Grf10 regulates the response to copper, iron, and phosphate in Candida albicans

Abstract The pathogenic yeast, Candida albicans, and other microbes must be able to handle drastic changes in nutrient availability within the human host. Copper, iron, and phosphate are essential micronutrients for microbes that are sequestered by the human host as nutritional immunity; yet high copper levels are employed by macrophages to induce toxic oxidative stress. Grf10 is a transcription factor important for regulating genes involved in morphogenesis (filamentation, chlamydospore formation) and metabolism (adenylate biosynthesis, 1-carbon metabolism). The grf10Δ mutant exhibited resistance to excess copper in a gene dosage-dependent manner but grew the same as the wild type in response to other metals (calcium, cobalt, iron, manganese, and zinc). Point mutations in the conserved residues D302 and E305, within a protein interaction region, conferred resistance to high copper and induced hyphal formation similar to strains with the null allele. The grf10Δ mutant misregulated genes involved with copper, iron, and phosphate uptake in YPD medium and mounted a normal transcriptional response to high copper. The mutant accumulated lower levels of magnesium and phosphorus, suggesting that copper resistance is linked to phosphate metabolism. Our results highlight new roles for Grf10 in copper and phosphate homeostasis in C. albicans and underscore the fundamental role of Grf10 in connecting these with cell survival.


Introduction
Candida albicans is a ubiquitous commensal of human microbiome and a major human fungal pathogen. In health and disease, C. albicans resides in and infects multiple niches, including the gastrointestinal tract, skin, mouth, and the female reproductive tract (Odds 1987). Morphological plasticity allows C. albicans to adapt to each niche; thus, the ability to change morphologies and physiology increases its survival and virulence (Noble et al. 2017). In addition to morphological plasticity, metabolic flexibility is essential for C. albicans during commensalism, infection, and disease progression. Versatile metabolic adaption of C. albicans allows the fungus to assimilate the available micronutrients that vary depending upon specific host niches and host immunological defense (Miramón and Lorenz 2017).
Complex transcriptional regulatory networks control morphological switching and metabolic adaptation in C. albicans (reviewed in (Brown et al. 2014;Noble et al. 2017). We previously characterized the transcription factor Grf10 as an important regulator of both morphogenesis and metabolism. Grf10 is necessary for filamentation and filamentation-related processes such as chlamydospore and biofilm formation and for virulence (Chauvel et al. 2012;Nobile et al. 2012;Romanowski et al. 2012;Ghosh et al. 2015;Rodriguez et al. 2020). Grf10 and Bas1 regulate expression of the ADE regulon, which includes purine nucleotide biosynthetic genes (ADE genes), a nucleoside permease-encoding gene (NUP), and 1-carbon metabolic genes which are important for yeast growth when limited for adenine (Homann et al. 2009;Wangsanut et al. 2017). Functional characterization demonstrated that Grf10 separately controls morphogenesis and adenine-regulated transactivation (Wangsanut et al. 2018). However, other genes regulated by Grf10 have not yet been identified.
Grf10 is linked to copper metabolism because the grf10Δ mutant is able to grow on high levels of copper (Homann et al. 2009), suggesting a role in copper uptake or storage. Transition metals such as copper and iron are nonorganic micronutrients that are essential to cell survival and infection processes (Gerwien et al. 2018). Through nutritional immunity, the host sequesters trace elements away from the microbe to reduce pathogenicity during infection, or alternatively the host overwhelms the fungal cells' ability to handle toxic levels of copper (reviewed in Hodgkinson and Petris 2012;Hood and Skaar 2012;Ding et al. 2014;Garcia-Santamarina and Thiele 2015;Besold et al. 2016;Mackie et al. 2016). Therefore, understanding the role of Grf10 in copper homeostasis can shed light on the host-pathogen interaction during the infection period and potentially lead to clinical significance.
Here, we focused on this copper resistance phenotype and found that growth rates were affected by gene dosage and that resistance did not extend to other metals. Prolonged incubation of the grf10Δ mutant under high copper levels triggered filamentous growth. We also showed that the conserved residues D302 and E305, previously shown to be necessary for activity by Grf10 during adenine limitation and filamentation (Wangsanut et al. 2018), are also critical for Grf10 to regulate copper toxicity. Examination of gene expression differences by RNA sequencing (RNA-seq) led to the identification of genes involved in copper, iron, and phosphate uptake and sequestration/storage. Using inductively coupled plasma mass spectrometry (ICP-MS), we found that the accumulation of copper and other metals in the grf10Δ mutant was no different from the wild type (WT) in low or high copper. However, we found lower accumulation of phosphorus by ICP-MS, in agreement with decreased expression of genes involved in phosphate uptake and polyphosphate synthesis. This study expands the repertoire of Grf10 target genes to include genes involved in copper, iron, and phosphate metabolism and further emphasizes that Grf10 controls multiple pathways in C. albicans.

