
==== Front
101573691
39703
Cell Rep
Cell Rep
Cell reports
2211-1247

39126650
10.1016/j.celrep.2024.114601
nihpa2019543
Article
Functional genomic analysis of genes important for Candida albicans fitness in diverse environmental conditions
Xiong Emily H. 16
Zhang Xiang 26
Yan Huijuan 3
Ward Henry N. 24
Lin Zhen-Yuan 5
Wong Cassandra J. 5
Fu Ci 1
Gingras Anne-Claude 15
Noble Suzanne M. 3
Robbins Nicole 1
Myers Chad L. 24*
Cowen Leah E. 17*
1 Department of Molecular Genetics, University of Toronto, Toronto, ON M5S 1A8, Canada
2 Department of Computer Science and Engineering, University of Minnesota, Minneapolis, MN 55455, USA
3 Department of Microbiology and Immunology, University of California San Francisco, San Francisco, CA 94143, USA
4 Bioinformatics and Computational Biology Graduate Program, University of Minnesota, Minneapolis, MN 55455, USA
5 Lunenfeld-Tanenbaum Research Institute, Mount Sinai Hospital, Sinai Health, Toronto, ON M5G 1X5, Canada
6 These authors contributed equally
7 Lead contact
AUTHOR CONTRIBUTIONS

Conceptualization, N.R., X.Z., C.L.M., and L.E.C.; methodology, E.H.X. and X.Z.; software, X.Z. and H.N.W.; validation, E.H.X., X.Z., and C.J.W.; formal analysis, E.H.X., X.Z., C.J.W., and H.Y.; investigation, E.H.X., H.Y., and Z.-Y.L.; resources, L.E.C., C.L.M., A.-C.G., S.M.N.; data curation, E.H.X., X.Z., and C.J.W.; writing – original draft, E.H.X., X.Z., and N.R.; writing – review & editing, all authors; visualization, E.H.X., X.Z., and H.Y.; supervision, C.F., C.L.M., N.R., and L.E.C.; project administration, N.R.; funding acquisition, L.E.C., C.L.M., A.-C.G., and S.M.N.

* Correspondence: chadm@umn.edu (C.L.M.), leah.cowen@utoronto.ca (L.E.C.)
9 9 2024
27 8 2024
08 8 2024
22 9 2024
43 8 114601114601
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/).
SUMMARY

Fungal pathogens such as Candida albicans pose a significant threat to human health with limited treatment options available. One strategy to expand the therapeutic target space is to identify genes important for pathogen growth in host-relevant environments. Here, we leverage a pooled functional genomic screening strategy to identify genes important for fitness of C. albicans in diverse conditions. We identify an essential gene with no known Saccharomyces cerevisiae homolog, C1_09670C, and demonstrate that it encodes subunit 3 of replication factor A (Rfa3). Furthermore, we apply computational analyses to identify functionally coherent gene clusters and predict gene function. Through this approach, we predict the cell-cycle-associated function of C3_06880W, a previously uncharacterized gene required for fitness specifically at elevated temperatures, and follow-up assays confirm that C3_06880W encodes Iml3, a component of the C. albicans kinetochore with roles in virulence in vivo. Overall, this work reveals insights into the vulnerabilities of C. albicans.

In brief

Xiong et al. perform a functional genomics screen to identify genes important for fitness in diverse environments in the fungal pathogen Candida albicans. Through their investigations, they characterize C1_09670C as subunit 3 of replication factor A (Rfa3) and C3_08880W as kinetochore component Iml3 with important roles in C. albicans virulence.

Graphical Abstract
==== Body
pmcINTRODUCTION

The fungal kingdom encompasses more than five million species that inhabit incredibly diverse ecological niches.1,2 For the few hundred fungal species capable of infecting a human host, the ability to sense and adapt to a myriad of environmental stressors is critical for survival and pathogenesis. One of the leading causes of fungal infections in humans is Candida albicans, an opportunistic pathogen that causes both superficial infections in immunocompetent individuals and life-threatening systemic disease in those with compromised immune systems.3,4 Invasive C. albicans infections are associated with mortality rates of approximately 40% despite treatment, a statistic that has remained unchanged since 1997.5–7 This is largely due to the fact that there are only three primary classes of antifungals for the treatment of invasive fungal infections: azoles, polyenes, and echinocandins, which each target functions essential for viability.8 Hence, the threat to human health posed by C. albicans demands a significant response in research efforts to expand the current knowledge about this pathogen.

C. albicans is a successful human pathogen at least in part due to its ability to survive and adapt to a diverse range of stresses encountered within the human host. Unlike the majority of species in the fungal kingdom, C. albicans can thrive at normal body temperature (37°C) as well as tolerate elevated temperatures that are encountered during febrile episodes.9,10 The metabolic flexibility of C. albicans also contributes to its success as a pathogen, as it can utilize alternative metabolic pathways in nutrient- and oxygen-limited host niches, such as within phagocytes or in the microbe-rich gastrointestinal tract.11 Furthermore, as a leading cause of bloodstream infections, C. albicans is able to survive in the presence of serum, a key component of human blood that has several antimicrobial properties, including low levels of essential nutrients and elements of the complement system that play an important role in the host innate immune response.9,12,13 However, despite the ability of this organism to grow in diverse environments, a substantial portion of the current literature examining various facets of C. albicans biology is based on studies performed under standard laboratory conditions (rich medium at 30°C), which has limited relevance to the diverse stresses in host environments.

Most antimicrobials in clinical use target functions essential for pathogen viability, and understanding the function of candidate drug targets is key to developing effective therapeutics. We recently leveraged an expanded C. albicans functional genomics resource termed the Gene Replacement and Conditional Expression (GRACE) collection to perform an arrayed phenotypic assessment of genes important for fitness under standard laboratory conditions, revealing 634 genes crucial for growth.14 While this study characterized several previously uncharacterized essential genes, including ones involved in kinetochore function (KRP1), mitochondrial integrity (EMF1), and translation initiation (TIF33), it was limited to genes required for fitness under standard laboratory conditions. It has been well documented in the model yeast Saccharomyces cerevisiae that the eukaryotic genome is highly buffered, with a significant number of genes only important for growth under particular environmental conditions.15,16 Most recently, comprehensive analysis of over 4,000 S. cerevisiae double deletion mutants in 14 growth conditions found that, while genetic interactions are generally robust to environmental perturbation, more than half of the double mutants surveyed showed significantly different fitness defects upon exposure to different environmental stresses.17 While the impact of environmental conditions on genes important for fitness has not been studied systematically in fungal pathogens, several C. albicans genes that are dispensable in standard laboratory conditions have been implicated in fitness in specific environmental conditions. Examples include genes encoding the Med7 subunit of the Mediator protein complex, which is only essential in media with alternative carbon sources18; the protein phosphatase calcineurin, which is required for C. albicans survival in serum19; and the transcription factor Tye7, which is specifically important for growth in hypoxic conditions.20,21

Here, we utilized a high-throughput pooled screen of a large-scale C. albicans conditional expression mutant library to identify genes important for fitness in diverse environmental conditions. The GRACE library includes a collection of 2,287 barcoded heterozygous deletion mutants, where expression of the remaining wild-type allele is regulated by a doxycycline (DOX)-repressible promoter.14,22 We leveraged a barcode sequencing approach to enable facile screening of strain abundance under different growth conditions. Screening of GRACE pooled mutants in eight distinct environmental conditions identified between 242 and 313 genes important for fitness in each condition, of which 171 genes were identified in all conditions (condition independent). Interestingly, dozens of genes with a variable impact on fitness between conditions (condition dependent) were identified, expanding the repertoire of C. albicans genes important for fitness in host-relevant conditions in an environmentally contingent manner. We prioritized two previously uncharacterized genes with either condition-independent or temperature-dependent importance for fitness and described their functions in DNA replication and repair (RFA3 or C1_09670C) and the inner kinetochore (IML3 or C3_06880W), respectively. Overall, this study employs a powerful functional genomic strategy to uncover genes with condition-dependent impacts on C. albicans fitness and yields insights into previously unexplored biology of this human fungal pathogen.

RESULTS

A high-throughput pooled screening pipeline identifies genes important for C. albicans fitness in diverse environmental conditions

Although strains in the GRACE collection have been evaluated previously for fitness under standard laboratory conditions,14,23 expansion of screening to include additional environments has been limited by the time and reagents needed to perform arrayed screens of this extensive collection. To enable systematic fitness comparisons between diverse growth conditions, we utilized a pooled functional genomics approach for massively parallel analysis of GRACE strains to assess each strain’s fitness profile through quantification of strain-specific barcodes by next-generation sequencing. Specifically, a pooled collection of 2,238 barcoded GRACE strains covering ~35% of the genome was grown in triplicate in the presence and absence of DOX for 24 h and then sub-cultured into matching fresh medium with or without DOX for an additional 48-h incubation. We screened the barcoded GRACE collection in the baseline condition of synthetic defined yeast nitrogen base (YNB) medium at 30°C, as well as six additional growth conditions at 30°C: nutrient-rich medium (yeast extract peptone dextrose [YPD]); YNB containing the osmotic stressor sodium chloride; YNB supplemented with serum, a key component in human blood; YNB supplemented with the membrane stressor SDS; YNB containing the osmotic stressor sorbitol; and iron-free YNB. We also screened in YNB at the elevated temperature of 37°C (Table S1). Screens in all conditions were performed in technical triplicate and biological duplicate apart from the baseline condition, which was tested in five biological replicates to assess the robustness of this approach (Figure S1A).

After growth of the cultures, strain-specific molecular barcodes were amplified from extracted gDNA, and pooled barcode PCR products were subjected to next-generation sequencing (Figures 1A and S2A). With data normalization (STAR Methods), a log2 fold change (log2FC) score was calculated from the average “no DOX”/DOX ratio of barcode reads for each strain from the technical triplicates, followed by moderated t tests to determine whether the score was significantly different from the median log2FC value of all strains from that specific condition (Figure 1B; Table S1). Genes represented by strains that met chosen thresholds for effect size (log2FC≧2) and statistical significance after multiple hypothesis correction (false discovery rate FDR ≦ 0.05) were deemed important for C. albicans fitness. Our screens captured sufficient barcode reads (>50 raw reads in “no DOX” samples) across all conditions tested for 2,168 strains or 96.9% of the 2,238 barcoded strains within the GRACE library. Importantly, a high correlation in log2FC scores was observed between technical and biological replicates (Figures S1B–S1D). This pipeline identified between 242 and 313 genes significant for fitness in each condition (Figure 1C), unveiling a comprehensive set of genes required for fitness in diverse environments.

To define the genes that were either universally important for fitness under all tested conditions (condition independent) or selectively important for growth under specific environmental conditions (condition dependent), we performed comparative analyses across our screens and identified 171 condition-independent hits and dozens of genes with condition-dependent requirements for C. albicans fitness (Figure 2A). Of the 171 condition-independent hits, 137 were previously annotated as essential14 when the GRACE library was grown in an arrayed format on YNB agar (Figure S1E). Transcriptional repression of 25 of the remaining genes had been annotated previously as conferring a growth defect, consistent with our pooled screen assessing fitness on a continuous scale. We also observed a positive linear correlation when comparing the log2 from the pooled screens with essentiality scores from the arrayed screen, with a Pearson correlation coefficient (R) of 0.88 (Figure S1F). Outside of condition-independent hits, 39 mutants displayed a significant fitness defect in YPD only, whereas 36 mutants showed a fitness defect in all conditions except rich medium, highlighting that a significant contributor to the environment-specific effects observed was the growth medium. Other interesting observations included the identification of 18 mutants that showed significant fitness defects at 37°C and seven mutants with fitness defects in the presence of serum. Thus, our pooled screening pipeline provides a powerful platform to identify C. albicans genes important for growth while enabling extensive multiplexing to increase the robustness of the results.

