rht1Δ/Δ cells were spread on YPD agar and an ETEST strip of fluconazole was applied. Plates were incubated at 30 or 35 °C and imaged after 3 days. G) Strains were spread on YPD agar and 50 µg of fluconazole was spotted on filter paper. Plates were incubated at 30 or 35 °C and imaged after 3 days. The data underlying Fig 5B, 5D, and 5E can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003409.g005

Identification of protein interaction partners can provide further insights into gene function. Hence, we performed affinity purification coupled with mass spectrometry (AP-MS) of a C-terminally GFP-tagged Rht1 strain to define the protein’s interactome using Eno1-GFP as a control for background [25,27]. The homozygous tagged protein was expressed and fully rescued the rht1Δ/Δ high-temperature essentiality phenotype, confirming functionality of Rht1-GFP (S7A and S7B Fig). AP-MS identified 62 high-confidence protein interactors (Bayesian False Discovery Rate [BFDR] = 0, average spectral counts ≥20, and fold change relative to Eno1 ≥50; S7 Data). Consistent with our genetic evidence, the majority of the interactors were found to function in chromatin remodeling, DNA replication and repair, cell cycle progression, transcription, and transcription regulation (Fig 5C). A total of 29 genes encoding Rht1-interacting proteins were included in the secondary fitness screen across six different temperatures. Twelve genes were found in cluster 1 (genes important for fitness across all temperatures) and 15 genes were found in clusters 2 or 6 (genes important for fitness at elevated temperatures), including six genes encoding proteins of the Remodels the Structure of Chromatin (RSC) complex (S7C Fig). Interestingly, five of the six RSC complex members are located on Chr1 or Chr2, chromosomes that increased in copy number upon growth of the rht1 mutant at elevated temperature.

Rht1 affects C. albicans virulence and influences fluconazole activity

The specificity of Rht1 to the Candida clade, its requirement for growth at high temperatures, and its implication in genome stability motivated assessment of its role in virulence and antifungal susceptibility. To determine the role of Rht1 in C. albicans virulence, we turned to the G. mellonella model of C. albicans infection. While injection of the wax moths with the wild-type and complemented strains resulted in profound lethality, the rht1Δ/Δ strain had significantly attenuated virulence only at 39 °C, confirming that Rht1 is required for C. albicans pathogenesis at febrile temperature (Fig 5D).

Genome stability modulates fluconazole susceptibility in C. albicans [59], and therefore, we wanted to investigate whether Rht1 was important for azole activity. Dose-response assays with fluconazole showed that loss of Rht1 rendered fluconazole cidal only at 35 °C (Fig 5E). Similarly, when we performed an ETEST strip assay and disc diffusion assay by spotting 50 µg of fluconazole onto a disc placed on a lawn of C. albicans, we confirmed that Rht1 is important for azole activity at 35 °C, as the deletion mutant grew in the absence of fluconazole at 35 °C and showed enhanced clearance around the fluconazole strip or disc relative to wild type (Fig 5F and 5G). The zone of clearance remained after 5 days of incubation at 35 °C (S8A and S8B Fig). Interestingly, growth around the fluconazole disc was restored in ByM10 at 35 °C but not in ByM1 (Fig 5G), suggesting aneuploidy could not fully mask the requirement of Rht1 in fluconazole susceptibility, but the substitution in Iml3 could.

Fungal strains and growth conditions

Strains used in this study are listed in S1 Table. Plasmids used in this study are listed in S2 Table. Strains were frozen in 25% glycerol at −80 °C for long-term storage. Active cultures were maintained on solid yeast extract peptone (YPD; 1% yeast extract, 2% bactopeptone, 2% glucose, and 2% agar) at 4 °C for no longer than 1 month. Liquid cultures were grown in YPD at 30 °C, unless otherwise indicated. To select for NAT-resistant mutants, NAT (Jena Bioscience, 96736-11-7) was solubilized in water and supplemented into YPD agar plates (150–250 μg/mL). Prototrophic colonies were selected on SD plates (0.674% yeast nitrogen base without amino acids with ammonium sulfate, 2% glucose, and 2% agar). Temperature-dependent fitness screens were performed using minimal medium plates (0.17% yeast nitrogen base without amino acids and without ammonium sulfate, with 2% glucose, 0.1% glutamic acid (MSG), 0.002% histidine, and 2% agar). DOX (Bio Basic, DB0889) stock was prepared at a concentration of 50 mg/mL in H₂O, filter-sterilized, and added to the media at the indicated concentration.