C. albicans strains and culturing conditions
The strains of C. albicans used in this study are listed in Table 1. Strains DAY185 and DAY286 were obtained from A. Mitchell (Davis et al. 2000(Davis et al. , 2002, strain JC1928 was obtained from J. Quinn (Ikeh et al. 2016), and strains OHWT and TF021 were obtained from the Fungal Genetics Stock Center (Homann et al. 2009). RAC strain series were previously described (Ghosh et al. 2015;Wangsanut et al. 2017Wangsanut et al. , 2018. Evolved (EV) strains were derived from WT (DAY286) by selection on 10 mM CuSO 4 as papillae growth; 3 colonies from separate papillae were cultured in YPD and stored at −80°C.
All strains were started from frozen glycerol stocks on solid YPD medium (Sherman 1991), grown at 30°C for 2 days, and maintained at room temperature for 10 days. YPD was supplemented with 10-13 mM CuSO 4 as described in the text, selecting the optimum concentration under the different growth conditions to see the phenotype. Solid and liquid YPD were also supplemented with 200-1000 mM CaCl 2 , 1.0-2.5 mM CoCl 2 , 0.3-5 mM FeCl 3 at pH 4.0, 1.0-15 mM MnCl 2 , 0.5-5 mM ZnSO 4 , or 0.32 µg/ml spermidine, as indicated in the text (Ikeh et al. 2016). Synthetic dextrose (SD) medium with 10 mM potassium phosphate or lacking phosphate (Sherman 1991) was supplemented with 0.8 mg/ml arginine.

Spot assays
C. albicans strains were grown overnight in 5-ml YPD broth at 30°C and diluted to an OD 600 of 0.2 using sterile deionized water or were inoculated to 0.2 OD 600 and allowed to grow to log phase and then diluted to OD 600 of 0.2 as indicated in the figure legends. Samples were serially diluted 1:10 in sterile water, and 5 µl of each dilution was spotted onto solid medium, as indicated in each figure. Samples were incubated at 30°C and photographed at 16-18 h or daily for 7 days as indicated, using an ImageQuant Imager. At least 3 biological replicates were performed for each spot assay.

Growth rate and doubling time determination
C. albicans strains were grown overnight at 30°C in 5-ml YPD broth and inoculated into 96-well plate containing 200 µl of YPD or YPD supplemented with the specified concentration of metal/cation at an OD 600 of 0.01 (Schwartz et al. 2013). The cultures were grown at 30°C with orbital shaking in a thermo-controlled GloMax plate reader. The OD 600 of each well was measured every 30 min for 24 or 48 h, and background OD 600 absorption from wells containing sterile supplemented media was subtracted from each sample. Doubling times for each strain at each growth condition were calculated from the growth curve, and standard deviation was calculated in Excel. Three biological replicates with 3 technical replicates were measured and averaged for each C. albicans strain and growth medium.

RNA-seq and data analysis
Two separate RNA-seq experiments were performed, both using 3 cultures of each C. albicans strains WT (OHWT) and grf10Δ (TF021) grown in 30°C. The first investigated the differences in gene expression between 1 and 4 h (#1) and the second between normal and excess copper conditions (#2).

Growth conditions
(#1) Strains were grown in YPD broth at 30°C for 1 and 4 h and harvested as described (Kadosh and Johnson 2005;Bruno et al. 2010). Briefly, a saturated overnight culture (OD 600 of ∼10) in YPD medium was used to inoculate a 1:10 dilution for 1-h cultures or a 1:50 dilution for 4-h cultures, into 5 ml of fresh YPD medium, and cultures were grown in a tube roller at 30°C in YPD medium. Cultures were quickly chilled in an ice water bath, cells were pelleted by centrifugation, and pellets were immediately stored at −80°C. (#2) For the high copper experiment, we followed the experimental design as described (Schwartz et al. 2013). Strains were grown overnight in 5-ml YPD at 30°C, inoculated into 100 ml of YDP medium at an OD 600 of 0.2, and grown at 30°C with orbital shaking until an OD 600 of ∼1 (∼4 h). The 100-ml cultures were split into 2 50-ml portions, 12 mM CuSO 4 was added to one of the cultures, and both were grown at 30°C for 30 min; the cells were harvested by pelleting for 10 min, and stored at −80°C until RNA extraction.

RNA extraction, library preparation, and RNA-sequencing
The RNA-seq data set has been deposited at Gene Expression Omnibus (Edgar et al. 2002;Barrett et al. 2013) with the SuperSeries record number GSE223218 (Wangsanut et al. 2023).
(#1) RNA was extracted from frozen pellets using the RiboPure Yeast Kit (Ambion). Copy DNA (cDNA) libraries were prepared from 1 µg of DNase I-treated RNA sample using TruSeq Stranded mRNA Sample Prep Kit, following the manufacturer's directions (Illumina, RS-122-2101 Set A). RNA samples were checked for quality using an Agilent Bioanalyzer, and DNA concentration of the cDNA libraries was determined using Qubit. The cDNA libraries (10 nM) were prepared for sequencing using the MiSeq Reagent Kit, following the manufacturer's instructions (Illumina,, and sequenced on a MiSeq instrument (Illumina). The raw FASTQ files were submitted to NCBI, accession number GSE223216. (#2) RNA was extracted using the RNeasy Plus Mini Kit (Qiagen) with an additional mechanical lysis using MP Bio Fast Prep-24 in 1.5-ml screw cap tubes containing RNase-free acid-washed beads and RLT Plus Buffer (Qiagen). Samples were shaken at 4 m/s for 40 s and then put on ice for 60 s, repeated 4 times. Samples were centrifuged at top speed in a microfuge for 10 min to remove beads and cellular debris. The aqueous phase was transferred to gDNA eliminator column (RNeasy kit), following the manufacturer's instructions. RNA sample quality was tested using the Bioanalyzer High Sensitivity RNA Analysis Pico Kit (Agilent) and an Agilent 2100 Bioanalyzer. Samples were sent to Novogene (Sacramento, CA, USA) for cDNA synthesis and sequencing. The raw FASTQ files were submitted to NCBI, accession number GSE223217.