To explore condition-independent and -dependent gene functions, we performed Gene Ontology (GO) enrichment analysis of biological processes on groups of condition-independent and -dependent genes (Table S2). Analysis of condition-independent genes revealed biological processes known to be fundamental for growth, such as progression of the cell cycle, translation, transcription, maintenance of cell shape, and chromosome segregation (Figure 2B). Additionally, GO enrichment was explored for smaller subsets of genes identified to be important for only one or a unique group of conditions (Figures 2C–2E). Analysis of genes important for fitness in all conditions except YPD revealed genes associated with amino acid and nutrient metabolism, which is consistent with the differences in nutrition between growth media (Figure 2C). In line with previous literature in both fission and budding yeast that reported differential regulation of respiration, metabolism, and signaling in response to nutrient levels,24–26 we observed enrichment of regulation of the cell cycle, phosphorylation, and protein modifications in more nutrient-rich growth conditions, such as YPD and YNB supplemented with serum (Figure 2E). Excitingly, GO analysis of the 18 genes exclusively important for fitness at the physiological temperature of 37°C identified enrichment of processes involved in interaction with the host (Figure 2D). Collectively, these serve as a proof of concept that our pooled screen has sufficient sensitivity to identify differences in the genes required for fitness between different growth conditions.

C1_09670C is required for fitness in all conditions and encodes Rfa3

While most of the condition-independent essential genes identified from our screen are characterized or assigned functional annotations based on homologs in model yeasts (S. cerevisiae or Schizosaccharomyces pombe), we identified one gene, C1_09670C, for which transcriptional repression impaired fitness across all conditions tested and for which an S. cerevisiae homolog remained unknown (Table S1). To validate the importance of C1_09670C for growth, we performed monotypic growth assays with the tetO-C1_09670C/C1_09670cΔ strain in the absence and presence of DOX in all conditions used for the primary screen. Consistent with the log2 scores from our pooled screen, we observed a significant decrease in growth upon transcriptional repression of C1_09670C in liquid growth assays relative to wild-type or “no DOX” controls; the growth defects were comparable to what was observed upon repression of HSP90, a well-characterized essential gene in C. albicans (Figures 3A and S2B).

Given the lack of an identifiable homolog in S. cerevisiae, we attempted to identify C1_09670C homologs in a broader range of species beyond the model yeast by Basic Local Alignment Search Tool (BLASTp).27 Homologs in more distantly related fungi, such as Cryptococcus neoformans, were identified to encode for Rfa3, a heterotrimeric single-stranded DNA binding protein involved in DNA replication, repair, and recombination.28,29 Conserved protein domain searches using Pfam30 and UniProt31 supported BLASTp results by identifying homology to the RFA3 domain family. Similarities in structure were observed when comparing the AlphaFold-predicted32 protein structures of C1_09670c with Rfa3 in S. pombe, C. neoformans, and S. cerevisiae (Figure 3B). Interestingly, C1_09670c is unique in its lack of significant similarity to Rfa3 in S. cerevisiae with no BLASTp similarity and an European Molecular Biology Open Software Suite (EMBOSS) Needle sequence similarity of 40.6%.33 The two other subunits of RFA in C. albicans share significant homology with their S. cerevisiae counterpart, Rfa1 (E = 7 × 10−175, % sequence similarity = 61.4%) and Rfa2 (E = 1 × 10−40, % sequence similarity = 52.2%) and have been annotated based on their sequence homology (Figures S3A–S3C).

Given the predicted role of C1_09670C as RFA3, a gene that encodes a protein involved in DNA replication,29 we examined the impact of transcriptional repression of C1_09670C (referred to as RFA3 hereafter) on C. albicans cell morphology. C. albicans polarized growth is induced upon perturbation of the cell cycle, and impeding DNA replication results in cell cycle arrest.34,35 Wild-type or tetO-RFA3/rfa3Δ strains were left untreated, treated with a low concentration (0.5 μg/mL) of DOX, or treated with the microtubule-disrupting agent nocodazole, as indicated. We observed that depletion of RFA3 resulted in formation of pseudohyphae, similar to that of wild-type cells treated with nocodazole (Figure 3C). We also monitored the cellular localization of a strain expressing homozygous C-terminal epitope-tagged Rfa3-GFP using fluorescence microscopy. In line with Rfa3 in S. cerevisiae,28,36 we observed localization of Rfa3 to the nucleus and occasional assembly into a single focus per nucleus (Figure 3D).

Given that Rfa3 has a role in DNA repair, we assessed the sensitivity of the tetO-RFA3/rfa3Δ mutant strain to genotoxic stress using spot dilution assays on medium containing the DNA damaging agent methyl methanesulfonate (MMS). A HOF1 mutant known to be hypersensitive to MMS was tested in parallel as a control.37 In the absence of DOX, growth of the tetO-HOF1/hof1Δ and tetO-RFA3/rfa3Δ strains was comparable to the wild type in both the presence and absence of 0.01% MMS. However, depletion of HOF1 and RFA3 with DOX resulted in hypersensitivity to 0.01% MMS (Figures 3E and S4A–S4C). Thus, transcriptional repression of RFA3 results in hypersensitivity to genotoxic stress, consistent with a role in DNA damage repair. We then compared the localization of GFP-tagged Rfa3 in response to nucleotide damage induced by MMS, as the RFA heterotrimer has been shown to rapidly associate with single-stranded DNA to initiate downstream checkpoint and repair mechanisms.36,38,39 Samples were left untreated or treated with 0.01% MMS for 3 h prior to imaging and quantification of cells displaying multiple foci, a single focus, and no foci per nucleus (Figure 3F). We observed instances of all three phenotypes in untreated cells, indicating basal levels of DNA damage. However, upon MMS treatment, we observed a marked increase of cells with multiple foci per nucleus, consistent with the recruitment of the RFA complex to the nucleus in response to DNA damage.

To provide further evidence that C. albicans Rfa3 participates in the RFA complex, we performed affinity purification coupled with mass spectrometry (AP-MS) of our strain expressing homozygous GFP-tagged Rfa3 alongside a strain expressing an unrelated GFP-tagged Eno1 as a control. Tagging both alleles of RFA3 confirmed functionality, as the strain was viable. We identified 130 high-confidence protein interactors (Bayesian FDR [BFDR] threshold of ≤ 0.01 and total spectral counts ≥ 5) (Figure 3G; Table S4). Among the interactors, we identified a large cluster of proteins involved in DNA replication and repair, including both remaining members of the RFA heterotrimer, Rfa1 and Rfa2, as well as all five members of the heteropentameric replication factor C (RFC), Rfc1–Rfc5. Overall, our data support that C1_09670C encodes a functional ortholog of Rfa3 with conserved roles in DNA replication and repair despite divergence in sequence.

Analysis of mutant sensitivity profiles reveals functionally coherent clusters

While our characterization of Rfa3 highlighted the power of functional genomic screens to identify previously uncharacterized essential genes in C. albicans, an additional advantage of our pooled screening pipeline is to reveal condition-specific essential genes and advance our understanding of C. albicans biology. We reasoned that the profile of quantitative phenotypes for each mutant across multiple screened conditions could provide the basis for the systematic discovery of gene function. Such profile similarity analysis of mutant phenotypes across different environmental conditions or genetic backgrounds has been applied successfully to S. cerevisiae chemical genomic and functional genomic data.16,40,41 To test the potential of our datasets for capturing functionally related genes, we applied FLEX (functional evaluation of experimental perturbations), a computational tool for measuring the functional coherence of the genetic screen data,42 and we focused on a subset of 507 mutants that exhibited a significant growth defect (log2FC≧1) in at least one of the YNB-based conditions examined. The primary goal of this analysis was to test whether profile similarity measured on our growth defect profiles could non-randomly associate functionally related gene pairs, which FLEX measures by producing a precision-recall curve (Figure 4A). Specifically, a Pearson correlation coefficient is computed between each gene pair to measure profile similarity, which is then tested for its ability to discriminate functionally related gene pairs from unrelated pairs based on annotations to specific GO terms. An uninformative similarity measure would produce a straight line centered at the background fraction of functionally related pairs to all gene pairs (as shown in Figure 4A). Indeed, we found enrichment for functionally related pairs among mutant pairs exhibiting high fitness profile correlations, with an ~3-fold enrichment for co-function among the most highly correlated pairs compared to the random baseline (Figure 4A). This enrichment was supported by genes belonging to a broad range of different biological processes, including strong representation from genes involved in amino acid and mRNA metabolism, cell cycle regulation, chromatin remodeling, and cytoskeleton organization (Figure 4B). These analyses confirm that these fitness profiles can associate genes with related functions.

Given the evidence that profile similarity can be used to identify functionally related genes, we performed a matrix clustering involving genes associated with at least one growth defect across all conditions but YPD (Figure 4C), applying a modified hierarchical clustering algorithm (see STAR Methods for details). Consistent with our FLEX analysis, several of the gene clusters identified were enriched for specific GO terms given an FDR cutoff of 0.05 (Table S3) and the specific pattern of fitness defects driving each cluster. This approach can not only implicate characterized genes in previously unreported biological processes but also generate functional predictions for previously uncharacterized genes. In our dataset, we observed an instance of the former in the identification of an E3 ubiquitin ligase (HRT1) within a cluster enriched for “endocytosis” (Figure 4C, inset #2). The role of this gene in endosome maturation has not yet been explored in C. albicans but has been documented for its homolog in S. cerevisiae.43 We also identified C2_04370W, a gene our group recently characterized as a translation initiation factor subunit (TIF33),14 within a cluster enriched for “translation initiation” (Figure 4C, inset #3). Interestingly, we identified a completely uncharacterized gene (C3_06880W) within a cluster of 15 genes grouped based on their shared importance for fitness in the high-temperature condition that was enriched for the GO terms of “mitotic cell process” and “filamentous growth of a population of unicellular organisms” (Figure 4C, inset #1). The distinct fitness pattern of C3_06880W across our screening conditions and consequent co-clustering with genes of known functions in the cell cycle, cytoskeleton organization, or filamentation-associated processes suggests that it may have a similar function. Thus, this analysis supports the utility of fitness similarity profiling to predict and uncover biology in C. albicans.

C3_06880W encodes kinetochore component Iml3, required for fitness in elevated temperatures

To further confirm that profile similarity analysis is a powerful tool that can predict gene function in C. albicans, we focused on the aforementioned gene C3_06880W (Figure 4C, inset #1). C3_06880W is an uncharacterized gene that was among those with the highest differences in log2 scores (differential log2 [dLFC] = 4.46, FDR = 4.09 × 107) when comparing fitness at physiological temperature (37°C) to the baseline condition (30°C) (Figure 5A). This gene was of particular interest, as it was completely uncharacterized, and common methods of annotating cellular localization or function based on signal peptides, coexpression data, conserved protein domains, and homologs in other fungal species failed to yield any promising clues. We confirmed that C3_06880W was important for fitness in a temperature-dependent manner, as transcriptional repression of C3_06880W with DOX resulted in significant growth defects only at 37°C or 42°C compared to wild-type or “no DOX” controls (Figures 5B and S4D). This was in contrast to the condition-independent essential gene HSP90, for which transcriptional repression led to significant growth defects at all temperatures (Figures 5B and S4C).

To confirm the prediction generated by our modified hierarchical clustering analysis that C3_06880W was involved in a mitotic cell-cycle-associated process, we used a predicted structure of C3_06880w generated by AlphaFold32 to query all 3D protein structures deposited in the PDB using PDBsum.44 Significant similarities were identified to the crystal structure of Iml3 in S. cerevisiae (Z score = 123.2).45 This prediction was further supported by favorable template modeling (TM) scores calculated by both TM-align46 and pairwise structure alignments47 of C3_06880w with S. cerevisiae Iml3 (TM = 0.62186) (Figure 5C), where a TM score greater than 0.5 suggests significant structural similarity of the protein pair.46 In S. cerevisiae, Iml3 is an inner kinetochore protein that forms a stable heterodimer with Chl4 to govern chromosome segregation.45 Encouragingly, we identified C. albicans CHL4 among the top five genes with temperature-specific importance for fitness (dLFC = 3.65; FDR = 7.80 × 105) (Figure 5A), and CHL4 was also a member of the cluster enriched for genes associated with the mitotic cell cycle in our hierarchical clustering analysis (Figure 4C, inset #1).