Manual curation of gene ontology (GO) annotation by evidence code

All genes in the C. albicans genome were analyzed for GO annotations by biological functions on CGD website [28]. Genes with GO terms assigned with codes IC (inferred by curator), IDA (inferred by direct assay), IEP (inferred from expression pattern), IGI (inferred from genetic interaction) in C. albicans, IMP (inferred from mutant phenotype), IPI (inferred from physical interaction), and TAS (traceable author statement) were categorized in the “experimentally characterized” group. Genes with GO terms assigned based on evidence in S. cerevisiae with codes IEA (inferred from electronic annotation), IGI, ISA (inferred from sequence alignment), ISM (inferred from sequence model), ISO (inferred from sequence orthology), and ISS (inferred from sequence or structural similarity) were categorized in the “putative based on S. cerevisiae ortholog” group. Genes with GO terms assigned based on evidence in S. pombe with code IEA were categorized in the ‘putative based on S. pombe ortholog’ group. Genes with GO terms assigned based on evidence in other fungal species with codes IEA, ISA, ISO, and ISS were categorized in the ‘putative based on other fungal species’ group. Genes with GO terms assigned based on conserved protein domains with codes IEA were categorized in the “putative based on conserved protein domains” group. Genes with GO terms assigned codes ISA based on nonfungal species, ISM, and NAS (nontraceable author statement) were categorized in the ‘other’ group. Genes with no GO terms assigned were categorized in the “no data available” group.

Strain construction

Screening GRACE collection for growth at 30 and 39 °C

C. albicans GRACE strains in the GRACEv3 library were screened for growth at 30 °C, as previously described [25]. Briefly, GRACE strains were pinned into 96-well plates (Sarstedt) containing liquid YPD medium and grown overnight at 30 °C. The next day, cells were transferred into liquid YPD medium in the absence and presence of 100 µg/mL DOX in technical duplicates and grown at 30 °C overnight. Approximately 2 μL of cultures were then transferred onto minimal media agar plates in the absence and presence of 100 µg/mL DOX using a metal replicator and grown for 2 days 30 °C. The plates were imaged on a ChemiDoc, and growth fitness was scored independently by two people, using the scoring system outlined in Fig 1C. The entire GRACE library was screened for growth fitness at 39 °C following the same screening pipeline except that cells were grown on minimal media agar plates in the absence and presence of 100 µg/mL DOX at 39 °C (S1B Fig).

Screening prioritized GRACE strains for growth fitness at six temperatures

C. albicans GRACE strains with a fitness score above 1 in the presence of DOX from the growth screens at 30 and 39 °C were prioritized for a secondary screen at six temperatures (S1D Fig). 947 strains were arrayed and pinned into 96-well plates (Sarstedt) containing liquid minimal medium in the absence and presence of 0.5 µg/mL DOX and grown overnight at 30 °C. The next day, 5 µL of mixed cells were diluted in 95 µL H₂O and then 5 µL diluted cells were inoculated onto minimal media agar plates in the absence and presence of 100 µg/mL DOX using a multi-channel pipettor and grown for 10 days at 16 °C, 4 days at 22 °C, 2 days at 30 °C, 2 days at 37 °C, 3 days at 39 °C, and 4 days at 40.5 °C. The plates were imaged on a ChemiDoc, and growth was scored independently by three people, using the scoring system outlined in Fig 1C. Four biological replicates were performed for growth screens at 16 and 22 °C, and three biological replicates were performed for growth screens at 30, 37, 39, and 40.5 °C.

Clustering and GO enrichment analysis

Genes represented by the 947 strains screened at the six distinct temperatures were clustered based on fitness scores in six temperatures by Euclidean distance (k = 6). GO enrichment analysis was performed using CGD for genes in each cluster against the background of 947 genes screened. P-value was adjusted by false discovery rate (FDR), and only terms with p-adjusted values greater than 0.05 were plotted. Count represents the total number of genes assigned to a given GO term.

BLAST analysis and protein structure analyses

Blast analyses were performed for C1_11680C and C6_00110C DNA and protein sequences using NCBI (https://blast.ncbi.nlm.nih.gov/Blast.cgi) [89], FungiDB (https://fungidb.org/fungidb/app/workspace/blast/new) [90], and SGD (https://www.yeastgenome.org/blast-sgd) [91] blast tools. AlphaFold [46] predicted structures were obtained from UniProt31 Website (https://www.uniprot.org/). Similarity between pairs of protein structures was assessed through TM-Align (https://zhanggroup.org/TM-align/) [47].

Spot dilution assay

Wild-type and GRACE strains were grown in YPD in the absence and presence of 0.05 µg/mL DOX overnight at 30 °C. Aliquots of each overnight culture (1 mL) were washed once in H₂O and diluted to an OD₆₀₀ of 0.8. Cells were 10-fold serially diluted four times in H₂O in a 96-well plate and 3 µL of cells from each dilution were spotted on YPD agar in the absence and presence of 100 µg/mL DOX. Cells from YPD overnight were spotted on YPD agar and cells from YPD supplemented with 0.05 µg/mL DOX overnight were spotted on YPD agar supplemented with 100 µg/mL DOX. YPD agar plates for box H/ACA ribonucleoprotein complex and SF3b complex GRACE strains were grown for 5 days at 30 °C. YPD agar plates for the tetO-RHT1/Δ GRACE strain were grown for 2 days at 30 °C. Spot dilution assays were performed the same way for wild type, deletion mutant, complemented strains, and bypass mutants without DOX.