Reference genome
Parental directory C_albicans_SC5314 Assembly 22 found on CGD website (candidagenome.org; Skrzypek et al. 2017) was used as a reference genome for both RNA-seq experiments.

Bioinformatics analysis
Using Geneious software, version Geneious Prime, FASTQ files of the single-end reads were paired, trimmed using the BBDuk plugin, and mapped to the reference genome. Raw reads from the 2 alleles were combined, and expression differences between the WT (OHWT) and mutant (TF021) were determined using the DESeq2 (v1.30.1) algorithm (Love et al. 2014) in R (v4.0.2). A cut-off of 10 reads total across the compared samples was the threshold for a gene to be included in the data set, and the apeglm algorithm was used as the common LFC shrinkage option to minimize variability of lowly expressed genes (Zhu et al. 2018). The P-values provided were adjusted with a Benjamini-Hochberg false discovery rate. Statistically significant differentially expressed genes had a P-value of 0.05 or less and a log2 fold change (up or down) of 0.58 or greater.

qRT-PCR analysis
Quantitative real-time polymerase chain reaction (qRT-PCR) assays were performed on the same samples as for RNA-seq experiment (Ghosh et al. 2015). Cell collection and RNA extraction was performed as listed above. Primers used for PHO100, FRE7, FRE30, FET31, FTR1, CFL2, OPT1, and TNA1 are listed in Table 2, and primers for ADE13 and NUP and the reference gene TEF1 were described (Wangsanut et al. 2017). The P-values for qRT-PCR were calculated using Student's t-test function in Excel. Pearson correlation coefficient (r) comparing relative fold changes obtained from RNA-Seq and qRT-PCR was calculated in Excel.

Inductively coupled plasma mass spectrometry
Samples were prepared for ICP-MS analysis as described (Rosenfeld et al. 2010). Three cultures of each strain were grown overnight at 30°C in YPD and diluted to OD 600 = 0.1 in 20 ml of YPD or YPD containing 13 mM CuSO 4 (Fig. 7) or in 20 ml of SD + Arg with and without 10 mM potassium phosphate (Fig. 8), as described (Ikeh et al. 2016). Cultures were placed in a 30°C shaker and grown to OD 600 of ∼1.0. A sample volume corresponding to 20 OD 600 units was pelleted by centrifugation, washed 3 times with 10 mM Tris-HCl, pH 8.0, 1 mM EDTA, and pellets were digested overnight in 1 ml of 70% nitric acid at 95°. Cell debris was removed twice via centrifugation at 18,000 × g, and the clarified solution was diluted to 2% nitric acid using Milli-Q water. ICP-MS analysis was performed using an Agilent 7700 Series ICP-MS ( Fig. 7) or an Agilent 7800 Series ICP-MS ( Fig. 8) according to manufacturer specifications. The ICP-MS standards for calcium, cobalt, copper, iron, magnesium, manganese, phosphate, potassium, sodium, and zinc were purchased from Fluka Analytical and were diluted in 5% nitric acid. Three biological samples were each measured in 3 technical replicates. The counts per second output for the samples were converted to nmol/10 9 cells after comparison with the standard curves. Averages from the biological and technical replicates were determined, and the standard deviations are presented in the graphs.

Whole-genome sequencing and analysis
Strains were inoculated from single colonies into 5 ml of YPD medium and grown overnight at 30°C. Cells were pelleted in a microfuge for 10 min, resuspended in 180-µl lysis buffer (1% SDS, 0.1 M NaCl, 10 mM Tris pH 8.0, 1 mM EDTA, and 2% Triton X-100), and transferred to a 1.5-ml screw cap tube containing ∼200 µl of 0.1-µm glass beads and 200 µl of phenol:chloroform:isopropanol (25:24:1). Samples were shaken in a MP-Fast Prep-24 at 4 m/s for 25 s at 4°C; 200 µl of 10 mM Tris-HCl, pH 8.0, 1 mM EDTA, was added; and samples were centrifuged for 10 min at top speed in a microfuge. The (top) aqueous layer was transferred to a new Eppendorf tube. Ethanol (1 ml of 100%) was added to the sample, mixed by inversion 10 times, and the DNA was pelleted for 10 min in a microfuge. The DNA pellet was resuspended in 400 µl of 10 mM Tris-HCl, and pH 7.5, 1 mM EDTA. RNase was added to 7.5 µg/ml, and samples were incubated for 45 min at 37°C. To precipitate the DNA, ammonium acetate was added to 100 mM, and ethanol was added to a final concentration of 70% prior to mixing. The samples were placed at −20°C for 30 min followed by centrifugation for 10 min. The pellets were washed in 1 ml of ice-cold 70% ethanol, and the DNA pellets were air-dried for 30 min. The pellets were each resuspended in 50 µl of sterile, molecular biology grade water.
DNA samples (of 6 µg or more) were sent to the SeqCenter (Pittsburgh, PA, USA) for whole-genome DNA sequencing using the Illumina 400 Mbp (2.67 M reads) Sequencing Package. The DNA sequencing was returned as fastq.gz files and also included the total reads, total read pairs, total base pair quality score, percent base pair coverage, and md5sum scores. Paired end reads were mapped in Y MAP (Abbey et al. 2014) using reference genome SC5314 (var. A21-s02-m09-r07), and corrections, GC content bias and chromosome-end bias, were applied, as described (Selmecki et al. 2006(Selmecki et al. , 2015.