To test whether C3_06880w performs a function similar to ScIml3 in the kinetochore, we used microscopy to assess filamentation in the absence of an inducing cue, as blocking kinetochore function prevents G2/M progression in the cell cycle and induces pseudohyphal morphology.34,35,48 While wild-type C. albicans and “no DOX” controls all grew as yeast in the absence of an inducing cue, transcriptional repression of the known kinetochore components MTW1 or CHL4 as well as C3_06880W (IML3) induced pseudohyphae, similar to what was observed when wild-type cells were treated with the M phase-disrupting agent nocodazole (Figure 5D). We then assessed the localization of Iml3 using fluorescence microscopy by constructing a strain expressing a homozygous C-terminal epitope-tagged Iml3-GFP with C-terminal epitope-tagged Mtw1-RFP. Functionality of the tagged proteins was confirmed by assessing growth at 37°C, where the strain displayed a mild defect after 24 h and growth comparable to the wild type after 48 h (Fig. S5A–B). Using fluorescence microscopy, we observed co-localization of Mtw1 (red) and Iml3 (green) puncta at the nuclear periphery, consistent with our model where Iml3 functions at the kinetochore (Figure 5E).

As a final assay to characterize the function of C. albicans Iml3, we sought to identify its physical binding partners. To do so, we performed affinity purification of Iml3-GFP using a GFP-trap resin followed by mass spectrometry. The significance analysis of interactome (SAINT) analysis tool49 was used to discriminate high-confidence interactors from the background by comparing results to an Eno1-GFP control. Using a stringent BFDR threshold of ≤0.01 and total spectral counts ≥5, we identified 65 high-confidence protein interactors for Iml3 (Table S4). Of the 10 known protein interactors of Iml3 in S. cerevisiae,50 only three homologs—Chl4, Mcm21, and Ctf3—have been annotated in C. albicans. Encouragingly, all three were identified with high confidence by AP-MS (Figures 5F and 5G). Our AP-MS experiment also yielded eight uncharacterized proteins, with two having predicted kinetochore localization (C3_04690c and C1_08150c) and the remaining six lacking functional annotations or predictions (Figure 5F). We employed similar techniques as used to predict the function of both Rfa3 and Iml3 by looking for homologs in more distantly related fungal species and querying structural databases for similarities in protein structure. Starting with proteins predicted to localize to the kinetochore, we identified conserved protein domains and characterized homologs in other fungal species, such as A. flavus and S. pombe, suggesting that C3_04690c and C1_08150c are homologs of Nkp1 and Nkp2, respectively. In addition, one of the uncharacterized proteins, C2_02660w, contained a conserved domain corresponding to Okp1, a subunit of the Ctf19/Okp1/Mcm21/Ame1 (COMA) complex within the S. cerevisiae kinetochore. We also observed significant structural similarities between C2_03890w and Ctf19, another subunit of the COMA complex. Further experimental characterization is needed to confidently annotate the functions of these uncharacterized proteins.

Next, we sought to evaluate the temperature-dependent role of Iml3 in C. albicans fitness in vivo. The invertebrate model of the greater wax moth Galleria mellonella has been used successfully to study the pathogenesis of many fungal species,51–56 with previous studies noting excellent correlation of the virulence of C. albicans when comparing insect and mouse models of infection.57 G. mellonella has also proven to be applicable for exploring temperature-dependent phenotypes in capsule formation and virulence in C. neoformans,58 as the larvae can be maintained at temperatures ranging from 25°C to 39°C.54 To determine the role of Iml3 in C. albicans fitness in this model, we infected larvae by proleg injection with wild-type or tetO-IML3/iml3Δ strains of C. albicans in the absence or presence of DOX and incubated the G. mellonella larvae at either low temperature (28°C) or elevated temperature (39°C) (Figure 6A). Upon transcriptional repression of IML3 by the addition of DOX in the inoculum, we observed a statistically significant increase in the survival of worms infected with tetO-IML3/iml3Δ when incubated for 3 days at 39°C but not at 28°C, confirming that Iml3 displays a temperature-dependent role for C. albicans fitness in vivo. Finally, we assessed the impact of IML3 on C. albicans virulence in a mouse model of infection. To achieve this, we infected BALB/c mice with either wild-type or tetO-IML3/iml3Δ C. albicans strains via retroorbital injection. IML3 expression was controlled in vivo by providing DOX in the drinking water, with no DOX treatment as a control (Figure 6B). Notably, we observed a complete attenuation of C. albicans virulence in mice infected with tetO-IML3/iml3Δ and given DOX-dosed water, with all mice surviving until the experimental endpoint of 25 days post infection. These results highlight the importance of identifying genes with condition-specific importance on C. albicans fitness, as they may have significant impacts on virulence in a mammalian host, as we have shown for IML3.

DISCUSSION

Collectively, the work presented in this study emphasizes the power of functional genomic screens to uncover crucial biological insights into fungal pathogens. The ability of C. albicans to adapt to diverse growth conditions is a key virulence trait that plays a role in its success as a human pathogen. Identifying genes important for fitness in an environmentally contingent manner not only improves our current understanding of known cellular processes but also enables the exploration of biological functions of uncharacterized portions of the genome. Although genes required for fitness in C. albicans have been studied in standard laboratory conditions,14,22,23,59 these datasets overlook genes that are only required for growth upon exposure to specific growth environments or stresses. Here, we optimized an approach that leverages the high-throughput nature of pooled screening to enable facile assessment and identification of genes required for fitness under multiple conditions. To our knowledge, our study was the first to perform genomic screens with stepwise alterations in environmental conditions with an extensive genetic mutant collection, enabling a more nuanced and comprehensive view of gene function. This resulted in the identification of two previously uncharacterized genes with either condition-independent or temperature-dependent importance for fitness. C1_09670C was shown to encode the C. albicans homolog RFA3, with critical roles in DNA repair, while C3_06880W encodes the inner kinetochore component IML3, with temperature-dependent importance for fitness. Downstream characterization of this gene confirmed its role in virulence within a mouse model of infection, highlighting that there exists a pool of genes important for C. albicans virulence that remains to be discovered.

In addition to identifying genes important for C. albicans virulence, it is also necessary to uncover biology and expand gene function annotations. Despite the complete genome sequence for C. albicans having been available for a decade, a large portion of the genome has yet to be functionally characterized.60 Comparing fitness profiles across multiple environmental conditions or in response to small molecules has been used successfully to probe gene function in S. cerevisiae,16,41 bacteria,61–63 and even humans64 by relying on the principle that genes associated with similar phenotypes across large-scale screening datasets are often functionally related. We highlight that this concept is also true for C. albicans, as comparing fitness profiles across seven of our growth conditions revealed clusters of genes that were involved in coherent biological processes and helped us to define the role of Iml3. This observation was remarkable, given that the barcoded GRACE collection is currently limited to mutant strains covering ~35% of the genome and our screens were limited to eight distinct conditions. Thus, this reinforces the importance of expanding functional genomic resources in fungal pathogens themselves as well as assessing strain fitness in an increasing number of environmental conditions to improve the power of this approach in describing gene function in this human fungal pathogen.

Through our work, we identified 171 condition-independent genes, which are enriched for genes with functions such as DNA replication, protein synthesis, and biosynthesis of macromolecules. In agreement with previous arrayed screens, the majority of these genes have been identified as essential under standard laboratory conditions. Despite the functions of many condition-independent genes having been well-described, C1_09670C remained uncharacterized. This delay in functional annotation is likely because prediction of gene function in C. albicans is conventionally reliant on sequence homology to S. cerevisiae and inference of gene function from the identified homolog, for which none could be found for C1_09670C. However, our expanded homology search of more distantly related fungal species (C. neoformans and S. pombe) yielded several homologs annotated to encode Rfa3. With increasing evidence showing that sequence and gene function can diverge substantially between C. albicans and S. cerevisiae,65–67 and a rapidly expanding repertoire of sequenced fungal genomes, a strategic pivot to expand the search scope beyond S. cerevisiae for prediction of gene function in C albicans is warranted.

In recent years, a number of powerful computational tools have been developed and are becoming more readily accessible. These advances are not only stand-alone achievements but can also be integrated with previous approaches to reveal additional avenues of hypothesis generation. For instance, although protein structure databases and search programs such as PDBsum and 3D BLAST have been around for years, the introduction of predicted protein structures generated by AlphaFold dramatically increases the utility of these tools by expanding the searchable collection of protein structures beyond the previously limited number of experimentally resolved structures. Leveraging these computational approaches, we were able to identify and experimentally validate C3_06880w as an Iml3 homolog. Collectively, this study emphasizes the need for characterizing fitness under diverse environmental conditions, highlights the importance of leveraging diverse fungal genomes and predicted protein structures to assign gene function for the fungal pathogen C. albicans and illustrates how a variety of computational approaches can be employed to define gene function. The condition-dependent and -independent genes identified and characterized in this study strongly support the use of pooled screening as a powerful tool for investigating previously unexplored biology of C. albicans.

Limitations of the study

One notable caveat to the interpretation of the pooled screening results we presented, is the use of a high screening concentration of DOX for gene repression, as a previous study demonstrated that this small molecule can chelate iron and disrupt iron homeostasis.68 As a result, there may be clusters of genes identified as hits in our screen that represent synthetic interactions between the repressed target gene and DOX rather than repression of the gene alone. To minimize the impact of this caveat, we used lower doses of DOX for assays in all follow-up characterizations. We also note that, while this study focuses broadly on genes with condition-dependent and -independent importance for fitness, the dense dataset generated by our screening efforts also identified alternative phenotypes such as genes that increase fitness upon transcriptional repression. Follow-up on these cases was not within the scope of this study, but the data we acquired could provide valuable information for understanding C. albicans biology in future studies and are provided in full in Table S1.

STAR★METHODS

RESOURCE AVAILABILITY

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Leah Cowen (leah.cowen@utoronto.ca).

Materials availability

C. albicans strains generated in this study will be provided by the lead contact without restriction as long as stocks remain available and reasonable compensation is provided by the requestor to cover processing and shipment.

Data and code availability

Next generation sequencing data will be deposited at SRA via GEO and are publicly available as of the date of publication. Accession number is listed in the key resources table. All affinity purification mass spectrometry data have been deposited at MassIVE and are publicly available as of the date of publication. Accession numbers are listed in the key resources table.

All original code and data used to generate figures are available at https://github.com/csbio/C.albicans_in_vitro_pipeline, deposited at Zenodo, and are publicly available as of the date of publication. DOI is listed in the key resources table.

Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Mouse model

All mouse experiments were compliant with the Animal Welfare Act, the Guide for the Care and Use of Laboratory Animals, and the Office of Laboratory Animal Welfare and were conducted with approval from UCSF Institutional Animal Care and Use Committee (protocol number AN189431-01). Animals were provided with fresh water and chow ad libitum. Housing was performed under a 12:12 light dark cycle in a well-ventilated room (10–15 air changes per hour) with stabilized temperature (68–79°F) and humidity (30–70%).

8- to 10-week-old female mice (purchased from Charles River Laboratories; Stock No: 028) were randomly grouped with 8 mice per treatment and 4 mice per cage. For the survival curve experiment, mice were given sterile drinking water containing 5% glucose with or without doxycycline hyclate (final concentration 250 μg/mL) for 10 days before infection. Glucose was included to ensure sufficient water uptake. C. albicans strains were injected retro-orbitally and mice were monitored daily for signs of illness and euthanized upon reaching a moribund state defined by the following criteria: hunched posture, decreased motor activity, respiratory distress, or BCS of 2 or less.

Galleria mellonella larvae model

Galleria mellonella larvae were obtained from Reptilia (Vaughan, Canada). Larvae were stored at room temperature (68–79°F) in the dark prior to infection. Larvae were weighed and sorted into groups of 20 in petri dishes containing woodchips. For the survival curves, only larvae weighing between 230 and 300 mg were used. C. albicans strains were injected into the last left proleg of the larvae and were monitored for 72 h for death as defined by a lack of response to physical touch. No food or water was provided during the experiment and pupae were censored from the dataset. Larvae and pupae were euthanized by freezing at −20°C.

C. albicans strains and growth conditions

All strains used in this study are listed in the key resources table and were derived from the CaSS1 or SN95 background strains of C. albicans. Archives of all strains were maintained in yeast extract peptone dextrose (YPD; 1% Bacto yeast extract, 2% Bacto peptone, 2% glucose) with 25% glycerol at −80°C. Active cultures were maintained on YPD supplemented with 2% agar at 4°C for no longer than 1 month. Liquid cultures were grown in yeast nitrogen base (YNB) medium at 30°C unless otherwise indicated (0.17% YNB without amino acids and ammonium sulfate, 2% glucose, 0.1% glutamic acid (MSG), 0.096 mM histidine). Where indicated, strains were grown in the presence of doxycycline (DOX) dissolved in filter-sterilized water.