Bypass mutant selection

To evolve bypass mutants, single colonies were cultured in 3 mL YPD under shaking conditions at 30 °C for 48 h and 24 h for CaLC10712 and CaLC10720, respectively. Cells were washed once in H₂O and 10⁸ cells were plated on YPD agar media and incubated at 42 °C for 2 weeks and 5 days, respectively. Colonies were propagated on YPD agar and grown at 30 °C to isolate single colonies. A single colony was used to validate bypass phenotype, isolate genomic DNA extraction, and archive in 25% glycerol for −80 °C storage. For rht1Δ/Δ bypass mutants, lineages with restored high temperature sensitivity were obtained from the YPD agar plate during isolation for single colonies (S4 Fig).

Quantitative reverse transcriptase PCR (qRT-PCR)

To confirm DOX-dependent transcriptional repression, wild-type and GRACE strains were grown in YPD in the absence and presence of 0.05 µg/mL DOX overnight at 30 °C, shaking. The next day, the cultures were sub-cultured to an OD₆₀₀ of 0.2 in 10 mL YPD in the absence and presence of 100 µg/mL DOX and grown for 4 h at 30 °C. GRACE mutants for essential genes were sub-cultured to an OD₆₀₀ between 0.3 and 0.4 in 10 mL YPD in the presence of 100 µg/mL DOX for 4 h at 30 °C. To confirm DOX-dependent transcriptional repression for YSF3 and EFG1 GRACE strains used in the filamentation assay, cells were grown as above except on the second day, cells were sub-cultured in YPD in the absence and presence of 20 µg/mL DOX. For ITS1 and ITS2 quantification, overnight cultures of wild-type, gar1Δ/Δ, GAR1A/A, and gar1Δ/Δ ByM1 strains were sub-cultured to an OD₆₀₀ of 0.2 in 10 mL YPD and grown for 4 h at 30 °C. Cells were pelleted, washed once with cold H₂O, flash frozen, and stored at −80°C. Cells were lysed by bead beating four times for 30 s, with one minute on ice between. RNA was extracted using the QIAGEN Rneasy kit and Dnase-treated using the Invitrogen DNA-free DNA removal kit. cDNA synthesis was performed using the iScript cDNA synthesis Kit (Biorad). Standard curve was performed for each pair of primers using wild-type samples to determine the optimal cDNA dilutions for qRT-PCR. cDNAs were diluted 80-fold for expression analysis and 800-fold for ITS1 and ITS2 level quantification. qRT-PCR was performed in technical triplicate with a 10 µL reaction volume in a 384-well plate, using Fast SYBR Green Master Mix (Applied Biosystems) and the BioRad CFX-384 Real Time System. The following cycling conditions were used: 95 °C for 3 min, then 95 °C for 10 s and 60 °C for 30 s, for 40 cycles. The melt curve was completed with the following cycle conditions: 95 °C for 10 s and 65 °C for 10 s with an increase of 0.5 °C per cycle up to 95 °C. The primers used to monitor expression are listed in S8 Data.

Translation assay

The fluorescent translation assay was performed with the Click-iT HPG Alexa Fluor 488 Protein Synthesis Assay Kit (Thermo Fisher C10428), as previously described [37]. Overnights of SC5314, CaLC10712, and CaLC10713 were sub-cultured to an OD₆₀₀ of 0.2 into 5 mL synthetic defined (SD) medium without amino acids, without ammonium sulfate supplemented with 2% glucose, monosodium glutamate, and uridine and grown at 30 °C, shaking, for 4 h. For positive and negative controls, 1 mL aliquots of SC5314 cells were spun down, and the pellet was resuspended in SD with 100 µM anisomycin or 4 µg/mL fluconazole, and the cultures were incubated under shaking conditions at 30 °C for 10 min. One milliliter aliquots of each sample and drug-treated SC5314 cells were pelleted and resuspended in 1 mL of SD with l-HPG at 1:1,000 dilution. SC5314 cells resuspended in SD without HPG were used as the no-HPG control. All samples were incubated at 30 °C, shaking, for 30 min. Cells were spun down, supernatant removed, and pellets washed once with 1× PBS before being resuspended in 500 µL PBS. Cells were fixed by adding 500 µL 70% ethanol in 1× PBS and rocking for 1 h at room temperature. Cells were then pelleted and washed twice with 3% BSA in PBS. Pellets were resuspended with 500 µL of the reaction cocktail containing the azide fluorophore (prepared according to the manufacturer’s protocol) and incubated for 30 min, shaking in the dark. The samples were washed once with the rinse buffer, pelleted, and resuspended in 20 µL of 1× PBS. Cells were imaged by differential interference contrast microscopy and the EGFP channel on a Zeiss Axio Imager M1 at the same exposure time and analyzed using a CytoFlex flow cytometer (Backman Coulter) to quantify fluorescence intensity. A total of 20,000 events were documented and median fluorescein isothiocyanate value was determined using the CytExpert software.