Results
The grf10Δ mutant shows resistance to copper toxicity and exhibits a gene dosage effect Strains that lack the transcription factor Grf10 exhibit resistance to excess copper (10 mM CuSO 4 ) (Homann et al. 2009). We note that these excess copper conditions represent toxic concentrations; nonetheless, we were interested in investigating this phenotype. We extended this observation by examining resistance in homozygous-null and heterozygous-restored strains (i.e. strains returned to heterozygosity by reintroducing 1 allele) in both the BWP17 and SN152 backgrounds (Fig. 1a). We confirmed that the grf10Δ mutant exhibits resistance as compared to the isogenic WT when presented with excess copper, and this phenotype is stronger in the BWP17 background than in SN152. To quantify the effect of copper on growth, we measured the doubling times of each strain growing in YPD liquid medium with or without the addition of 12 mM CuSO 4 (Fig. 1b). The doubling times were the same for all of the strains grown in YPD (∼1.14 h); however, all of the strains grown in excess copper exhibited an increase in doubling time. The doubling time for the WT strain (DAY286) in the presence of high copper was 12.52 ± 2.02 h, growing significantly more slowly, and cell growth ceased, plateauing at an OD 600 of ∼0.2. Strikingly, the grf10Δ mutant grew much better than the WT strain in the presence of copper and showed a gene dosage effect (Fig. 1b); the heterozygous-restored strain RAC120 doubled at ∼5 h (5.35 ± 2.67 h), and the null mutant RAC117 doubled ∼1.8 h (1.81 ± 0.09 h), which is only 1.6-times slower than when copper was not present. We found similar results with the strains in the SN152 background (Fig. 1c). These results suggest that the levels of Grf10 may be important for the expression of genes associated with copper uptake or sequestration.

The metal resistance is specific to copper
It is possible that the strong copper resistance phenotype of the grf10Δ mutant was due to a general defect in overall metal homeostasis. To test this, we assessed growth of the WT, grf10Δ mutant, and the heterozygous-restored strains on solid YPD medium supplemented with 450 mM calcium chloride, 5 mM iron chloride, 10 mM manganese chloride, 1 mM zinc sulfate, and 0.32 µg/ml spermidine (Fig. 2a). Spermidine is an organic cation that was used to determine whether this resistance effect was due to overall positive charge, as opposed to being specific to metals (Ikeh et al. 2016). There was no difference in colony growth with any of these cations.
To test a broader range of concentrations and to ensure that we could detect growth differences, we assessed growth of the WT, grf10Δ mutant, and restored strains from both strain backgrounds in liquid medium supplemented with various metals. Figure 2b shows the growth curves for the SN152 background at 1 concentration for each metal tested; we supplemented YPD with 200 mM CaCl 2 , 1.5 mM CoCl 2 , 1 mM FeCl 3 (at pH 4), 12 mM MnCl 2 , and 4 mM ZnSO 4 . The growth curves for the full range of concentrations tested and in both strains are found in Supplementary Figs. 1-5. All strains grew worse in the presence of the tested metal than in its absence, with slower doubling times and reaching stationary phase at a lower OD 600 . Importantly, all 3 strains responded in the same way to each addition, indicating that there was no general defect in metal homeostasis. Thus, the phenotype is specific to copper, suggesting that Grf10 is involved in expression of genes important for copper metabolism.

Copper toxicity response is dependent on Grf10 conserved residues D302 and E305
Grf10 contains a conserved interaction region, located from amino acids 270-353, that is important for it to activate expression, likely due to a co-regulator interaction (Wangsanut et al. 2017;Wangsanut et al. 2018). The grf10-D302A mutant is completely defective in adenine regulation and in the response to filamentation, and the grf10-E305A mutant is partially defective for filamentation and is normal for adenine regulation (Wangsanut et al. 2018); neither mutation affects protein stability. We examined whether conserved residues D302 and E305 are important for the copper sensitivity of the GRF10 strain.
To test this, we used previously constructed strains, grf10-D302A or grf10-E305A, which contain 1 allele of GRF10 with either the D302A or E305A point mutation. We grew these strains, the isogenic WT, the grf10Δ mutant, and the restored strain GRF10R on solid YPD medium with or without 10 mM copper sulfate (Fig. 3a). We found that both of the mutant grf10 alleles completely failed to reverse the copper sensitivity; in other words, the substitutions behaved as a nonfunctioning null allele. This result indicated that regulation of copper toxicity is dependent on both D302 and E305 of Grf10 and suggested that these residues mediate the interaction between Grf10 and a specific protein partner in response to excess copper.
We observed that with prolonged incubation on YPD containing copper sulfate (30°C for 7 days), peripheral hyphae were produced from the colonies that expressed the defective alleles of GRF10-the grf10Δ, grf10-D302, and grf10-E305-whereas the WT counterpart and the heterozygous-restored strains did not form hyphae (Fig. 3b). This intriguing result suggests that Grf10 normally inhibits hyphal formation when the cell encounters toxic levels of copper.