METHOD DETAILS

Grace pool growth

Individually cultured C. albicans GRACE mutants were grown in YPD to saturation in 96-well plates and then combined in equal volumes and archived in 25% glycerol. Stock pools were thawed, diluted into 24-well plates (Falcon, 353047) in triplicate 1 mL cultures to an OD600 of 0.05 with or without 100 μg/mL DOX. Cultures were incubated for 24 h in a static manner in the conditions of interest or until OD600 of cultures reached approximately 5.0 (Table S1). The cultures were each sub-cultured into triplicate 1 mL cultures with fresh medium at an OD600 of 0.05 with or without 100 μg/mL DOX and incubated for an additional 48 h or until OD600 of cultures reached approximately 5.0 (Table S1). All OD600 values were measured using the Ultrospec 3100 pro UV/visible spectrophotometer.

Screens in all conditions were performed in technical triplicate and in biological duplicate with the exception of the baseline condition, which was conducted with five biological replicates. Entire 1 mL volumes of culture were collected, pelleted by centrifugation, and stored at −80°C until processed.

Genomic DNA preparation and sequencing

Genomic DNA was purified using the PureLink Genomic DNA Mini Kit as per instructions with minor modifications, as described. Cell pellets were resuspended in 400 μL of lysis buffer (from PureLink Kit) and physically lysed by adding approximately 200 μL of acid-washed glass beads and bead beating six times for 30 s. Extracted genomic DNA was quantified with the Quant-iT PicoGreen dsDNA Assay Kit. Barcode amplification by PCR was performed using 105.6 ng of genomic DNA via the Ex-Taq enzyme using the following thermal cycler program.

94°C, 120 sec

94°C, 20 sec

53°C, 20 sec

72°C, 14 sec

Go To Step 2, x28 Cycles

72°C, 60 sec

4°C, Hold

UPTAG and DNTAG primer sequences are included in Table S5. Separate UPTAG and DNTAG multiplexed pools were formed by combining equal amounts of PCR product from each sample. Pooled UPTAG and DNTAG DNA pools were extracted from a 1% agarose gel using the QIAquick Gel Extraction Kit. Equal quantities of UPTAG and DNTAG pools as determined by PicoGreen assay were combined to form a library, which was sequenced on an Illumina MiSeq v2 Regular or NextSeq v2.5 Mid-Output instrument (single-end flow cell) using specific primers to sequence and index the UPTAG and DNTAGs (Table S5).

Barcode sequence reads were mapped to an artificial genome containing known UPTAG and DNTAG sequences via Bowtie.74 Read frequency for the UPTAG and DNTAG of each strain were compiled for each indexed sample. Correlation of the UPTAG and DNTAG reads are demonstrated in Fig. S1C–D. Screens in all conditions were performed in biological duplicate and technical triplicate, with the exception of the baseline condition, which was performed with five biological replicates. Correlations between technical and biological replicates found in Fig. S1B–D.

Statistical scoring of mutant growth effects

Scores measuring the growth effects of each mutant were measured by computing the Log2FC fold-change (Log2FC) of the NO DOX control sample with respect to the corresponding DOX sample. Before this Log2FC calculation, total read count normalization was performed for each replicate sample (read counts were multiplied by 106/total read counts in that condition) and pseudocounts were added to each barcode as described in the formula below: log2FC=log2normalizedreadsof(NODOX)normalizedreadsof(+DOX)

normalizedreads=reads×106totalreadsincondition+1

Strains for which the NO DOX sample read counts were <50 were excluded from the analysis. The Log2FC scores for each condition were then normalized to guarantee equal variance across all replicate screens per tag (either UPTAG or DNTAG). Specifically, we computed a per-screen standard deviation, SDs, then computed a median standard deviation across all screens, SDm, and multiplied each screen by the ratio SDmSDs. For each strain in each condition, we then applied a moderated t test to assess the significance of the difference from the median effect in that condition. Specifically, we tested the hypothesis that MeanSi,j,t-REF=0 for each strain, Si, in each condition, where j refers to the replicate, t refers to either UPTAG or DNTAG, and REF was computed by first averaging per-strain effects across m replicates in that condition and then computing the median across all the n strains (i.e., REF=Median(Mean1,Mean2,…Meann), where Meani=1m∑j=1m Sij,t for each tag t). The R package limma75 was used to perform this moderated t test. From the results for each condition, mutant strains that satisfied Log2FC≧2 and FDR ≦ 0.05 were annotated as having a significant growth defect in that condition. In addition, to investigate the strains with significantly stronger defects than the baseline condition (YNB at 30°C), we also conducted a moderated t test on the difference between the Log2FC of each condition and the baseline condition where the same cutoffs were applied (Log2FC≧2 and FDR ≦ 0.05) to define condition-specific hits. More specifically, we calculated the differential Log2FC (dLFC) values by subtracting the mean Log2FC values of baseline replicates (we used YNB 30°C as this reference) from the Log2FC values of each condition’s replicates. We report the effect size, which represents the difference between the mean Log2FC values of the NO DOX vs. DOX screens within each condition, and we also provide the dLFC of each condition in comparison with YNB 30°C.

Liquid growth assay

Individual mutant strains were grown in the condition of interest under static conditions in triplicate 1 mL cultures in 24-well plates (Falcon, 353047) with or without 100 μg/mL DOX. Optical density at 600 nm (OD600) was measured after 24 h then cultures were sub-cultured into fresh medium with the same concentration of DOX at a starting OD600 of 0.05. Endpoint optical density was measured after 48 h of growth.

Methyl methanesulfonate sensitivity assay

Strains were grown overnight in YPD in the absence or presence of 0.05 μg/mL DOX, then spotted onto YPD agar supplemented with DOX (0, 0.05, or 20 μg/mL) and/or MMS (0 or 0.01%). Cultures were spotted in 2-fold dilutions starting at an OD600 of 10. Images taken after 48 h of growth at 30°C. MMS supplemented agar plates were stored in the dark at 4°C and used within 24 h of preparation.

Protein function and structure analyses

Gene Ontology (GO) enrichment analysis was performed based on Biological Process annotations84,85 from CGD60 with identifiable barcoded genes as the background set. Conserved protein domain searches were performed using Pfam30 integrated in InterPro76 (https://www.ebi.ac.uk/interpro/). Protein Basic Local Alignment Search Tool (BLASTp)77 searches were performed on NCBI web tool (https://blast.ncbi.nlm.nih.gov/Blast.cgi?PAGE=Proteins), FungiDB78 (https://fungidb.org/fungidb/app/workspace/blast/new), and CGD60 (http://www.candidagenome.org/cgi-bin/compute/blast_clade.pl). Similarity between protein sequences were determined using EMBOSS Needle33 (https://www.ebi.ac.uk/Tools/psa/emboss_needle/) with default parameters. AlphaFold32 predicted structures obtained from UniProt31 Website (https://www.uniprot.org/). Query for similar protein structures performed using 3D BLAST79 (http://3d-blast.life.nctu.edu.tw) and PDBSum44 (http://www.ebi.ac.uk/thornton-srv/databases/pdbsum/). Similarity between pairs of protein structures assessed through TM scores calculated by TM-Align46 (https://zhanggroup.org/TM-align/). Images of overlayed structures generated through Pairwise Structural Alignment tool on RCSB PDB47 Website (https://www.rcsb.org/alignment) with jFATCAT (rigid) parameters using structural files downloaded from UniProt.31 Cytoscape80 was used to visualize high-confidence protein interactors identified by AP-MS.

Functional evaluation of mutant profile similarity and clustering analysis

We applied FLEX (FunctionaL Evaluation of eXperimental perturbations)42 to leverage GO Biological Process functional annotations for assessing the functional coherence of mutant profiles across the set of screens we performed (Table S6). For these analyses, we focused on the subset of mutants that showed at least one effect in at least one of the conditions. Specifically, we selected 23.4% of the tested strains exhibiting Log2FC≧1 in at least one of the seven YNB-based conditions. FLEX generates a Precision-Recall curve that measures the extent to which gene pairs’ profile similarity across the conditions is predictive of co-annotation to a functional standard (in our case, annotations to specific GO terms). Strong precision-recall performance indicates that profile similarity is identifying genes known to have common functions. FLEX also produces a corresponding diversity plot that describes the relative contributions of different functional categories to the true positive pairs identified (pairs that have high profile correlation and were found to be co-annotated). FLEX demonstrates the consistency of GRACE profiles with known gene functions and the intrinsic value of profile similarity in uncovering functional relationships. More details on FLEX evaluations can be found in the original FLEX paper.42 In addition to the FLEX evaluations, we implemented a modified hierarchical clustering algorithm to cluster the mutant effect profiles. A standard average linkage hierarchical clustering algorithm with a Pearson correlation metric was modified to achieve more uniform cluster sizes across several different layers of the resulting dendrogram. Specifically, at each layer of the clustering, an iterative process is applied to identify the clusters at that layer. At each iteration, clusters are nominated using average linkage scoring and assessed for cluster size and the cluster signal-to-noise (SNR) ratio, which is computed as the average within-cluster PCC divided by the within-cluster PCC standard-deviation. At each iteration, clusters that satisfy a minimum size criteria (≧ 3 members in this application) and score in the top specified percentile in terms of SNR (we used 1%) are selected and removed from further consideration until the next layer of clustering. Remaining unclustered elements at that layer are then reclustered and reassessed for cluster size and SNR. The iterations for a single layer of clustering continue until there is ≦ 1 element remaining to be clustered, or no additional clusters meet the required criteria (min. size and min. SNR cutoff). Once iterations at a given layer are complete, all clusters formed at that layer are then considered as single elements for the next layer of clustering (as with standard hierarchical clustering). This iterative process within each layer prevents a small number of highly coherent clusters from dominating the cluster hierarchy and results in more uniform clusters. For our analysis of these mutant profiles, we chose the 48 clusters that resulted from the second layer of this modified hierarchical clustering. Details of the cluster members and the hierarchy are provided in Table S3. The implementation of the modified hierarchical clustering algorithm and all the other parameters can be accessed in the provided source code.

Reverse transcription quantitative PCR

YPD overnights of the wild-type strain and the GRACE mutants were subcultured to an OD600 of 0.1 in fresh YPD in the absence and presence of 0.05 μg/mL DOX. Cultures were grown overnight at 30°C, shaking. The next day, cultures were subcultured again to an OD600 of 0.2 in fresh YPD in the absence and presence of 0.05 μg/mL, 5 μg/mL, and 20 μg/mL DOX. Cultures were grown at 30°C, shaking for ~4 h or until in mid-log phase. Cells were pelleted, washed once with cold 1× phosphate buffered saline (PBS), flashfrozen, and stored at −80°C. Cells were lysed by bead beating four times for 30 s, with 1 min on ice in between. RNA was extracted using the RNeasy kit and DNase treated using the RNase free DNase Set. cDNA synthesis was preformed using the iScript cDNA synthesis Kit. qRT-PCR was performed in technical triplicate with a 10 μL reaction volume in a 384-well plate, using Fast SYBR Green Master Mix and the BioRad CFX-384 Real Time System. The following cycling conditions were used.

95°C, 120 sec

95°C, 10 sec

60°C, 30 sec

Go To Step 2, x 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 Table S5.

Filamentation assay

Overnight cultures of the indicated strains were grown in YPD at 30°C, shaking, and subcultured to an OD600 of 0.1 in fresh YPD in the absence and presence of DOX in the indicated concentrations. Cultures were grown at 30°C for 4 h before imaging on a Zeiss Axio Imager.MI. A wild-type strain was also subcultured into medium containing 50 μM nocodazole and grown for 4 h prior to imaging.

C. albicans strain construction

CaLC9181. This strain has both copies of C3_06880W (IML3) C-terminally tagged with GFP. It was made using a transient CRISPR approach adapted from Min et al.86 The GFP-NAT cassette was PCR amplified from pLC389 using oLC10k662 and oLC10k663. The CaCAS9 cassette was amplified from pLC963 using oLC6924 and oLC6925. The sgRNA fusion cassette was PCR amplified from pLC963 with oLC5978 and oLC10k661 (fragment A) and oLC5980 and oLC10k660 (fragment B), and fusion PCR was performed on fragments A and B using the nested primers oLC5979 and oLC5981. The GFP-NAT cassette, sgRNA, and Cas9 DNA were transformed into CaLC239. Upstream integration was PCR tested using oLC600 and oLC10k664, and downstream integration was tested using oLC274 and oLC10k665. Lack of a wild-type allele was PCR tested using oLC10k665 and oLC10k666.