Dose-response assays

Drug susceptibility was measured using a modified broth microdilution protocol in 96-well plates (Sarstedt) format, as previously described [92]. Drugs fluconazole, HU, NOC, and MMS were tested in liquid YPD. Cultures were grown in 2-fold dilutions of drug for 48 h as indicated. To assess cidality, 5 µL of mixed cultures were pipetted onto YPD agar following OD₆₀₀ measurements, and the agar plates were incubated at 30 °C for 2 days. Data were displayed quantitatively as heat maps using Java TreeView software (version 1.1.6r4). Reported results are representative of two independent experiments, each performed in technical duplicate.

Illumina whole-genome sequencing analysis

Genomic DNA was isolated using the phenol-chloroform extraction method as previously described [78]. Libraries were prepared using NEB Ultra II FS DNA Kit and sequenced with Illumina Novaseq X Plus PE150bp to obtain 50–1,000× coverage. Sequencing data were processed as previously described, with minor modifications [77,93]. For genomes with 1,000× coverage, quality control and trimming were performed using fastp (v0.23.4) [94] with minimum qualified phred quality of 20, a 40% limit on unqualified bases, a mean quality requirement of 20, a cut window size of 5, both front and tail cutting, and with correction enabled. For genomes with 50–100× coverage, adaptor sequences and low-quality reads were trimmed using Trimmomatic (v0.39 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:36 TORPHRED33) [95]. Trimmed reads were aligned to the C. albicans reference genome (SC5314_version_A21-s02-m09-r10) [28] using BWA-MEM (v.0.7.17) [96] with default parameters. Aligned reads were sorted, duplicate reads were marked, and the resulting BAM file was indexed with Samtools (v.1.17) [97] and picard (RRID:SCR_006525, v.2.26.3). The quality of trimmed FASTQ and BAM files was assessed with FastQC (v.0.12.1) [98] and Qualimap (v.2.2.1) [99]. Chromosome copy number changes were visualized using the Yeast Mapping Analysis Pipeline (YMAP v.1.0) [100]. Aligned BAM files were uploaded to YMAP and mapped to the reference genome C. albicans SC5314 (A21-S02-m09-r10) with GC-content bias correction for read-depth. Read depth along the x axis indicating chromosome position (Chr1–ChrR) was plotted as the log2 ration that was converted to 0–4 as the y axis indicating chromosome copy number [100]. Haplotypes are color-coded with gray, magenta, and cyan indicating heterozygous AB, homozygous B, and homozygous A, respectively.

Variant calling and filtering were performed for bypass mutants that failed to revert with Mutect2 and FilterMutectCalls within the GATK package (v.4.6.1.0) [101]. Pairwise variant calling was run for each bypass mutant and its parental strain with the bypass mutant as “tumor” and the parental strain as “normal” using the reference genome C. albicans SC5314 (A21-S02-m09-r10). The resulting VCF files were further filtered using bcftools (v.1.21) [97]. Specifically, the parental strain was removed from the VCF file and variant calls were filtered for a quality status of “PASS“ and exclusion of calls in repeat regions as marked in (SC5314 v.A21-s02-m09-r10_features.gff) and 5,000 bp subtelomeric regions. Hard filtering was performed with the following parameters: minimum of 5 supporting reads, at least one supporting read in both forward and reverse direction, minimum alternate allele frequency of 0.2. Identified variants were compared among bypass mutants and only unique calls were kept. Variants were annotated with SnpEff (v.4.3t; database built from SC5314 v.A21-s02-m09-r10, with alternate yeast nuclear codon table) [102], formatted using bcftools [97], and verified in IGV [103,104]. Verified variants were compiled into S3 Table.

Galleria mellonella model of C. albicans infection

A Galleria mellonella larvae model of infection was performed to test C. albicans temperature-dependent virulence as previously described [27]. Briefly, wild type (SC5314), deletion mutants, and complemented strains were grown overnight shaking in YPD at 30 °C. Cells were counted using a hemocytometer to prepare inoculum of 1 × 10⁸ cells per mL. Trypan blue was added to all inoculum stocks to a final concentration of 0.04% as a visual indicator of proper injection in the larvae. Larvae were screened using a weight filter of 230–300 mg and lack of dark spots indicative of injury and visible signs of pupation. Twenty larvae per group were injected in the last left proleg with 1 × 10⁶ cells in a 10 μL volume using a 500 μL gas-tight 1750 series syringe (Hamilton, Ref: CAL81230) with the PB600 repeating syringe dispenser attachment (Hamilton, 83700) for each strain. A group injected with PBS without cells were used as control. Larvae were incubated in petri dishes by treatment group at indicated temperatures and monitored for 144 h for mortality as defined by lack of response to physical touch. Mortality was first checked at 12 h postinfection, then every 4 h between 12 and 24 h postinfection, every 6 h between 24 and 60 h postinfection, and finally every 12 h thereafter until the endpoint of 144 h. Larvae and pupae were euthanized by freezing at −20 °C.