Evolution of WT demonstrates copper resistance when grown in excess copper
The WT colony on YPD + 10 mM CuSO 4 at 7 days of incubation had an irregular colony morphology as compared to growth on YPD (Fig. 3b). We observed papillae growing above the background at earlier times (see Fig. 3a and Supplementary Fig. 6). Furthermore, we noted that the WT strain had late growth in liquid YPD + CuSO 4 medium ( Fig. 1b and c). We hypothesized that the papilla were individual yeast cells that had mutated under selection to resist the high levels of copper; those colonies continued to grow and eventually merged over the week to become the , and heterozygous-restored strains (RAC256, RAC120) were serially diluted and spotted on solid YPD medium with or without 10 mM copper and incubated at 30°C for 16 h. b and c) BWP17 and SN152 background growth (OD 600 ) of the same cultures as in (a) was followed in YPD or YPD with 12 mM CuSO 4 for 24 h. Graphs show the average OD 600 reading from 3 biological replicates. See labels on graph; key: YPD: wild type in yellow; grf10Δ, light blue; restored, green; YPD + copper, wild type in dark blue; grf10Δ, orange; restored, gray.
lumpy colony (Fig. 3b). If so, the trait of growth in high copper should be stable due to these mutations.
To test this, we selected several papillae from the growth of the WT strain on YPD + 12 mM CuSO 4 and grew them on solid YPD medium-i.e. growth conditions without high copper selection. Three of these "evolved" strains and 3 colonies from the original (nonselected) WT strain were used to inoculate cultures with and without high copper, and growth was measured in liquid culture over time (Fig. 4a). We found that the EV strains (EV1-3) were able to grow immediately, whereas the WT cultures grown in the presence of excess copper exhibited a growth delay, similarly to what was seen in Fig. 1.
The most likely explanation for the EV phenotype is aneuploidy with increased copy number of genes important to the copper selection (Berman 2016). To examine this, we performed wholegenome sequencing on the parental WT (DAY286) and 2 EV strains, EV1 and EV2. As shown in Fig. 4b, the EV strains EV1 and EV2 showed increased aneuploidy of chromosome 5; EV2 also exhibited aneuploidy of chromosomes 1 and 6. Chromosome 5 is 1.19 kb long and carries 523 ORFs (Skrzypek et al. 2017); 2 of the genes located on chromosome 5 and related to copper homeostasis are CUP1 that encodes the copper metallothionein and CCC2 that encodes a metallochaperone. These genes may have driven selection for increased copy number.

Identification of global GRF10-dependent genes
To gain insights into misregulated genes involved in copper metabolism, we analyzed an RNA-seq data set comparing differential gene expression from the grf10Δ mutant (TF021) and isogenic WT (OHWT) strains grown in YPD medium for 1 and 4 h. Grf10-dependent genes were identified as those in which the expression was altered by at least 1.5-fold (log2 fold change = 0.58) and an adjusted P-value of 0.05 or lower (Tables 3 and 4; the complete data sets are found in Supplementary Tables 1 and 2).
As expected, we saw a significant difference in expression for GRF10 and for marker genes LEU2 and HIS1 integrated at GRF10 rather than their native loci. Of the remaining genes, 52 genes showed differential expression in the grf10Δ mutant at 1 h following inoculation (43 genes with decreased expression and 9 genes with increased expression), and 82 showed differential expression in the grf10Δ mutant at 4 h (60 genes with decreased and 22 genes with increased expression); see volcano plots in Fig. 5a and 5b and GO term analyses in Supplementary Tables 3 and 4. Twelve of these genes were in common between the 2 time points, and half of these are involved in phosphate metabolism and ion transport (FET31, IRO1, PHO84, PHO87, VTC3, and VTC4). Finally, we verified the expression data using qRT-PCR for 10 genes: PHO100, FRE7, FRE30, FET31, FTR1, CFL2, OPT1, TNA1, ADE13, and NUP (Fig. 5c); expression differences were correlated between the 2 methods (Pearson correlation coefficient = 0.983).

