Characterization of endoplasmic reticulum-associated degradation in the human fungal pathogen Candida albicans

Generation of mutants

Mutations to Ca_HRD1, Ca_DOA10, and Ca_UBC7 were made using CRISPR-mediated genome editing. Guide RNAs to the 5′ region of Ca_HRD1, Ca_DOA10, and Ca_UBC7 were cloned into plasmid pV1093 (Vyas, Barrasa & Fink, 2015), which also contains Cas9 as well as nourseothricin (Nat) resistance marker (Nat^R). Repair templates that introduce a stop codon and restriction site upon homologous recombination with the cut chromosome were generated. C. albicans wild type strain SC5314 was transformed with Cas9 guide constructs and repair templates using lithium acetate. Transformants were selected on 200 µg/ml Nat. Correct homozygous transformants were identified by colony PCR followed by restriction digestion. Sanger sequencing was used to confirm the mutants. Homozygous mutants will be henceforth referred to as Ca_hrd1, Ca_doa10, and Ca_ubc7.

Growth analyses

Spot assays: Yeast growth assays were performed using a modified version of a previously published protocol (Watts et al., 2015). Yeast were grown to saturation overnight at 30 °C in yeast extract-peptone-dextrose media (YPD). Cells were diluted to an OD₆₀₀ of 1.0. Fivefold dilutions were prepared in a 96-well plate in YPD. Cells were plated using a pin replicator onto YPD agar lacking or containing increasing concentrations of hygromycin B (Brodersen et al., 2000; Ganoza & Kiel, 2001). Cells were grown at 30 and 37 °C for 5 days and at 25 and 40 °C for 10 days.

Inhibition assays: A modified drug susceptibility assay (Bauer et al., 1966) was performed. Yeast were grown to saturation overnight at 30 °C in YPD. 1 × 10⁷ cells were spread on 20 ml of synthetic complete media in 150 mM plates and were allowed to dry. A sterile filter paper disc was placed on the cells, and 15 µl of 1 M dithiothreitol (DTT) suspended in water, 33% (v/v) β-mercaptoethanol (βME) in water (Jia et al., 2019), or sterile water were directly pipetted onto the paper disc. Plates were incubated for 3 days at 37 °C before visualization.

Mass spectrometry and data analysis

Sample preparation, mass spectrometry analysis, bioinformatics, and data evaluation for quantitative proteomics experiments were performed in collaboration with the Indiana University Proteomics Center for Proteome Analysis at the Indiana University School of Medicine (IUSM) similarly to previously published protocols (Kumar et al., 2022; Morris et al., 2022; Soundararajan et al., 2022; Stanhope et al., 2023).

Sample preparation: 12 samples (n = 3 of wild type, Ca_hrd1/Ca_hrd1, Ca_doa10/Ca_doa10, and Ca_ubc7/Ca_ubc7 yeast) were submitted to the IUSM Center for proteome analysis, where proteins were denatured in 8 M urea, 100 mM Tris-HCl, pH 8.5 with sonication using a Bioruptor^® sonication system (Diagenode Inc, Denville, NJ, USA) with 30 s/30 s on/off cycles for 15 min in a water bath at 4 °C. After subsequent centrifugation at 14,000 rcf for 20 min, protein concentrations were determined by Bradford protein assay (BioRad Cat No: 5000006). A total of 100 µg equivalent of protein from each sample were reduced with 5 mM tris (2-carboxyethyl) phosphine hydrochloride (TCEP, Sigma-Aldrich Cat No: C4706) for 30 min at room temperature and alkylated with 10 mM chloroacetamide (CAA, Sigma Aldrich Cat No: C0267) for 30 min at room temperature in the dark. Samples were diluted with 50 mM Tris.HCl, pH 8.5 to a final urea concentration of 2 M for Trypsin/Lys-C based overnight protein digestion at 37 °C (1:70 protease:substrate ratio, Mass Spectrometry grade, Promega Corporation, Cat No: V5072.).