Cell growth and RNA preparation for ACT1 splicing assay and RNA sequencing

Overnight cultures of wild-type, tetO-YSF3/Δ, and tetO-PRP19/Δ strains in YPD in the absence and presence of 0.05 µg/mL DOX were sub-cultured to an OD₆₀₀ of 0.2 and 0.4 in 10 mL YPD in the absence and presence of 20 µg/mL DOX and grown for approximately 3 h at 22 or 37 °C. RNA was extracted, Dnase-treated, and reverse transcribed as described above with the iScript cDNA synthesis Kit (Biorad), which uses a blend of oligo (dT) and random primers to avoid 5′ and 3′ bias. ACT1 splicing assay was performed as previously described [49]. Briefly, 1 µL of cDNA product was mixed with 0.1 μL of each ACT1 primer at 100 μM (oLC9781 and oLC10309), 2.5 μL of the supplied 10× reaction buffer, 2 μL of the supplied 2.5 mM dNTP mix, 0.25 μL of TAKARA ExTaq DNA polymerase, and 19 μL of ddH₂O and subjected to a 30-cycle PCR reaction of 94 °C for 30 s, 52 °C for 45 s, and 72 °C for 60 s. PCR products were run on a 2% agarose gel at 100 V for 60 min, and the SYBR Safe-stained gels were imaged on a ChemiDoc.

Illumina Stranded mRNA Ligation Kit was used to prepare libraries for wild-type and tetO-YSF3/Δ RNAs. Libraries were sequenced with Illumina Novaseq X Plus PE150bp to obtain >50 million reads per sample.

RNA-Seq data processing and mapping

Following quality control by FastQC/0.11.9 [98] using the following command: fastqc -o QC_$OUTNAME $R1 $R2 # “$OUTNAME”_R1t.fastq.gz “$OUTNAME”_R2t.fastq.gz, detected Nextera adapter sequences were trimmed off of the paired-end 150 bp reads using cutadapt/2.10 [105] with the following flags and settings: --cores=0 -b CTGTCTCTTATACACATCT -B CTGTCTCTTATACACATCT -e 0.1 -O 10 -m 16 –output=”$OUTNAME”_R1t.fastq.gz –paired-output= ”$OUTNAME”_R2t.fastq.gz $R1 $R2. Another FastQC quality control was performed after trimming. HISAT2/2.2.1 [106] was used to map the trimmed reads to the Candida albicans SC5314 genome version A22 with the following specifications: hisat2 –summary-file “$OUTNAME”_A22_summary.txt –max-intronlen 2000 -x $GENOME2 –rna-strandness FR -1 $R1 -2 $R2 -S “$OUTNAME”_A22.sam –novel-splicesite-outfile “$OUTNAME”_novelSS.txt. Unmapped reads were removed, and the generated output of the mapped reads was converted into tdf and bedgraph files using SAMtools/1.16/1 [107], BEDtools/2.30.0 [108] and IGVtools [104] for IGV visualization and further analyses.

Expression analysis and principal component analysis

Read counts per gene for differential gene expression analysis were obtained using a custom bash script involving SAMtools/1.9, BEDTools/2.27.1 and a custom R script, as described in Lash and colleagues [49]. Specifically, the resulting bam files from mapping were split into first and second reads with SAMtools view -h -b -f 0x40 $BAM > r1.bam/ 0x80 $BAM > r2.bam. Following intersection with the genome annotation by bedtools intersect, reads overlapping with annotated features were counted using awk and bash commands. Resulting files contained reads per gene, as well as total mapped reads (including intergenic regions). Differential gene expression analysis was performed with custom R scripts using the DESeq2 package [109] from Bioconductor. DESeqDataSetFromMatrix was used to create the DESeqDataSet object to run DeSeq, comparing every condition with each other (S2 Data). Principal component analysis was performed by using vst and plotPCA functions. To create custom colors and shapes based on condition, ggplot2 [110] was used for better visualization of the PCA plot.

For the analysis towards Figs 3H and S6E, intron-containing gene expression values were adjusted to consider exclusively the CDS for gene expression and excluding reads covering introns that are expected to differ strongly in counts due to decreased pre-mRNA splicing upon YSF3 depletion. First, the reads per intron were calculated using the same approach as for obtaining reads per gene, except for using the intron annotation for bedtools intersect. The obtained intron counts, as well as the total gene counts were normalized for library size by dividing by the number of total mapped reads for each sample. Following normalization, the average counts were calculated between replicates of the same condition. Next, the normalized, averaged number of reads per intron was subtracted from the number of reads per (intron-containing) gene to obtain ‘reads per exon’, to get a measure for expression of transcripts that could actually be subject to translation. Combined exon length was calculated by subtracting the intron length from gene length. Then, reads per kilobase (RPK) values were calculated by dividing number of reads per exon by the exon length in kilobases. To obtain spliced and unspliced intron-containing gene expression, RPK values were multiplied by the respective IR fraction for that gene and condition (S5 Data). Genes with multiple introns were excluded from this analysis.