Gene expression in the grf10Δ strain shows a largely WT response to a copper challenge
To investigate the responses to copper in the grf10Δ mutant and WT strains, we examined differential gene expression after challenging cells with 12 mM copper sulfate using RNA-seq. The response to copper altered the expression of 1,283 in WT cells (adjusted P-value < 0.05 and log2 fold change of >0.58) and 1,164 in the grf10Δ mutant (Supplementary Tables 5 and 6, respectively). There was no significant difference in the expression of GFR10 in WT cells upon copper treatment (Supplementary Table 5).
Given the copper resistance phenotype, we looked closely at the expression of genes known to be involved in copper, iron, and superoxide metabolism (Askwith et al. 1994;Morrissey et al. 1996;Chibana et al. 2005;Douglas et al. 2011;Cheng et al. 2013;Garcia-Santamarina and Thiele 2015;Li et al. 2015;Smith et al. 2017;Gerwien et al. 2018;Khemiri et al. 2020). The grf10Δ mutant strain responded to the copper challenge nearly the same as the WT strain (Fig. 6), upregulating key survival genes such as the CUP1 metallothionein, the ATX1 copper metallochaperone,  Supplementary Figs. 1-5). Graphs show the average OD 600 from 3 biological replicates. The key is the same as in Fig. 1 (legend is bottom right corner).
the CRP1 copper extrusion pump, and the SOD1 superoxide dismutase, as well as repressing the copper transporter CTR1. Three genes showed significant differential expression in the 2 strains in high copper: the iron utilization gene IRO1 was expressed at 36% of the WT, the multicopper oxidase gene FET31 was expressed at about half the level of the WT, and the intracellular copper transporter gene CCC2 was expressed 1.6-fold higher. All of the genes showing differential expression in the grf10Δ mutant in high copper are listed in Supplementary Table 7).
Beyond the genes noted above, 224 genes are misregulated in the grf10Δ mutant under high copper conditions (110 genes have increased expression and 114 with decreased expression) (Supplementary Table 7). Not surprisingly, genes with lower expression included those involved in adenylate and 1-carbon metabolism as well as 2 genes (SNZ1 and SNO1) involved in pyridoxal phosphate biosynthesis. The gene showing the greatest increase in expression in the mutant was CHA1, which encodes a Ser/Thr dehydratase and is induced in low iron (Lan et al. 2004).

ICP-MS reveals similar intracellular copper accumulation in mutant and WT cells
Given the difference in expression of copper and iron transporters (Fig. 5) as well as the growth resistance to copper (Fig. 1), we asked whether the grf10Δ mutant strain would accumulate lower levels of copper. We determined the intracellular metal accumulation using ICP-MS. We assessed the levels of copper and 8 additional elements-calcium, cobalt, iron, magnesium, manganese, potassium, sodium, and zinc (Rosenfeld et al. 2010) (Fig. 7). Under normal YPD culturing conditions, we found no difference in the intracellular copper accumulation in the grf10Δ mutant and WT strains, with both strains accumulating ∼20 nmol per billion cells. Magnesium accumulated to 57% of the WT levels in the grf10Δ mutant under normal growth conditions (P = 0.02). There were no changes in the steady-state levels of the other metals examined (Fig. 7).
When cells were grown in YPD with 13 mM copper sulfate added, the WT and grf10Δ mutant strains had no significant difference in copper accumulation (P = 0.08). Cobalt levels increased in a pattern that was similar to that of copper, higher in the presence of copper in both strains (P ≤ 0.02), but not significantly different between mutant and WT (P = 0.08). We detected lower levels of iron, magnesium, potassium, and zinc (P ≤ 0.02) in both the grf10-E305A

(a)
YPD YPD + 10 mM CuSOИ Fig. 3. Amino acids D302 and E305 of Grf10 are critical for copper sensitivity and morphology. a) Overnight cultures of the WT strain (DAY286), grf10Δ mutant (RAC117), heterozygous-restored strains expressing GRF10 (RAC120), or GRF10 with substitution mutations, grf10-D302A (RAC259) and grf10-E305A (RAC260), were spotted onto YPD with or without 10 mM CuSO 4 and grown as in Fig. 1. b) The same plates from (a) were cultured for 7 days at 30°C and photos were taken of a single whole colony or edges of a colony.  . 4. Evolution of WT demonstrates copper resistance when grown in excess copper and contains chromosomal abnormalities. a) Overnight cultures of the WT strain (DAY286) and putatively force evolved papilla (EV; originating from DAY286) were grown overnight in 5 ml of YPD, normalized to OD 600 of 0.2, and grown in YPD or YPD with 12 mM CuSO 4 . Three biological replicates of each WT (squares of gray, blue and red) and EV (circles of brown, navy and green) strains were used for the YPD plus copper addition, and only 1 biological replicate was used for the WT (blue triangle) and EV (orange triangle) strains for the YPD control. b) YMAP plots illustrating genome-wide changes occurring in EV1 and EV2 strains as compared to the parental WT (DAY286) strain. The gray lines indicate equal representation of both alleles (A and B) of each gene across the chromosome, and the aqua and pink indicate previously described loss-of-heterozygosity events on chromsomes 2 and 3. In the evolved strains (EV1, EV2), the blue regions and allele counts above the diploid levels of 2 indicate aneuploidy (trisomy and tetrasomy) within the population. mutant and WT strains (Fig. 7). Given this, we conclude that the WT and the grf10Δ mutant do not differ in their accumulation of intracellular metals. Furthermore, these results indicate that the grf10Δ mutant and WT cells accumulate the same amount of copper in spite of its copper toxicity resistance phenotype.

The grf10Δ mutant accumulates lower levels of phosphate
The identification that many genes in the PHO regulon showed lower expression in the grf10Δ mutant led us to hypothesize that   Table 4. Grf10-dependent genes at 4 h.