CaLC9182. This strain has both copies of C3_06880W (IML3) C-terminally tagged with GFP and both copies of MTW1 C-terminally tagged with RFP. It was made using a transient CRISPR approach adapted from Min et al.86 The GFP-NAT cassette was PCR amplified from pLC389 using oLC10k662 and oLC10k663. The CaCAS9 cassette was amplified from pLC963 using oLC6924 and oLC6925. The sgRNA fusion cassette was PCR amplified from pLC963 with oLC5978 and oLC10k661 (fragment A) and oLC5980 and oLC10k660 (fragment B), and fusion PCR was performed on fragments A and B using the nested primers oLC5979 and oLC5981. The GFP-NAT cassette, sgRNA, and Cas9 DNA were transformed into CaLC7414. Upstream integration was PCR tested using oLC600 and oLC10k664, and downstream integration was tested using oLC274 and oLC10k665. Lack of a wild-type allele was PCR tested using oLC10k665 and oLC10k666.

CaLC9687. This strain has both copies of C1_09670C (RFA3) C-terminally tagged with GFP. It was made using a transient CRISPR approach adapted from Min et al.86 The GFP-ARG cassette was PCR amplified from pLC1205 using oLC12270 and oLC12271. The CaCAS9 cassette was amplified from pLC963 using oLC6924 and oLC6925. The sgRNA fusion cassette was PCR amplified from pLC963 with oLC5978 and oLC12268 (fragment A) and oLC5980 and oLC12269 (fragment B), and fusion PCR was performed on fragments A and B using the nested primers oLC5979 and oLC5981. The GFP-ARG cassette, sgRNA, and Cas9 DNA were transformed into CaLC239. Upstream integration was PCR tested using oLC11313 and oLC10k761, and downstream integration was tested using oLC4722 and oLC12273. Lack of a wild-type allele was PCR tested using oLC11313 and oLC12273.

Plasmids used for strain construction are listed in the key resources table and primers are listed in Table S5.

Fluorescence microscopy

For co-localization, overnight cultures of CaLC9181, CaLC7414, CaLC9182, and CaLC9687 were grown in YPD at 30°C with shaking. Cells were subcultured to an OD600 of 0.1 in fresh YPD and grown for 4 h at 30°C with shaking. Cultures were then washed once with PBS, resuspended in 1 mL PBS, then further incubated with 5 μg/mL Hoechst 33342 and ProLong Live AntiFade on a rotating platform for 10 min in the dark at room temperature. Cells were imaged by differential interference (DIC) microscopy, GFP channel (C3_06880w-GFP and C1_09670c-GFP), dsRed channel (Mtw1-RFP), and DAPI channel (Hoechst 33342) on a Zeiss Axio Observer.Z1 with Zeiss ApoTome Attachment (for CaLC9181, CaLC7414, and CaLC9182) or Zeiss Axio Imager.MI. (for CaLC9687) with consistent exposure time between samples.

Quantification of foci within nuclei in response to MMS treatment

CaLC9687 was grown, stained, and imaged as described above with the modification that cells were grown in YPD with and without 0.01% MMS for the 4-h subculture. In three images of non-overlapping fields of view, the numbers of cells with no foci, a single focus, and multiple foci per nucleus were tabulated and a percentage based on total cells was calculated for each category. Quantification results are plotted as mean of three images with SD.

Mass spectrometry

The GFP-tagged Iml3 (CaLC9181) or GFP-tagged Rfa3 (CaLC9687) 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 1720 × g for 30 min at 4°C, washed with ice-cold water, and snap-frozen in liquid nitrogen bath. As a control, GFP-tagged Eno1 (CaLC4449) was prepared in the same manner. For protein extraction, samples were diluted 1:1 by weight in lysis buffer (50 mM HEPES [pH 7.5], 150 mM NaCl, 5 mM EDTA, 5 mM 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 1 min at 100 × g with a 27½-gauge needle and clarified by centrifugation at 16,110 × g for 20 m in 4 °C in a microcentrifuge. To affinity purify the GFP-interacting proteins, we used the GFP trap affinity resin (ChromoTek, Martinsried, Germany). The resin was 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]), using 25 μL resin for each 1 L culture. The protein extract was then added to the resin and rotated for 2 h at 4 °C. The beads were 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 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. The beads were then magnetized, and the supernatant was transferred to a fresh tube and incubated with 0.5 μg trypsin without rotation at 37 °C for 4 h. Trypsin digested peptides were stored in 2% formic acid at −40°C until MS acquisition.

AP-MS samples and controls were analyzed by MS in two biological replicates, in data-dependent acquisition (DDA) LC-MS/MS on a 90-min gradient. Previously analyzed candidate ions were dynamically excluded for 7 s. MS data generated were stored, searched, and analyzed using the ProHits laboratory information management system (LIMS) platform and searched using Mascot (v2.3.02)83,87 and Comet (v2016.01)83,88 against the C. albicans RefSeq database (txid5476[Organism:exp]) acquired from NCBI, supplemented with “common contaminants” from the Global Proteome Machine82 (GPM; ftp://ftp.thegpm.org/fasta/cRAP/crap.fasta) and forward & reverse sequences (labeled “DECOY”) for a total of 30,018 entries. Database parameters were set to search for tryptic cleavages, allowing up to 2 missed cleavages sites per peptide with a mass tolerance of 35 ppm for precursors with charges of 2+ to 4+ and a tolerance of 0.15 amu for fragment ions. Variable modifications were selected for deamidated asparagine and glutamine and oxidized methionine. Results from each search engine were analyzed through TPP83 (the Trans-Proteomic Pipeline, v.4.7 POLAR VORTEX rev 1) via the iProphet75 pipeline. SAINTexpress (v3.3)49 was used as a statistical tool to calculate the probability value of each potential protein–protein interaction from background contaminants using default parameters and a ProteinProphet cutoff of 0.95 and a minimum of two unique peptides. A total spectral count greater than or equal to 5 and a BFDR of lower than or equal to 0.01 was required for proteins to be classified as significant interaction partners (Table S4). Data have been deposited as a complete submission to the MassIVE repository (https://massive.ucsd.edu/ProteoSAFe/static/massive.jsp) and assigned the accession number MSV000092994. The ProteomeXchange accession is PXD045766.

Mouse model of C. albicans infection

8- to 10-week-old female BALB/c mice (Charles River) were grouped with 8 mice per group and 4 mice per cage. They were initially provided with sterile drinking water for one day after arrival. On the subsequent day, 16 mice were given sterile drinking water containing 5% glucose with doxycycline hyclate (final concentration 250 μg/mL) for 10 days before infection, while another 16 mice received sterile drinking water with 5% glucose as a control. Glucose was included to ensure sufficient water uptake. C. albicans strains were grown in YPD overnight at 30°C. Cells were diluted to OD600 0.1 in 100 mL YPD broth and incubated at 30°C for 4 h with shaking. Following two washes with 0.9% saline, cells were counted (4 × 105 cells per mouse) and injected retro-orbitally. Mice were monitored daily for signs of illness and euthanized upon reaching a moribund state defined by the following criteria: hunched posture, decreased motor activity, respiratory distress, or BCS of 2 or less. Animal experiments were conducted with approval from UCSF Institutional Animal Care and Use Committee (protocol number AN189431–01).

Invertebrate model of C. albicans infection

C. albicans wild type (CaSS1) and the tetO-IML3/iml3Δ mutant were grown overnight shaking in YPD at 30°C with or without 20 μg/mL doxycycline. Cells were counted using a hemocytometer to prepare inoculum stocks 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. Doxycycline was added to stocks where applicable to a final concentration of 500 μg/mL. Larvae were screened using a weight filter of 230 mg–300 mg and lack of dark spots indicative of injury and visible signs of pupation. Each larvae was 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) with either PBS containing no cells as a control, wild type, or tetO-IML3/iml3Δ mutant cells with 20 worms per treatment group (strain, dox/no dox, temperature). Larvae were incubated in petri dishes by treatment group at indicated temperatures and monitored for 72 h for mortality as defined by lack of response to physical touch. Mortality was first checked at 12 h post infection, then every 4 h between 12 and 24 h post infection, every 6 h between 24 and 60 h post infection, and finally every 12 h thereafter until the endpoint of 72 h. Larvae and pupae were euthanized by freezing at −20°C.

QUANTIFICATION AND STATISTICAL ANALYSIS

Statistical significances are reported in the methods sections as well as Figure and Figure Legends. Data are considered to be statistically significant when p < 0.05 for two-way ANOVA with Bonferroni correction for multiple comparisons and p < 0.0332 for Log rank (Mantel-Cox) tests. Asterisks denote statistical significance in figures analyzed using two-way ANOVA (*p < 0.05, ***p < 0.0005, ****p < 0.0001) and Log rank (Mantel-Cox) (**p < 0.0021, ***p < 0.0001). Statistical analyses were performed in RStudio81 and GraphPad Prism.

Supplementary Material

1

2

3

4

5

6

7

ACKNOWLEDGMENTS

We thank Merck and Genome Canada for making the original C. albicans GRACE mutant collections available. We thank all past and current Cowen lab members for helpful discussions and assistance in expanding the GRACEv2 collection. Proteomics was performed at the Network Biology Collaborative Center at the Lunenfeld-Tanenbaum Research Institute, a facility supported by Calgary Foundation for Innovation Funding, by the Ontario Genomics, and by Genome Canada and Ontario Genomics (OGI-139). A.-C.G. is supported by the Canadian Institutes of Health Research (FDN 143301) and a Canada Research Chair, Tier 1, in Functional Proteomics. X.Z., H.Y., C.L.M., S.M.N., and L.E.C. were supported by grant R01AI127375 from the National Institutes of Health, and H.N.W. was supported by a Doctoral Dissertation Fellowship from the University of Minnesota and grant R01HG005084 from the National Institutes of Health. E.H.X. is supported by an Natural Sciences and Engineering Research Council of Canada: CREATE EvoFunPath Fellowship. L.E.C. is also supported by a Canadian Institutes of Health Research Foundation grant (FDN-154288) and is a Canada Research Chair (Tier 1) in Microbial Genomics & Infectious Disease and Co-director of the CIFAR Fungal Kingdom: Threats & Opportunities program. BioRender was used to generate some schematics in the graphical abstract.

Figure 1. Pooled screening of the GRACE library in diverse growth conditions identifies genes important for fitness in condition-independent and condition-dependent manners

(A) Simplified schematic of the pooled screening experimental pipeline.

(B) Simplified schematic of the pooled screening computational analysis pipeline. A fitness score was calculated for each strain using the log2FC in barcode read counts between “no DOX” compared to DOX conditions. Significance is determined by comparing each strain’s fitness score to the median fitness score in each condition using a moderated t test with Benjamini-Hochberg correction. Strains with LFC ≧ 2.0 and FDR ≦ 0.05 are deemed hits.

(C) Pooled screening in eight growth conditions identifies hundreds of genes important for fitness in each condition. Numbers reflect the number of genes identified as hits in each condition. All screening was performed with technical triplicates in biological duplicate, except for the baseline condition performed in biological quintuplicate.

See also Figures S1 and S2.

Figure 2. Identification of condition-specific hits and functional enrichment analysis

(A) Upset plot highlighting genes important for fitness in specific conditions. Horizontal bars represent the total number of hits for each unique combination of conditions, indicated by dots below the column. Vertical bars indicate total number of hits for each specific condition.

(B–E) Dot plots representing GO enrichment results for (B) 171 condition-independent hits, (C) 36 genes important for fitness in all conditions except YPD, (D) 18 hits exclusively identified at high temperature, and (E) 12 genes important for fitness in YPD and YNB supplemented with 10% serum. FDR is shown on the color scale, with black being the most significant.

Figure 3. C1_09670C is important for C. albicans fitness in all conditions and encodes RFA3

(A) Strains were grown in various conditions of interest in the absence or presence of 100 μg/mL DOX for 24 h and then subcultured into fresh medium to a starting optical density 600 (OD600) of 0.05. OD was measured after an additional 48 h of growth after subculture. Heatmap values represent the mean of three biological replicates.