IR quantification

For splicing and IR quantification, the intron annotation used contained annotated introns from CGD, including 5′UTR introns. Analysis was carried out as in Lash and colleagues [49] using custom bash and R scripts. Specifically, bam files were split into forward and reverse reads again using samtools view -h -b -f 0x40 $BAM > r1.bam/0x80 $BAM > r2.bam. Afterwards, reads were overlapped with the annotated introns maintaining strand information using bedtools intersect. A minimum overlap of three bases was required to consider a read overlapping. The block count shows the number of continuous aligned segments in a read. Reads with a block count of 1, indicating alignment without gaps, were considered unspliced and were further categorized into intronic or unspliced, depending on the start and end position of the read; If the block count is > 1, indicating a gap in the aligned read and thus an exon-exon junction, it was classified as spliced. Split reads were further divided into spliced (exactly matching annotated intron) or alternatively spliced (no perfect match). For each intron, reads for each of the categories were counted. For each intron, IR fraction was calculated by dividing the number of unspliced reads by the sum of the number of unspliced and 2 × the number of spliced reads (S3 and S4 Data). We introduced factor 2 for “spliced” junction reads to account for the aspect that one unspliced transcript can give rise to two “unspliced” reads, but a spliced transcript can only yield one “spliced” read. A minimal read cutoff of 20 was required for further analysis. IR difference was calculated as the difference between the fraction of IR of each induced (+ DOX) and uninduced (− DOX) sample.

Branchpoint sequence identification

To computationally identify likely BPSs in C. albicans introns we made use of previously annotated 7-nt branchpoint motifs in S. cerevisiae and C. albicans [52,111]. First, we calculated the frequency, position weight, and logOdds matrix from the frequency of nucleotides at each position within the 7-nt motif and the average nucleotide frequency across S. cerevisiae introns using a custom R script and the Biostrings package. Using the S. cerevisiae LogOdds matrix we calculated a branchpoint score for each 7-nt running window across C. albicans introns by summing up LogOdds matrix entries at the respective motif position (column) and base identity (row). To select likely BPSs we required a score above 4. Three introns did not have a score above 4, filtering for a score greater than 0 yielded a likely branchpoint for all three introns. Overall, 333 introns showed the consensus BPS of 5′ UACUAAC 3′ and 76 introns showed slight variations of that motif (non-consensus BPSs) (S3 Data). Seventy-nine introns had more than one possible branchpoint, whereas we considered the BPSs matching the median distance between BPS and 3′ SS best as the primary branchpoint for our analysis. The median distance between BPS and 3′ SS is 24 nts and matches previous estimates [51]; the distance includes the 7-nt BPS. Introns were categorized by their BP motif (consensus/non-consensus) and into four ranges of BPS to 3′ SS distance using quantile to ensure equal distribution of the data.

Intron feature analysis

Intron lengths were obtained through the start and end intron coordinates. As intron lengths underlie a bimodal distribution (S6C Fig), we determined a cutoff between short and long introns by analyzing the distribution of intron lengths in our dataset using kernel density estimation (density() function in R). The density plot revealed two prominent peaks, corresponding to two major intron length modes. We used the findpeaks() function (pracma package) to identify these peaks and then determined the local minimum (valley) in the density curve between them. This valley was used as a threshold to classify introns into “short” (≤valley position) and “long” (>valley position). The first 6 nucleotides of an intron were considered for the 5′SS analysis and the last 3 intronic nucleotides to call 3′SS sequences. The conserved 5′SS motif consists of the sequence GUAUGU (GTATGT in the DNA sequ“nce),”and we defined the 3′SS consensus as the most conserved TAG motif, while also including CAG and AAG, which are common 3′SSs in C. albicans as well [51]. We found that all introns for which we measured IR by RNA-seq had a consensus 3′SS, with TAG being the most abundant comprising 298 introns (73%), AAG was found in 58 introns (14%) and 50 introns had a CAG 3′SS (12%). There was some variation within the 5′ SS, while most of them (336 introns) had the full consensus sequence, which is almost 83%, and 62 introns had only the strongly conserved first three nucleotides GTA of the motif, which in combination makes up almost all introns (398/98%). Thus, we categorized the SSs into consensus or variable 5′ SS (S4 Data). When plotting the Δ IR fraction (IR difference) for the categories intron length, BP motif, SS motif and BP–3′SS distance, we verified that the observed effects are consistent when all other variables are kept constant (optimal motifs/distance). Similar analysis for GC-content, polypyrimidine tract length and CU-fraction within the polypyrimidine tract did not produce significant differences in IR upon Ysf3 depletion.