Down or up in grf10Δ
Gene Locus Average Log2 change

Classification and comments
Decreased HGT17 phosphate metabolism is affected. We measured differences in phosphate accumulation using ICP-MS in the WT, grf10Δ and pho4Δ strains when cells were cultured in phosphate replete (10 mM potassium phosphate) and limiting (no addition) media.
In addition to phosphorus, we examined accumulation of magnesium and iron (see Fig. 8, and Rosenfeld et al. 2010;Ikeh et al. 2016).
The grf10Δ mutant accumulated ∼75% of phosphorus relative to the WT (P = 0.03) when the growth medium was supplemented with phosphate (Fig. 8). The pho4Δ mutant accumulated only about 30% of the WT levels of phosphorus (P = 0.006) under these conditions. When limited for phosphate, all 3 strains showed lower accumulation; only the pho4Δ was significantly different from the WT (P = 6.9 × 10 −19 ). Magnesium, the typical counterion to phosphate, also showed a similar pattern as phosphate. The grf10 strain accumulated magnesium to 82% of the WT (P = 0.05), and the pho4 strain accumulated it to 54% of WT (P = 0.001).
We found no significant difference in the accumulation of iron between any of the 3 strains (Fig. 8). These results of lower phosphorus accumulation in the grf10Δ mutant are consistent with the decrease in PHO and VTC gene expression. Given this, it is possible that Grf10 and Pho4 transcription factors interact to regulate PHO genes.

Discussion
Grf10 is a homeodomain transcription that plays roles in developmental processes such as morphogenesis, filamentation, biofilm formation, and white-opaque switching as well as in biosynthetic pathways such as adenylate synthesis and 1-carbon metabolism (Ghosh et al. 2015;Wangsanut et al. 2017Wangsanut et al. , 2018Rodriguez et al. 2020;Qasim et al. 2021). In this study, we examined the copper resistance phenotype, first described by Homann et al. (2009). We extended the initial observation, examined metal accumulation by ICP-MS, and identified novel Grf10-dependent genes involved in the uptake, sequestration, and storage of copper, iron, and phosphate. Resistance was restricted to high levels of copper, and it did not extend to other metals such as iron, calcium, manganese, or zinc, leading us to focus on copper uptake and/or sequestration as a mechanism for resistance. The grf10Δ mutant did not accumulate less copper than the WT strain, suggesting that the bioavailability of copper  Adjusted p-value (negative log10)

Log2 Fold Change
Differential Gene Expression at 4-hr Fig. 5. Differential gene expression in the grf10Δ mutant. Volcano plots (Log2 differential expression vs. absolute value of the confidence score) showing both alleles of the genes. Differentially expressed genes were defined as those with a confidence score >6. Alleles of the genes associated with processes discussed in the text are denoted as red squares for copper, iron, and phosphate metabolism; blue diamonds for purine nucleotide and 1-carbon metabolism; and gray circles for the remainder. a) 1-h post-inoculation. b) 4-h post-inoculation. Note, HIS1 is not shown (point at −1.9, 101). c) Correlation between RNA-seq and qRT-PCR experiments. is different between the strains. The copper specific toxicity response was dependent on conserved residues D302 and E305 of Grf10 (Wangsanut et al. 2018). The Grf10 residues D302 and E305 are in protein interaction region that has been conserved through fungal diversification (Wangsanut et al. 2018). In Saccharomyces cerevisiae, this region of ScPho2 is necessary for interactions with 3 co-regulators: ScPho4 for regulation of phosphate uptake and utilization, ScBas1 for regulation of adenylate and 1-carbon metabolism, and ScSiw5 for mating type switching (Vogel et al. 1989;Daignan-Fornier and Fink 1992;Brazas et al. 1995;Barbaric et al. 1996;Magbanua et al. 1997;Zhang et al. 1997;Pinson et al. 2000;Bhoite et al. 2002;Hannum et al. 2002;Som et al. 2005). This same region is required for Grf10 to control the response to copper as well as filamentation and adenine prototrophy ( Fig. 3 and Wangsanut et al. 2018). While the D302A mutation was defective in all phenotypes tested, the E305A mutation exhibited variable phenotypes-completely defective for copper resistance, partially defective for filamentation, and normal for ADE prototrophy-consistent with a model that Grf10 regulates these phenotypes with different co-regulators (Wangsanut et al. 2017(Wangsanut et al. , 2018. We hypothesize that Grf10 forms a ternary complex consisting of Grf10, a protein partner, and DNA to regulate genes involved in multiple cellular processes including adenine biosynthesis and 1-carbon metabolism, filamentation, and metal homeostasis. Mac1 and Cup2 are the characterized transcription factors that regulate copper responsive genes during copper depletion and copper excess, respectively (Marvin et al. 2004;Schwartz et al. 2013;Smith et al. 2017). Grf10 likely works separately from Mac1 and Cup2 because deletions of each these genes show distinct copper phenotypes. We did not see any changes in the levels of MAC1 or CUP2 transcripts comparing the WT and grf10Δ mutant. Unlike the grf10Δ mutant that exhibits a resistance to excess copper, the mac1Δ mutant shows growth sensitivity under copper limitation, and the cup2Δ mutant shows growth sensitivity under copper excess (Homann et al. 2009). An extensive phenotypic analysis of transcription factor mutants revealed that several transcription factors conferred a similar copper resistance phenotype to that seen in the grf10Δ mutant: rlm1Δ, crz2Δ, pho4Δ, cph2Δ, wor2Δ, orf19.2961Δ, isw2Δ, hap31Δ, efg1Δ, hap2Δ, and hap5Δ strains (Homann et al. 2009). Grf10 could interact with 1 or more of these transcription factors to regulate the copper response.
Grf10 may play an important role in linking the response to copper with the hyphal developmental program by working downstream of or with Efg1 in Ras/cAMP/PKA pathway. The G-protein subunit Gpa2, cAMP, Efg1, and Grf10 are each required for both filamentation and the copper toxicity response (this study and Maidan et al. 2005;Homann et al. 2009;Schwartz et al. 2013;Ghosh et al. 2015;Azadmanesh et al. 2017). Likewise, Marvin et al. (2003) demonstrated that perturbation of copper acquisition due to mutations in the copper transporter gene CTR1 promoted filamentation in C. albicans, consistent with our observation that filaments were produced in the grf10Δ mutant after prolonged incubation with excess copper. Importantly, Efg1 induces both GRF10 expression during biofilm formation, possibly also during hyphal formation, and expression of copper and iron uptake genes (Stichternoth and Ernst 2009;Nobile et al. 2012;Rodriguez et al. 2020). Thus, Grf10 may work with Efg1 through the Ras/cAMP/ PKA pathway to coordinate a copper response with hyphal growth.