(B) Protein structures of AlphaFold-generated C. albicans C1_09670c (UniProt: Q5APD5), S. pombe Ssb3 (UniProt: Q92374), and C. neoformans Rfa3 (UniProt: J9VL31) and protein structure of S. cerevisiae Rfa3 (PDB: 6I52) determined by cryoelectron microscopy. The color of regions in AlphaFold-predicted structures indicates the predicted local distance difference test (pLDDT), with a scale of dark blue to yellow to red indicating high to low confidence. TM score was calculated by TM-align. The TM score range is between 0 and 1, with 1 indicating a perfect match and <0.2 corresponding to randomly chosen unrelated structures.

(C) Strains were grown overnight in YPD in the absence or presence of 0.05 μg/mL DOX and then subcultured into YPD supplemented with 0 or 0.5 μg/mL DOX. Images were taken after 4 h of growth at 30°C. The wild type (CaSS1) was also subcultured into YPD supplemented with 50 μM nocodazole. Scale bar corresponds to 10 μm. The experiment was performed in biological duplicates with similar results.

(D) Strains were subcultured to an OD600 of 0.1 in fresh YPD and allowed to grow for 4 h before visualization by fluorescence microscopy, with Rfa3 shown in green and nuclei stained with Hoechst 33342 shown in blue. Insets show enlarged images of boxed area, with the cell outline overlayed in gray dashed lines. Scale bars corresponds to 10 μm or 5 μm for insets. The experiment was performed in biological duplicates with similar results.

(E) Strains were grown overnight in YPD in the absence or presence of 0.05 μg/mL DOX, at which point they were spotted onto YPD agar supplemented with combinations of 0, 0.05, or 20 μg/mL DOX and 0 or 0.01% MMS, as indicated, in 2-fold dilutions starting from an OD600 of 10. Images were taken after 48 h of growth at 30°C. The experiment was performed in biological duplicates with similar results.

(F) A strain expressing GFP-tagged Rfa3 was grown overnight in YPD at 30°C prior to subculture in fresh medium with or without 0.01% MMS, and images were taken after 4 h. The number of foci per nucleus was quantified from three images of non-overlapping fields of view, and the percentages of cells with no foci, a single focus, and multiple foci per nucleus were calculated based on total cells in each image. The scale bar corresponds to 10 μm. Representative images are shown and quantification results plotted as mean ± SD of three images from a single biological replicate. The experiment performed in biological duplicates with similar results.

(G) AP-MS of GFP-tagged Rfa3. Cells were grown in YPD at 30°C, and statistically significant interactions were defined through SAINTexpress analysis compared to the Eno1-GFP control. All proteins represented in nodes were identified with high confidence (Bayesian false discovery rate [BFDR] threshold of <0.01, total spectral counts ≥ 5, and fold-change relative to Eno1 ≥ 50) and colored based on GO term annotations. Edge color reflects the fold change in peptide count of Rfa3 relative to Eno1, with darker colors representing greater fold change. Members of the RFA complex and RFC complex are outlined in red and orange, respectively.

See also Figures S2–S4.

Figure 4. Functional evaluation of mutant profile similarity and analysis of clusters

(A) Precision-recall (PR) performance analysis of the mutant profiles in comparison to established functional annotations (GO biological processes) using the FLEX software. The PR curve measures the extent to which gene pairs’ profile similarity across the conditions is predictive of co-annotation to a functional standard (in this case, annotations to specific GO terms). Strong PR performance indicates that profile similarity is identifying genes known to have common functions. The x axis represents the logarithmic scale of the total count of true positives (TPs), and the y axis measures precision. The dashed line corresponds to a random baseline expected of an uninformative similarity measure.

(B) Diversity plot showing the contribution of GO biological processes to TP pairs identified in the PR analysis (i.e., pairs that are highly correlated in their profile and also co-annotated to a common GO term). The plot is generated by systematically adjusting a precision cutoff from high to low (corresponding to cutoffs high to low on the similarity score), as indicated on the y axis. At each cutoff point, a stacked bar plot is created along the x axis to report the number of gene pairs (TP) coming from each of the corresponding process terms. The legend includes the top 10 contributing GO terms, while the light gray category encompasses all other terms that appear less frequently. For example, the term capturing the most TP gene-pairs across a range of cutoffs is “amino acid metabolic process,” especially at higher correlation thresholds. This plot details which processes are responsible for the functionally related pairs correctly captured by the highest profile similarities.

(C) Heatmaps visualizing the clusters from the modified hierarchical clustering algorithm. For visualization purposes, 48 clusters from this algorithm are shown in a single heatmap and were ordered by a standard hierarchical clustering of the cluster centroids (the resulting dendrogram is shown). Each mutant strain’s profile was Z score normalized across the conditions. The three clusters highlighted at the bottom are those that yield significant GO enrichment results given a cutoff of FDR < 0.05. Systematic and standard names of each gene are provided. S.cer represents the name of its S. cerevisiae ortholog.

Figure 5. C3_06880W encodes kinetochore component Iml3 and is specifically required for growth at physiological temperature

(A) Differential log2FC comparison between 37°C and 30°C revealed C3_06880W among top genes required for growth at 37°C. The differential log2FC (dLFC) indicates the mean difference in log2FC for each gene between 30°C and 37°C. Genes with dLFC that differ significantly from zero were identified using a moderated t test, generating a p value with Benjamin-Hochberg correction (padj). Significance and effect size thresholds are applied at FDR = 0.05 and dLFC = |2.0|, respectively.

(B) Strains were grown at 30°C, 37°C, and 42°C in the absence or presence of 100 μg/mL DOX for 24 h and then subcultured into fresh medium to a starting OD600 of 0.05. OD was measured after an additional 48 h of growth after subculture. The plot shows mean ± SD of growth relative to the wild type in the absence of DOX for biological triplicates. Statistical significance was assessed using two-way ANOVA with Bonferroni correction for multiple comparisons: *p < 0.05, ***p < 0.0005, ****p < 0.0001; n = 3.

(C) Pairwise structure alignment and TM-align reveal a high degree of overlap in superimposed protein structures of AlphaFold-predicted C3_06880w (orange; UniProt: Q5ADM8) and crystallized Iml3 of S. cerevisiae (blue; PDB: 4JE3) and favorable template modeling (TM) scores.

(D) Strains were grown overnight in the absence or presence of 0.05 μg/mL doxycycline (DOX). Strains were subsequently subcultured at a starting OD600 of 0.1 in fresh YPD in the absence or presence of 5 μg/mL DOX as indicated. The wild-type strain in the absence of DOX was treated with 50 μM nocodazole. Scale bar corresponds to 10 μm. The experiment was performed in biological duplicates with similar results.

(E) Strains were subcultured to an OD600 of 0.1 in fresh YPD and allowed to grow for 4 h before visualization by fluorescence microscopy. Shown are Iml3 (green), Mtw1 (red), and nuclei (blue). Insets show enlarged images of boxed areas, with the cell outline overlayed in gray dashed lines. The scale bars correspond to 5 μm or 2 μm for the inset. The experiment was performed in biological duplicates with similar results.

(F) AP-MS of affinity-tagged Iml3 identified kinetochore components. Cells were grown in YPD at 30°C, and statistically significant interactions were defined through SAINTexpress analysis compared with the Eno1-GFP control. All proteins were identified by AP-MS with high confidence (BFDR threshold < 0.01, total spectral counts ≥ 5, and fold change relative to Eno1 ≥ 50), with proteins as nodes colored based on GO term annotations. Edge color reflects the fold change in peptide count of Iml3 relative to Eno1, with darker colors representing greater fold change. Nodes are outlined if protein is a confirmed or putative homolog of an inner kinetochore component in S. cerevisiae. Outline color corresponds to protein complexes in (G). Source data are provided in Table S4.

(G) Simplified schematic of the S. cerevisiae inner kinetochore. Colored circles represent proteins for which homologs have been identified and characterized in C. albicans, including Iml3 encoded by C3_06880W.

See also Figure S4 and S5.

Figure 6. Depletion of C3_06880W attenuates virulence in a temperature-dependent manner

(A) Depletion of IML3 reduces virulence of C. albicans in a temperature-dependent manner in a G. mellonella larva model of infection. Transcriptional repression of IML3 by the addition of DOX to the inoculum of 1 × 106 cells significantly improved survival of larvae when incubated at 39°C but not 28°C. Log rank (Mantel-Cox) test, **p < 0.0021.

(B) Depletion of IML3 attenuates virulence of C. albicans in a mouse model of infection. Repression of Iml3 expression by the addition of DOX to the drinking water significantly improved survival relative to other conditions. Log rank (Mantel-Cox) test, ***p < 0.0001. Log rank test for trend, p = 0.0003.

KEY RESOURCES TABLE REAGENT OR RESOURCE	SOURCE	IDENTIFIER	
	
Biological samples			
	
Fetal Bovine Serum - Performance Plus	Gibco	16000–044	
	
Chemicals, peptides, and recombinant proteins			
	
Doxycycline hydrochloride	BioBasic	DB0889	
Methyl methanesulfonate	Sigma Aldrich	129925	
Nocodazole	Sigma Aldrich	M1404	
Hoescht 33342	Sigma Aldrich	14533	
ProLong™ Live AntiFade	Invitrogen	P36975	
Trypan Blue (0.4%)	Gibco	15250	
	
Critical commercial assays			
	
Purelink™ Genomic DNA Mini Kit	Invitrogen	K182002	
Quant-iT™ PicoGreen dsDNA Assay Kit	Invitrogen	P7589	
QIAquick Gel Extraction Kit	Qiagen	28704	
RNeasy Mini Kit	Qiagen	74104	
RNase-Free DNase Set	Qiagen	79254	
iScript cDNA Synthesis Kit	BioRad	1708890	
Fast SYBR Green Master Mix	Applied Biosystems	4385612	
ChromoTek GFP-Trap® Magnetic Agarose	Chromotek	gtma	
	
Deposited data			
	
Mass Spectrometry Data	MassIVE; ProteomeXchange	MSV000092994; PXD045766	
Next Generation Sequencing Data	NCBI SRA Via GEO	GSE266247	
All original code and data used to generate figures	GitHub; Zenodo	https://github.com/csbio/C.albicans_in_vitro_pipeline;
https://doi.org/10.5281/zenodo.10041523	
	
Experimental models: organisms/strains			
	
Mouse: BALB/c	Charles River Laboratories	Stock No. 028	
Wax Worm Larvae	Reptilia	N/A	
CaLC6106 - C. albicans CaSSI, (GRACE Library Parent)	Roemer et al.22	ura3::imm434/ura3::imm434 his3::hisG/his3::hisG leu2::tetRGAL4ADURA/LEU2	
tetO-RFA3/rfa3AΔ (GRACEv2 Strain)	Fu etal.14	As CaSS1, SAT1::tetO-RFA3/RFA3::HIS3	
tetO-HOF1/hof1Δ (GRACE Strain)	Roemer et al.22	As CaSS1, SAT1::tetO-HOF1/HOF1::HIS3	
tetO-HSP90/hsp90Δ (GRACE Strain)	Roemer et al.22	As CaSS1, SAT1::tetO-HSP90/HSP90::HIS3	
tetO-IML3/iml3Δ (GRACEv2 Strain)	Fu etal.14	As CaSS1, SAT1::tetO-IML3/IML3::HIS3	
tetO-MTW1/mtw1Δ (GRACEv2 Strain)	Fu etal.14	As CaSS1, SAT1::tetO-MTW1/MTW1::HIS3	
tetO-CHL4/chl4Δ (GRACE Strain)	Roemer et al.22	As CaSS1, SAT1::tetO-CHL4/CHL4::HIS3	
CaLC239 - C. albicans SN95	Noble et al.69	arg4/arg4 his1/his1 URA3/ura3::imm434
IRO1/iro1::imm434	
CaLC9687 - Strain expressing C-terminally GFP-tagged Rfa3	This study	As SN95, C1_09670C-RFP-ARG/C1_09670C-RFP-ARG	
CaLC7414 - Strain expressing C-terminally RFP-tagged Mtw1	Fu etal.14	As SN95, C1_09670C-RFP-ARG/C1_09670C-RFP-ARG	
CaLC9181 - Strain expressing C-terminally GFP-tagged Iml3	This study	As SN95, C3_06880W-GFP-NAT/C3_06880W-GFP-NAT	
CaLC9182 - Strain expressing both C-terminally RFP-tagged Mtw1 and C-terminally GFP-tagged Iml3	This study	As SN95, MTW1-RFP-ARG/MTW1 -RFP-ARG C3_06880W-GFP-NAT/C3_06880W-GFP-NAT	
CaLC4449 - Strain expressing C-terminally GFP-tagged Eno1	O'Meara et al.70	As SN95, ENO1/ENO1-GFP-NAT	
All other strains in pool used in this study are members of the GRACE or GRACEv2	Roemer et al.22 and Fu et al.14	N/A	
	