Peptide purification and labeling: Digestions were acidified with trifluoroacetic acid (TFA, 0.5% v/v) and desalted on Sep-Pak^® Vac cartridges (Waters^TM Cat No: WAT054955) with a wash of 1 ml 0.1% TFA followed by elution in 70% acetonitrile 0.1% formic acid (FA). Peptide concentrations were checked by Pierce Quantitative colorimetric assay (Cat No: 23275) and confirmed to be consistent across all samples. 50 µg peptides were then labeled with 0.25 mg Tandem Mass Tag pro (TMTpro) reagent (Thermo Fisher Scientific, TMTpro^™ Isobaric Label Reagent Set; Cat No: A44520, Lot VL313890; see Table S1) for 2 h at room temperature, quenched with a final concentration v/v of 0.3% hydroxylamine at room temperature for 15 min. Labeled peptides were mixed and dried by speed vacuum.

High pH basic fractionation: For high pH basic fractionation, peptides were reconstituted in 0.1% trifluoroacetic acid and half of the mixture was fractionated on a 50 mg Sep-Pak^® Vac cartridge using methodology and reagents from Pierce^™ High pH reversed-phase peptide fractionation kit (Thermo Fisher Cat No: 84868).

Nano-LC-MS/MS: Global proteomics were performed on an EASY-nLC 1200 HPLC system (SCR: 014993; Thermo Fisher Scientific, Waltham, MA, USA) coupled to Lumos Orbitrap^™ mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). A total of 1/5 of each fraction was loaded onto a 25-cm EasySpray column (Thermo Fisher Scientific ES902A, Waltham, MA, USA) at 400 nl/min. Peptides were eluted from 4–30% with mobile phase B (Mobile phases A: 0.1% FA, water; B: 0.1% FA, 80% Acetonitrile (Thermo Fisher Scientific Cat No: LS122500, Waltham, MA, USA)) over 160 min, 30–80% B over 10 min, and dropping from 80–10% B over the final 10 min. The mass spectrometer was operated in positive ion mode with a 4-s cycle time data-dependent acquisition method with advanced peak determination and Easy-IC (internal calibrant) on. Precursor scans (m/z 375–1,600) were done with an orbitrap resolution of 120,000, RF lens% 30, maximum inject time 50 ms, AGC target of 100% (4e5), MS2 intensity threshold of 2.5e4, MIPS mode to peptide, including charges of 2 to 7 for fragmentation with 30 s dynamic exclusion. MS2 scans were performed with a quadrupole isolation window of 0.7 m/z, 37% HCD CE, 50,000 resolution, 200% normalized AGC target, dynamic maximum IT, fixed first mass of 100 m/z.

Mass spectrometry data analysis: Resulting RAW files were analyzed in Proteome Discover^™ 2.5 (Thermo Fisher Scientific, Waltham, MA, USA) with a C. albicans UniProt FASTA (downloaded 03/15/2022, 6,030 entries) plus common contaminants. SEQUEST HT searches were conducted with a maximum number of three missed cleavages, precursor mass tolerance of 10 ppm, and a fragment mass tolerance of 0.02 Da. Static modifications used for the search were carbamidomethylation on cysteine residues and TMTpro label on lysine residues. Dynamic modifications included oxidation of methionine, TMTpro label on peptide N terminus, and acetylation, methionine-loss, or methionine-loss plus acetylation on protein N terminus. Percolator False Discovery Rate (FDR) was set to a strict peptide spectral match FDR setting of 0.01 and a relaxed setting of 0.05. In the consensus workflow, peptides were normalized by total peptide amount with no scaling. Co-isolation thresholds of 50% and average reporter ion S/N cutoffs of five were used for quantification. All peptides were used for normalization and protein roll-up and modified peptides were excluded in the pairwise ratio calculation. Protein abundance-based ratio calculations were done with no imputation. Protein FDR validator node was set to a strict target FDR of 0.01 and relaxed of 0.05. Resulting normalized abundance values for each sample type, abundance ratio and log2 (abundance ratio) values, and respective p-values (protein abundance-based ratio calculation and individual protein ANOVA) from Proteome Discover^™ were exported to Microsoft Excel.

Structural analyses of ERAD enzymes

We identified putative C. albicans genes encoding ERAD enzymes CaHrd1, CaDoa10, and CaUbc7. A summary of amino acid sequence identity and similarity between homologous proteins in C. albicans, S. cerevisiae, and H. sapiens is presented in Table S2. Domain organization and AlphaFold-predicted (Jumper et al., 2021) or experimentally determined (Arai et al., 2006; Cook et al., 1997; Wu et al., 2020) structures of these enzymes are presented in Fig. 2.