Grf10 may be involved in phosphate regulation
The identification of misregulation of genes involved in phosphate uptake and polyphosphate formation was unexpected. Several groups have reported that the grf10Δ mutant has no growth phenotype on media with limiting phosphate (Homann et al. 2009;Kerwin and Wykoff 2009;Wangsanut et al. 2017 . 8. The grf10Δ and pho4Δ mutants show differences in the intracellular accumulation of phosphorus and magnesium. WT (OHWT; derived from SN152), grf10Δ (TF021), and pho4Δ (JC1928; isogenic WT is SN152) strains were grown in SD + arginine or SD + arginine supplemented with 10 mM phosphate. Samples were prepared for ICP-MS as described in Fig. 7. Solid black bars indicate samples from phosphate-supplemented medium, and gray bars indicate samples lacking phosphate. Bars represent the averages from 3 independent cultures and 3 technical replicates, and the error bars indicate standard deviations. Statistical analysis was conducted using Student's t-test for comparisons between WT, grf10Δ, and pho4Δ strains, and significant P-values are indicated above the graphs. These data were collected using a different set of samples and different ICP-MS instrument from the data found in Fig. 7.
only CgPho4; therefore, as the ascomycetes diverged, the PHO pathway was rewired, and S. cerevisiae has the novel dependence on a second transcription factor (Kerwin and Wykoff 2009). Thus, the dependence on Grf10 for expression of PHO genes in C. albicans was unexpected.
On the other hand, there are connections between phosphate metabolism and metal homeostasis. We and others found that both the pho4Δ and grf10Δ mutants show copper resistance (Homann et al. 2009;Ikeh et al. 2016). In C. albicans, C. glabrata, and Cryptococcus neoformans, the transcription factor Pho4 regulates the PHO regulon during phosphate limitation, and genes involved with nutrient transporters and metal homeostasis under phosphate replete conditions (Toh-e et al. 2015;Ikeh et al. 2016;He et al. 2017;Lev and Djordjevic 2018). It has been suggested that an expanded set of Pho4 targets, beyond the PHO regulon, could have evolved from a reduced requirement for the co-activator Grf10/Pho2 (Kerwin and Wykoff 2009;He et al. 2017). Thus, our results highlight the need to reconsider the relationship between Grf10 and Pho4 in C. albicans.

Summary
We have expanded the understanding of Grf10-dependent target genes to multiple pathways: purine biosynthesis and uptake, yeast to hyphae switching, copper and iron response, and uptake of inorganic phosphate. Many genes in these pathways are required for virulence in murine models of C. albicans infection (Ramanan and Wang 2000;Donovan et al. 2001;Jezewski et al. 2007;Carlisle et al. 2009;MacCallum et al. 2009;Correia et al. 2010;Jiang et al. 2010;Mackie et al. 2016). It is possible that the combination of effects on these multiple pathways in the grf10Δ mutant led to strong attenuated virulence in mouse models of infection (Romanowski et al. 2012;Ghosh et al. 2015). The involvement of Grf10 in the copper response is novel, as this phenotype had not been reported in S. cerevisiae, providing new exploratory areas for orthologs of Grf10 in other fungal species. Since copper and iron availability lie at the center of the host-pathogen interface, Grf10 may play an essential role in sensing host nutrient levels and governing defenses against host nutrient immunity. Overall, our results add to the emerging concept that metabolism, morphogenesis, and nutritional immunity are intricately linked in C. albicans and that Grf10 participates in this transcriptional coordination.

Data availability
Strains are listed in Table 1 and are available upon request. The RNA-seq data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus (Edgar et al. 2002) and are accessible through GEO Series accession number GSE223218 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSExxx). Supplementary materials are available at figshare (https://doi.org/ 10.25387/g3.14597676): Supplementary Tables 1 and 2 contain differential gene expression genome-wide at 1-and 4-h postinoculation, respectively; Supplementary Tables 3 and 4 contain  GO term analyses; and Supplementary Tables 5-7 contain differential  gene  expression  in  high  copper conditions. Supplementary Figs. 1-5 contain growth data from cells challenged with FeCl 3 , CaCl 2 , MnCl 2 , ZnSO 4 , and CoCl 2 , respectively, over a range of concentrations; and Supplementary Fig. 6 shows papillae growth over time. The authors affirm that all data necessary for confirming the conclusions of the article are present within the article, figures, and tables.