GO analysis of differentially expressed genes

As in Lash and colleagues [49] GO term enrichment was performed using the combined GO annotations from CGD (https://geneontology.org/, version from 08/10/2023, gaf-version: 2; and version from 17/01/2024, gaf-version: 2.2) and GO annotations from S. cerevisiae orthologs (https://geneontology.org/, version from 18/01/2024, gaf-version: 2.2). S. cerevisiae orthologous genes were matched according to the information on CGD. GO term annotations were extended to all parents of individual GO terms using the “get_parent_nodes” function from the GOfuncR package. Occurrences of GO terms were counted in each gene group of interest (up- or down-regulated genes upon YSF3 depletion at 22 or 37 °C, respectively) and compared to their frequency in the gene background. For GO term enrichment among groups of differentially expressed genes we distinguished between upregulated genes (log2-fold change of differential expression > 0.5 and adjusted P-value < 0.00001, 592 genes at 22 °C and 239 genes at 37 °C) and downregulated genes (log2-fold change of differential expression >0.5 and adjusted P-value <0.00001, 616 genes at 22 °C and 342 genes at 37 °C). In this case, the gene background was formed by all genes that were quantified as expressed in the RNA-seq data. To assess the significance of GO term enrichment, we performed the Fisher’s Exact Test and multiple testing correction (Bonferroni correction). A P-value cutoff of <0.01 was used to call positively enriched GO terms for “Biological Process.” Enriched GO terms for “Biological Process,” “Molecular Function,” and “Cellular Component” across up- and down-regulated genes are listed in S6 Data.

Morphogenesis assay

To determine the impact of YSF3 on morphogenesis, CaSS1, YSF3, and EFG1 GRACE strains were grown in YPD in the presence and absence of 0.05 µg/mL DOX overnight at 30 °C, shaking, and sub-cultured to an OD₆₀₀ of 0.2 in YPD supplemented with 10% newborn calf serum in the presence and absence of 20 µg/mL DOX at 37 °C, shaking, for 6 h. Cells were then imaged on a Zeiss Axio Imager MI.

To determine the impact of RHT1 on morphogenesis, overnight cultures of SC5314, CaLC10720, and CaLC10721 were grown in YPD at 30 °C, shaking, and sub-cultured to an OD₆₀₀ of 0.2 in YPD. SC5314 cells treated with 25 mM HU and 50 µM NOC were used as controls. Cultures were grown at 30 °C for 4 h before imaging on a Zeiss Axio Imager MI.

Immunoprecipitation

Immunoprecipitation of GFP-tagged Rht1 was performed to assess RHT1-GFP expression, as previously described [112]. Briefly, overnight cultures were sub-cultured to an OD₆₀₀ of 0.2 in 50 mL YPD and grown for 4 h at 30 °C. Cells were pelleted and washed in 1 mL of F300 buffer (50 mM HEPES-KOH pH 7.2, 300 mM KOAc, and 10% glycerol) and stored at −80 °C. To prepare whole cell lysate, each pellet was resuspended in 1 mL F300 and disrupted by bead-beating (2 min on; 1 min rest on ice) for three cycles. Crude lysates were mixed with a half-volume of F300 supplemented with 0.06% IGPAL-CA630 and sonicated (20 s  × 4) at 30% amplitude using a 3-mm diameter probe. Sonicated lysates were treated with ~30 U Benzonase (Millipore) for 20 min after the addition of MgCl₂ to the final centration of 2 mM and clarified by centrifugation at 20,000 g for 20 min at 4 °C. Total protein (~6 mg) in supernatant was incubated with 20 µL of anti-GFP magnetic beads (GTMA-20; Chromotek) for 90 min on a revolving rotator. Bead-bound proteins were captured using a magnetic rack and washed with F300 + 0.02% IGEPAL-CA630 × 3 before boiling in 30 µL 1 × SDS-PAGE loading buffer. All eluted proteins were resolved on 10% precast SDS-gels and blots probed with HRP-conjugated anti-GFP antibody (118144600012; Roche; 1:3,000) followed by HRP-conjugated goat anti-mouse secondary antibody (170-65169; Bio-Rad; 1:5,000). All F300 buffers were supplemented immediately prior to use with 1× PhosSTOP (Roche) and 1× Complete inhibitor cocktails (Roche) to inhibit phosphatase and proteinase activity.

Mass spectrometry

AP-MS experiments were performed as previously described [27]. Briefly, Rht1-GFP (CaLC10748) and Eno1-GFP (CaLC4449) strains were grown overnight at 30 °C in YPD. Stationary phase cultures were diluted to an OD₆₀₀ of 0.1 in 1 L YPD and grown to an OD₆₀₀ between 0.6 and 0.8. Cells were harvested at 3,000 rpm for 30 min at 4 °C, washed with ice-cold water, and snap-frozen in liquid nitrogen bath. For protein extraction, samples were diluted 1:1 by weight in lysis buffer (50 mM HEPES [pH 7.5], 150 mM NaCl, 5mM EDTA, 5mM DTT, 0.1% NP-40, 1× ROCHE protease inhibitor cocktail tablet [Roche Diagnostics, Mississauga, ON, Canada]) and vortexed with glass beads (0.5 mm) for 4 × 1 min. Lysates were collected by stacked transfer for 5 min at 100g with a 18-gauge needle and clarified by centrifugation at 16,110g for 20 min in 4 °C in a microcentrifuge.