All three (H. sapiens, S. cerevisiae, and C. albicans) Hrd1 homologs possess an N-terminal transmembrane domain with eight predicted membrane-spanning segments, a catalytic Really Interesting New Gene (RING) domain, and a largely unstructured C-terminal extension (CTE), which likely contributes to substrate and cofactor interactions (Figs. 2A and 2B) (Omura et al., 2006; Schulz et al., 2017). The predicted CaHrd1 CTE possesses H. sapiens-like and S. cerevisiae-like features. The CaHrd1 CTE includes an alpha helix with amphipathic character (green helix in Fig. 2B) resembling the human C-terminal HAF-H domain implicated in HsHRD1 complex formation (Schulz et al., 2017). Like the S. cerevisiae enzyme, the CaHrd1 CTE includes a predicted two-stranded beta-sheet (cyan strands in Fig. 2B) with uncharacterized function.

AlphaFold-predicted structures of H. sapiens, S. cerevisiae, and C. albicans Doa10/MARCHF6 proteins include N-terminal RING domains, large C-terminal transmembrane domains, and short N- and C-terminal extensions (Figs. 2A and 2C). The transmembrane portions of these proteins include a conserved three-transmembrane segment TEB4-Doa10 (TD) domain (cyan in Fig. 2C) (Kreft & Hochstrasser, 2011). The large transmembrane portion of Doa10 homologs has been proposed to function as a retrotranslocation channel for ER export of transmembrane substrates, with the first (yellow) and fourteenth (green) transmembrane segments forming a lateral gate for substrate entry (Mehrtash & Hochstrasser, 2022; Schmidt, Vasic & Stein, 2020). The ScDoa10 and HsMARCHF6 CTEs promote substrate ubiquitylation (Mehrtash & Hochstrasser, 2022; Zattas et al., 2016); AlphaFold-guided mutational analyses indicate interactions between the CTE and N-terminal RING domain are essential for optimal ScDoa10 function (Mehrtash & Hochstrasser, 2022). AlphaFold structural predictions indicate this interaction is likely also present in CaDoa10 and HsMARCHF6. Alignments of Hrd1 and Doa10 homolog RING domains demonstrate conservation of zinc-coordinating cysteine and histidine residues as well as a tryptophan residue commonly found in ubiquitin ligase catalytic domains (Figs. 2B and 2C). Complete alignments of Hrd1 and Doa10 homologs are presented in Supplemental Files 1 and 2.

H. sapiens, S. cerevisiae, and C. albicans Ubc7 homologs exhibit strong conservation of primary and predicted tertiary structure (Figs. 2A and 2D). Comparison of crystal structures of ScUbc7 and HsUBE2G2 with AlphaFold-predicted CaUbc7 reveals all three adopt a characteristic ubiquitin-conjugating enzyme fold; the position of the invariant catalytic cysteine is indicated.

Generation of mutants

We generated functional deletions of Ca_HRD1, Ca_DOA10, or Ca_UBC7 using CRISPR-mediated genome editing. Stop codons were inserted at the 5′ ends of each of these genes, effectively creating knockouts of Ca_HRD1, Ca_DOA10, and Ca_UBC7. PCR and Sanger sequencing were used to confirm the correct mutations were made and that mutations were homozygous, preventing translation of both alleles’ transcripts. Heterozygous Ca_HRD1/Ca_hrd1 mutants were also isolated. Introduction of these mutations did not impact growth on YPD (Fig. 3).

C. albicans lacking HRD1, DOA10, or UBC7 exhibit sensitivity to proteotoxic stress

A quality control function for C. albicans genes encoding ERAD factors has not previously been reported. S. cerevisiae with mutations in protein quality control genes (including those encoding ERAD enzymes) exhibit elevated sensitivity to hygromycin B (Crowder et al., 2015; Runnebohm et al., 2020a; Verma et al., 2013; Woodruff et al., 2021). Hygromycin B distorts the ribosomal A site, thereby reducing translational fidelity, which is predicted to increase the abundance of aberrant proteins (Brodersen et al., 2000; Ganoza & Kiel, 2001). We compared the growth of wild type yeast and yeast with homozygous disruptions of Ca_HRD1, Ca_DOA10, or Ca_UBC7 in the presence of increasing concentrations of hygromycin B across a range of temperatures (Fig. 3). Homozygous deletion of Ca_HRD1, Ca_DOA10, or Ca_UBC7 caused a pronounced growth defect in the presence of hygromycin B. Sensitivity to hygromycin B was markedly enhanced at elevated temperatures.