Oligonucleotides	
	
See Table S5 for list of oligonucleotides.	N/A	N/A	
	
Recombinant DNA	
	
pLC389	Gerami-Nejad et al.71	GFP-NAT - To PCR amplify C-terminal GFP fusion constructs	
pLC963	Veri et al.72	pV1393–1 - Recyclable solo system vector for sgRNA cloning and downstream CRISPR guided genome editing by expression of sgRNA and CaCas9	
pLC1205	Zhang et al.73	pFA-GFP-ARG - C-terminal GFP tagging with ARG marker	
	
Software and algorithms	
	
Bowtie	Langmead and Salzberg74	RRID:SCR_005476	
LIMMA	Ritchie et al.75	RRID:SCR_010943	
Candida Genome Database	Skrzypek et al.60	RRID:SCR_002036	
Pfam	Sonnhammer et al.30	RRID:SCR_004726	
InterPro	Paysan-Lafosse et al.76	RRID:SCR_006695	
BLASTP	Camacho et al.77	RRID:SCR_001010	
FungiDB	Basenko et al.78	RRID:SCR_006013	
EMBOSS Needle	Needleman and Wunsch33	RRID:SCR_008493	
AlphaFold Protein Structure Database	Jumper et al.32	RRID:SCR_023662	
UniProt	The UniProt Consortium31	RRID:SCR_002380	
3D BLAST	Tung et al.79	http://3d-blast.life.nctu.edu.tw	
PDBsum	Laskowski et al.44	RRID:SCR_006511	
TM-align	Zhang, Y46	RRID:SCR_024390	
Research Collaboratory for Structural Bioinformatics Protein DataBank (RCSB PDB)	Prlić et al.47	RRID:SCR_01282	
Cytoscape	Shannon et al.80	RRID:SCR_003032	
RStudio	RStudio Team81	RRID:SCR_000432	
FunctionaL Evaluation of eXperimental perturbations (FLEX)	Rahman et al.42	https://github.com/csbio/FLEX_R.	
SAINTexpress	Teo et al.49	RRID:SCR_018562	
Global Proteome Machine Database (GPM DB)	Craig et al.82	RRID:SCR_006617	
Trans-Proteomic Pipeline (TPP)	Deutsch et al.83	http://www.tppms.org/	
GraphPad Prism	N/A	RRID: SCR_002798	
Original Code for Pooled Screening Analysis	This study	https://github.com/csbio/C.albicans_in_vitro_pipeline	

Highlights

Genomic screen reveals genes important for C. albicans growth in diverse conditions

C1_09670C encodes Rfa3 with roles in DNA damage repair

C3_08880W is required for C. albicans fitness and virulence at elevated temperature

C3_08880W encodes Iml3, a component of the inner kinetochore

DECLARATION OF INTERESTS

L.E.C. is a co-founder and shareholder in Bright Angel Therapeutics, a platform company for the development of novel antifungal therapeutics. L.E.C. is a Science Advisor for Kapoose Creek, a company that harnesses the therapeutic potential of fungi.

SUPPLEMENTAL INFORMATION

Supplemental information can be found online at https://doi.org/10.1016/j.celrep.2024.114601.
==== Refs
REFERENCES