To affinity purify the GFP-interacting proteins, we used 25 µL of GFP-TRAP_M magnetic beads (ChromoTek, Martinsried, Germany) for each 1 L culture. Beads were equilibrated three times with 1 mL of lysis buffer (50 mM HEPES [pH 7.5], 150 mM NaCl, 5 mM EDTA, 5 mM DTT, 0.1% NP-40, 1× ROCHE protease inhibitors [Roche Diagnostics, Mississauga, ON, Canada]). The protein extract was then added to the resin and rotated for 2 h at 4 °C. The beads were magnetized then washed with 1 mL of lysis buffer and then 1 mL of wash buffer (20 mM Tris [pH 8.0], 2 mM CaCl₂). The dried beads were resuspended in 20 mM Tris-HCl, and incubated with 0.5 μg trypsin (Sigma, T7575) with rotation at 37 °C overnight (~15 h). The samples were digested on-bead with 0.75 μg trypsin (0.2 μg/μL in 20 mM Tris-HCl, pH 8.0) at 37 °C overnight with rotation. Magnetization and digestion were repeated once more with a shorter incubation of 3–4 h without rotation. Formic acid was added to a final concentration of 2% and stored at −40 °C until used for MS.

AP-MS samples and controls were analyzed by MS in two biological replicates. 1/16th of each sample was acquired in data-dependent acquisition (DDA) via LC–MS/MS. Samples were loaded onto Evotip Pure cartridges and eluted from the Performance EV1109 column (8 cm × 150 µm with 1.5 µm beads), heated at 40 °C, with the 60SPD gradient on an Evosep One. The column was coupled to the timsTOF Pro 2 using a 20 µm diameter emitter tip (Bruker 1811107). Peptides were eluted from the column with an acetonitrile gradient generated by an Eksigent ekspert nanoLC 425 and analyzed on a TripleTOF 6600 instrument (AB SCIEX, Concord, Ontario, Canada). The total DDA protocol is 22 min. The MS1 scan had a mass range of 100–1,700 Da in PASEF mode. TIMS settings were accumulation and ramp time of 100 ms (with 4 PASEF ramps and active exclusion at 0.4 min), and within the mobility range (1/K0) of 0.85–1.3 V·s/cm². This was at a cycle time of 0.53 s. The target intensity was set to 17,500 and the intensity threshold set to 1,750. 1+ ions are excluded from fragmentation using a polygonal filter. The auto calibration was off.

MS data generated were stored, searched, and analyzed using the ProHits laboratory information management system (LIMS) platform and searched using MSFragger v4.1 [113] against the C. albicans UniProt database (UP000000559) acquired from NCBI, supplemented with common contaminants, decoy entries were added. Acetylated protein N-term and oxidated methionine were set as variable modifications. Precursor mass tolerance was set to 20 ppm on either side. Fragment mass tolerance was set to 20 ppm. Enzymatic cleavage was set to trypsin with 2 missed cleavages. MSBooster and Percolator were turned on. Percolator required a minimum probability of 0.5 and did not remove redundant peptides. The target-decoy competition method was used to assign q-values and PEPs. For ProteinProphet, the maximum peptide mass difference was set to 30 ppm. When generating the final report, the protein FDR filter was set to 0.01. FDR was estimated by using both filtered PSM and protein lists. Razor peptides were used for protein FDR scoring. All other parameters were default. Results were analyzed with SAINTexpress (v3.3) [114] with Eno1-GFP samples as negative controls. A total spectral count greater than or equal to 20 and a BFDR value equal to 0 were required for proteins to be classified as significant interaction partners (S7 Data). Data has been deposited as a complete submission to the MassIVE repository (https://massive.ucsd.edu/ProteoSAFe/static/massive.jsp) and assigned the accession number MSV000097442. The ProteomeXchange accession is PXD062215. The dataset is currently available for reviewers at ftp://MSV000097442@massive.ucsd.edu. Please login with username MSV000097442_reviewer; password: k5Zd8dQmZ8qZz70E. The dataset will be made public upon acceptance of the manuscript. Cytoscape [115] was used to visualize high-confidence protein interactors identified by AP-MS.

E-test strip and disc diffusion assays

E-test strip assay and disc diffusion assays were performed to test fluconazole tolerance for C. albicans strains, as previously described [116]. Wild type, rht1Δ/Δ mutant, complemented strain, ByM1, ByM1R, and ByM10s were grown in liquid YPD at 30 °C overnight, washed once with sterile H₂O, and diluted to a cell density of 10⁸ cells/mL. About 100 µL cells were spread on YPD agar using a sterile cotton swab. Fluconazole E-test strip (bioMérieux) or a 0.5 cm precut filter disc was placed at the center of the YPD agar plates. A volume of 5 µL of 10 mg/mL fluconazole stock was spotted on the filter disc. Plates were incubated upside down at 30 or 35 °C and imaged on a ChemiDoc after 3 days.