1. Blackwell M (2011). The fungi: 1, 2, 3 . 5.1 million species? Am. J. Bot. 98 , 426–438.21613136
2. Taylor LH , Latham SM , and Woolhouse ME (2001). Risk factors for human disease emergence. Philos. Trans. R. Soc. Lond. B Biol. Sci. 356 , 983–989.11516376
3. Brown GD , Denning DW , Gow NAR , Levitz SM , Netea MG , and White TC (2012). Hidden killers: Human fungal infections. Sci. Transl. Med. 4 , 165rv13.
4. Gow NAR , van de Veerdonk FL , Brown AJP , and Netea MG (2011). Candida albicans morphogenesis and host defence: discriminating invasion from colonization. Nat. Rev. Microbiol. 10 , 112–122.22158429
5. Pfaller MA , and Diekema DJ (2007). Epidemiology of invasive candidiasis: a persistent public health problem. Clin. Microbiol. Rev. 20 , 133–163.17223626
6. Kullberg BJ , and Arendrup MC (2015). Invasive candidiasis. N. Engl. J. Med. 373 , 1445–1456.26444731
7. Pappas PG , Lionakis MS , Arendrup MC , Ostrosky-Zeichner L , and Kullberg BJ (2018). Invasive candidiasis. Nat. Rev. Dis. Primers 4 , 18026.29749387
8. Robbins N , Caplan T , and Cowen LE (2017). Molecular evolution of antifungal drug resistance. Annu. Rev. Microbiol. 71 , 753–775.28886681
9. Alves R , Barata-Antunes C , Casal M , Brown AJP , Van Dijck P , and Paiva S (2020). Adapting to survive: How Candida overcomes host-imposed constraints during human colonization. PLoS Pathog. 16 , e1008478.32437438
10. Robert VA , and Casadevall A (2009). Vertebrate endothermy restricts most fungi as potential pathogens. J. Infect. Dis. 200 , 1623–1626.19827944
11. Barelle CJ , Priest CL , Maccallum DM , Gow NAR , Odds FC , and Brown AJP (2006). Niche-specific regulation of central metabolic pathways in a fungal pathogen. Cell Microbiol. 8 , 961–971.16681837
12. Duggan S , Leonhardt I , Hu€nniger K , and Kurzai O (2015). Host response to Candida albicans bloodstream infection and sepsis. Virulence 6 , 316–326. 10.4161/21505594.2014.988096.25785541
13. Polvi EJ , Li X , O’Meara TR , Leach MD , and Cowen LE (2015). Opportunistic yeast pathogens: reservoirs, virulence mechanisms, and therapeutic strategies. Cell. Mol. Life Sci. 72 , 2261–2287.25700837
14. Fu C , Zhang X , Veri AO , Iyer KR , Lash E , Xue A , Yan H , Revie NM , Wong C , Lin ZY , (2021). Leveraging machine learning essentiality predictions and chemogenomic interactions to identify antifungal targets. Nat. Commun. 12 , 6497.34764269
15. Giaever G , Chu AM , Ni L , Connelly C , Riles L , Véronneau S , Dow S , Lucau-Danila A , Anderson K , André B , (2002). Functional profiling of the Saccharomyces cerevisiae genome. Nature 418 , 387–391.12140549
16. Hillenmeyer ME , Fung E , Wildenhain J , Pierce SE , Hoon S , Lee W , Proctor M , St Onge RP , Tyers M , Koller D , (2008). The chemical genomic portrait of yeast: Uncovering a phenotype for all genes. Science 320 , 362–365.18420932
17. Costanzo M , Hou J , Messier V , Nelson J , Rahman M , VanderSluis B , Wang W , Pons C , Ross C , Ušaj M , (2021). Environmental robustness of the global yeast genetic interaction network. Science 372 , eabf8424.33958448
18. Tebbji F , Chen Y , Richard Albert J , Gunsalus KTW , Kumamoto CA , Nantel A , Sellam A , and Whiteway M (2014). A functional portrait of Med7 and the Mediator complex in Candida albicans. PLoS Genet. 10 , e1004770.25375174
19. Blankenship JR , Wormley FL , Boyce MK , Schell WA , Filler SG , Perfect JR , and Heitman J (2003). Calcineurin is essential for Candida albicans survival in serum and virulence. Eukaryot. Cell 2 , 422–430.
20. Askew C , Sellam A , Epp E , Hogues H , Mullick A , Nantel A , and Whiteway M (2009). Transcriptional regulation of carbohydrate metabolism in the human pathogen Candida albicans. PLoS Pathog. 5 , e1000612.19816560
21. Pérez JC , Kumamoto CA , and Johnson AD (2013). Candida albicans commensalism and pathogenicity are intertwined traits directed by a tightly knit transcriptional regulatory circuit. PLoS Biol. 11 , e1001510.23526879
22. Roemer T , Jiang B , Davison J , Ketela T , Veillette K , Breton A , Tandia F , Linteau A , Sillaots S , Marta C , (2003). Large-scale essential gene identification in Candida albicans and applications to antifungal drug discovery. Mol. Microbiol. 50 , 167–181.14507372
23. O’Meara TR , Veri AO , Ketela T , Jiang B , Roemer T , and Cowen LE (2015). Global analysis of fungal morphology exposes mechanisms of host cell escape. Nat. Commun. 6 , 6741.25824284
24. Lie S , Banks P , Lawless C , Lydall D , and Petersen J (2018). The contribution of non-essential Schizosaccharomyces pombe genes to fitness in response to altered nutrient supply and target of rapamycin activity. Open Biol. 8 , 180015.29720420
25. Pir P , Gutteridge A , Wu J , Rash B , Kell DB , Zhang N , and Oliver SG (2012). The genetic control of growth rate: a systems biology study in yeast. BMC Syst. Biol. 6 , 4.22244311
26. Conrad M , Schothorst J , Kankipati HN , Van Zeebroeck G , Rubio-Texeira M , and Thevelein JM (2014). Nutrient sensing and signaling in the yeast Saccharomyces cerevisiae. FEMS Microbiol. Rev. 38 , 254–299.24483210
27. Altschul SF , Gish W , Miller W , Myers EW , and Lipman DJ (1990). Basic local alignment search tool. J. Mol. Biol. 215 , 403–410.2231712
28. Belanger KD , Griffith AL , Baker HL , Hansen JN , Kovacs LAS , Seconi JS , and Strine AC (2011). The karyopherin Kap95 and the C-termini of Rfa1, Rfa2, and Rfa3 are necessary for efficient nuclear import of functional RPA complex proteins in Saccharomyces cerevisiae. DNA Cell Biol. 30 , 641–651.21332387
29. Longhese MP , Plevani P , and Lucchini G (1994). Replication factor A is required in vivo for DNA replication, repair, and recombination. Mol. Cell Biol. 14 , 7884–7890.7969128
30. Sonnhammer EL , Eddy SR , Birney E , Bateman A , and Durbin R (1998). Pfam: Multiple sequence alignments and HMM-profiles of protein domains. Nucleic Acids Res. 26 , 320–322.9399864
31. UniProt Consortium (2021). UniProt: The universal protein knowledgebase in 2021. Nucleic Acids Res. 49 , D480–D489.33237286
32. Jumper J , Evans R , Pritzel A , Green T , Figurnov M , Ronneberger O , Tunyasuvunakool K , Bates R , Žídek A , Potapenko A , (2021). Highly accurate protein structure prediction with AlphaFold. Nature 596 , 583–589.34265844
33. Needleman SB , and Wunsch CD (1970). A general method applicable to the search for similarities in the amino acid sequence of two proteins. J. Mol. Biol. 48 , 443–453.5420325
34. Berman J (2006). Morphogenesis and cell cycle progression in Candida albicans. Curr. Opin. Microbiol. 9 , 595–601.17055773
35. Bachewich C , Nantel A , and Whiteway M (2005). Cell cycle arrest during S or M phase generates polarized growth via distinct signals in Candida albicans. Mol. Microbiol. 57 , 942–959.16091036
36. Ramonatxo A , and Moriel-Carretero M (2021). Microscopy analysis of the smallest subunit of the RPA complex, Rfa3p, prompts consideration of how RPA subunits gather at single-stranded DNA sites. MicroPubl. Biol. 2021 . 10.17912/micropub.biology.000493.
37. Vylkova S , Carman AJ , Danhof HA , Collette JR , Zhou H , and Lorenz MC (2011). The fungal pathogen Candida albicans autoinduces hyphal morphogenesis by raising extracellular pH. mBio 2 , e00055.e11.21586647
38. Harrison JC , and Haber JE (2006). Surviving the breakup: the DNA damage checkpoint. Annu. Rev. Genet. 40 , 209–235.16805667
39. Wong RP , García-Rodríguez N , Zilio N , Hanulová M , and Ulrich HD (2020). Processing of DNA polymerase-blocking lesions during genome replication Is spatially and temporally segregated from replication forks. Mol. Cell 77 , 3–16.e4.31607544
40. Costanzo M , Baryshnikova A , Bellay J , Kim Y , Spear ED , Sevier CS , Ding H , Koh JLY , Toufighi K , Mostafavi S , (2010). The genetic landscape of a cell. Science 327 , 425–431.20093466
41. Costanzo M , VanderSluis B , Koch EN , Baryshnikova A , Pons C , Tan G , Wang W , Usaj M , Hanchard J , Lee SD , (2016). A global genetic interaction network maps a wiring diagram of cellular function. Science 353 , aaf1420.27708008
42. Rahman M , Billmann M , Costanzo M , Aregger M , Tong AHY , Chan K , Ward HN , Brown KR , Andrews BJ , Boone C , (2021). A method for benchmarking genetic screens reveals a predominant mitochondrial bias. Mol. Syst. Biol. 17 , e10013.34018332
43. Genschik P , Sumara I , and Lechner E (2013). The emerging family of CULLIN3-RING ubiquitin ligases (CRL3s): cellular functions and disease implications. EMBO J. 32 , 2307–2320.23912815
44. Laskowski RA , Jabłońska J , Pravda L , Vařeková RS , and Thornton JM (2018). PDBsum: Structural summaries of PDB entries. Protein Sci. 27 , 129–134.28875543
45. Hinshaw SM , and Harrison SC (2013). An Iml3-Chl4 heterodimer links the core centromere to factors required for accurate chromosome segregation. Cell Rep. 5 , 29–36.24075991
46. Zhang Y , and Skolnick J (2005). TM-align: a protein structure alignment algorithm based on the TM-score. Nucleic Acids Res. 33 , 2302–2309.15849316
47. Prlić A , Bliven S , Rose PW , Bluhm WF , Bizon C , Godzik A , and Bourne PE (2010). Pre-calculated protein structure alignments at the RCSB PDB website. Bioinformatics 26 , 2983–2985.20937596
48. Roy B , Burrack LS , Lone MA , Berman J , and Sanyal K (2011). CaMtw1, a member of the evolutionarily conserved Mis12 kinetochore protein family, is required for efficient inner kinetochore assembly in the pathogenic yeast Candida albicans. Mol. Microbiol. 80 , 14–32.21276093
49. Teo G , Liu G , Zhang J , Nesvizhskii AI , Gingras AC , and Choi H (2014). SAINTexpress: Improvements and additional features in significance analysis of interactome software. J. Proteomics 100 , 37–43.24513533
50. Szklarczyk D , Gable AL , Nastou KC , Lyon D , Kirsch R , Pyysalo S , Doncheva NT , Legeay M , Fang T , Bork P , (2021). The STRING database in 2021: Customizable protein–protein networks, and functional characterization of user-uploaded gene/measurement sets. Nucleic Acids Res. 49 , D605–D612.33237311
51. Romera D , Aguilera-Correa JJ , García-Coca M , Mahillo-Fernández I , Viñuela-Sandoval L , García-Rodríguez J , and Esteban J (2020). The Galleria mellonella infection model as a system to investigate the virulence of Candida auris strains. Pathog. Dis. 78 , ftaa067.33098293
52. Fuchs BB , O’Brien E , Khoury JBE , and Mylonakis E (2010). Methods for using Galleria mellonella as a model host to study fungal pathogenesis. Virulence 1 , 475–482.21178491
53. Li D-D , Deng L , Hu GH , Zhao LX , Hu DD , Jiang YY , and Wang Y (2013). Using Galleria mellonella-Candida albicans infection model to evaluate antifungal agents. Biol. Pharm. Bull. 36 , 1482–1487.23995660
54. Mylonakis E , Moreno R , El Khoury JB , Idnurm A , Heitman J , Calderwood SB , Ausubel FM , and Diener A (2005). Galleria mellonella as a model system to study Cryptococcus neoformans pathogenesis. Infect. Immun. 73 , 3842–3850.15972469
55. Scully LR , and Bidochka MJ (2005). Serial passage of the opportunistic pathogen Aspergillus flavus through an insect host yields decreased saprobic capacity. Can. J. Microbiol. 51 , 185–189.
56. Cotter G , Doyle S , and Kavanagh K (2000). Development of an insect model for the in vivo pathogenicity testing of yeasts. FEMS Immunol. Med. Microbiol. 27 , 163–169.
57. Brennan M , Thomas DY , Whiteway M , and Kavanagh K (2002). Correlation between virulence of Candida albicans mutants in mice and Galleria mellonella larvae. FEMS Immunol. Med. Microbiol. 34 , 153–157.12381467
58. García-Rodas R , Casadevall A , Rodríguez-Tudela JL , Cuenca-Estrella M , and Zaragoza O (2011). Cryptococcus neoformans capsular enlargement and cellular gigantism during Galleria mellonella infection. PLoS One 6 , e24485.21915338
59. Segal ES , Gritsenko V , Levitan A , Yadav B , Dror N , Steenwyk JL , Silberberg Y , Mielich K , Rokas A , Gow NAR , (2018). Gene essentiality analyzed by in vivo transposon mutagenesis and machine learning in a stable haploid isolate of Candida albicans. mBio 9 , 020488.
60. Skrzypek MS , Binkley J , Binkley G , Miyasato SR , Simison M , and Sherlock G (2017). The Candida genome database (CGD): Incorporation of assembly 22, systematic identifiers and visualization of high throughput sequencing data. Nucleic Acids Res. 45 , D592–D596.27738138
61. Ding D.w. , Huang W.f. , Lei L.l. , and Wu P (2022). Co-fitness analysis identifies a diversity of signal proteins involved in the utilization of specific c-type cytochromes. Ann. Microbiol. 72 , 38.
62. Kim P-J , and Price ND (2011). Genetic co-occurrence network across sequenced microbes. PLoS Comput. Biol. 7 , e1002340.22219725
63. Leshchiner D , Rosconi F , Sundaresh B , Rudmann E , Ramirez LMN , Nishimoto AT , Wood SJ , Jana B , Buján N , Li K , (2022). A genome-wide atlas of antibiotic susceptibility targets and pathways to tolerance. Nat. Commun. 13 , 3165.35672367
64. Kim E , Dede M , Lenoir WF , Wang G , Srinivasan S , Colic M , and Hart T (2019). A network of human functional gene interactions from knockout fitness screens in cancer cells. Life Sci. Alliance 2 , e201800278.30979825
65. Braun BR , van Het Hoog M , d’Enfert C , Martchenko M , Dungan J , Kuo A , Inglis DO , Uhl MA , Hogues H , Berriman M , (2005). A human-curated annotation of the Candida albicans genome. PLoS Genet. 1 , 36–57.16103911
66. Kellis M , Birren BW , and Lander ES (2004). Proof and evolutionary analysis of ancient genome duplication in the yeast Saccharomyces cerevisiae. Nature 428 , 617–624.15004568
67. O’Meara TR , and O’Meara MJ (2021). DeORFanizing Candida albicans genes using coexpression. mSphere 6 , e01245–20.33472984
68. Fiori A , and Van Dijck P (2012). Potent synergistic effect of doxycycline with fluconazole against Candida albicans is mediated by Interference with iron homeostasis. Antimicrob. Agents Chemother. 56 , 3785–3796.
69. Noble SM , and Johnson AD (2005). Strains and strategies for large-scale gene deletion studies of the diploid human fungal pathogen Candida albicans. Eukaryot. Cell 4 , 298–309.
70. O’Meara TR , O’Meara MJ , Polvi EJ , Pourhaghighi MR , Liston SD , Lin ZY , Veri AO , Emili A , Gingras AC , and Cowen LE (2019). Global proteomic analyses define an environmentally contingent Hsp90 interactome and reveal chaperone-dependent regulation of stress granule proteins and the R2TP complex in a fungal pathogen. PLoS Biol. 17 , e3000358.31283755
71. Gerami-Nejad M , Berman J , and Gale CA (2001). Cassettes for PCR-mediated construction of green, yellow, and cyan fluorescent protein fusions in Candida albicans. Yeast 18 , 859–864.11427968
72. Veri AO , Miao Z , Shapiro RS , Tebbji F , O’Meara TR , Kim SH , Colazo J , Tan K , Vyas VK , Whiteway M , (2018). Tuning Hsf1 levels drives distinct fungal morphogenetic programs with depletion impairing Hsp90 function and overexpression expanding the target space. PLoS Genet. 14 , e1007270.29590106
73. Zhang A , Petrov KO , Hyun ER , Liu Z , Gerber SA , and Myers LC (2012). The Tlo Proteins Are Stoichiometric Components of Candida albicans Mediator Anchored via the Med3 Subunit. Eukaryot. Cell 11 , 874–884.
74. Langmead B , and Salzberg SL (2012). Fast gapped-read alignment with Bowtie 2. Nat. Methods 9 , 357–359.22388286
75. Ritchie ME , Phipson B , Wu D , Hu Y , Law CW , Shi W , and Smyth GK (2015). limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43 , e47.25605792
76. Paysan-Lafosse T , Blum M , Chuguransky S , Grego T , Pinto BL , Salazar GA , Bileschi ML , Bork P , Bridge A , Colwell L , (2023). InterPro in 2022. Nucleic Acids Res. 51 , D418–D427.36350672
77. Camacho C , Coulouris G , Avagyan V , Ma N , Papadopoulos J , Bealer K , and Madden TL (2009). BLAST+: architecture and applications. BMC Bioinf. 10 , 421.
78. Basenko EY , Pulman JA , Shanmugasundram A , Harb OS , Crouch K , Starns D , Warrenfeltz S , Aurrecoechea C , Stoeckert CJ Jr. , Kissinger JC , (2018). FungiDB: An Integrated Bioinformatic Resource for Fungi and Oomycetes. J. Fungi 4 , 39.
79. Tung C-H , Huang J-W , and Yang J-M (2007). Kappa-alpha plot derived structural alphabet and BLOSUM-like substitution matrix for rapid search of protein structure database. Genome Biol. 8 , R31.17335583
80. Shannon P , Markiel A , Ozier O , Baliga NS , Wang JT , Ramage D , Amin N , Schwikowski B , and Ideker T (2003). Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 13 , 2498–2504.14597658
81. RStudio Team (2020). RStudio: Integrated Development Environment for R (Boston, MA: RStudio, PBC).
82. Craig R , Cortens JP , and Beavis RC (2004). Open source system for analyzing, validating, and storing protein identification data. J. Proteome Res. 3 , 1234–1242.15595733
83. Deutsch EW , Mendoza L , Shteynberg DD , Hoopmann MR , Sun Z , Eng JK , and Moritz RL (2023). Trans-Proteomic Pipeline: Robust Mass Spectrometry-Based Proteomics Data Analysis Suite. J. Proteome Res. 22 , 615–624.36648445
84. Gene Ontology Consortium; Aleksander SA , Balhoff J , Carbon S , Cherry JM , Drabkin HJ , Ebert D , Feuermann M , Gaudet P , Harris NL , (2023). The gene ontology knowledgebase in 2023. Genetics 224 , iyad031.36866529
85. Ashburner M , Ball CA , Blake JA , Botstein D , Butler H , Cherry JM , Davis AP , Dolinski K , Dwight SS , Eppig JT , (2000). Gene ontology: Tool for the unification of biology. The gene ontology consortium. Nat. Genet. 25 , 25–29.10802651
86. Min K , Ichikawa Y , Woolford CA , and Mitchell AP (2016). Candida albicans gene deletion with a transient CRISPR-Cas9 system. mSphere 1 , e00130–16.27340698
87. Perkins DN , Pappin DJ , Creasy DM , and Cottrell JS (1999). Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 20 , 3551–3567.10612281
88. Eng JK , Jahan TA , and Hoopmann MR (2013). Comet: An open-source MS/MS sequence database search tool. Proteomics 13 , 22–24.23148064