We also compared hygromycin B sensitivity of yeast possessing homozygous and heterozygous mutations in Ca_HRD1 (Fig. 4). Heterozygous Ca_HRD1/Ca_hrd1 yeast exhibited wild type resistance to hygromycin B at 25, 30, and 37 °C, indicating a single copy of Ca_HRD1 is sufficient to confer protection from hygromycin B at these temperatures. We observed a subtle growth defect of Ca_HRD1/Ca_hrd1 yeast in the presence of hygromycin B at 40 °C.

We analyzed susceptibility of wild type yeast and ERAD mutants to drugs that cause ER stress (disulfide-reducing agents beta-mercaptoethanol (βME) and dithiothreitol (DTT)) (Jia et al., 2019; Wu, Ng & Thibault, 2014). Homozygous mutation of Ca_HRD1, Ca_DOA10, or Ca_UBC7 enhanced sensitivity to βME (Fig. 5). By contrast, wild type and ERAD mutant strains exhibited similar susceptibility to DTT. Together, our results suggest Ca_HRD1, Ca_DOA10, or Ca_UBC7 are important for C. albicans growth under proteotoxic stress.

Proteomic analysis of ERAD mutant yeast strains

To identify candidate physiological substrates and interactors of ERAD enzymes, we identified proteins with altered abundance in yeast lacking Ca_HRD1, Ca_DOA10, or Ca_UBC7 relative to wild type yeast using quantitative tandem mass tag (TMT) mass spectrometry. The complete results of these experiments are presented in Supplemental File 3 and summarized in Fig. 6, Tables 1–3 and Tables S2–S8. In total, 229 proteins exhibited significant changes in abundance in at least one of the three mutants analyzed (Fig. 6A; Table 1). Compared to wild type yeast, 42 proteins exhibited a statistically significant increase and 19 proteins demonstrated a significant decrease in abundance in Ca_hrd1/Ca_hrd1 yeast. In Ca_doa10/Ca_doa10 yeast, 39 proteins were significantly elevated, and 24 were significantly downregulated relative to wild type yeast. Consistent with a broader role for CaUbc7, a greater number of C. albicans proteins exhibited significant alterations in Ca_ubc7/Ca_ubc7 yeast (82 increased, 63 decreased) compared to wild type yeast. The proteins with the largest significant increases in abundance in each mutant strain are presented in Table 2. 52% of proteins altered in Ca_hrd1/Ca_hrd1 yeast, and 11% of proteins altered in Ca_doa10/Ca_doa10 yeast, exhibited coordinated shifts in abundance in Ca_ubc7/Ca_ubc7 yeast (Fig. 6B; Tables S3 and S4). Seven proteins exhibited significant, opposite-direction changes in abundance in multiple mutants (Table S5). Loss of HRD1 caused a greater enrichment of predicted ER-targeted proteins (i.e., proteins with signal peptides and/or transmembrane segments) than did loss of DOA10 (Table 3).

We performed Gene Ontology (GO) analysis of genes encoding proteins with statistically significant changes in abundance in ERAD mutants (Tables S6–S8). Among proteins exhibiting increased abundance in Ca_hrd1/Ca_hrd1 yeast, genes with products that function in or are predicted to function in the ER and secretory pathway were enriched, including CaPdi1, CaKar2, CaDpm1, and CaSec61, homologs of proteins that are upregulated by the ER homeostatic unfolded protein response (UPR) in other organisms, including S. cerevisiae (Chapman, Sidrauski & Walter, 1998; Travers et al., 2000). GO analysis indicates a statistically significant enrichment of proteins mediating sterol synthesis in Ca_ubc7/Ca_ubc7 yeast (Table S8). The C-5 sterol desaturase ScErg3 is a bona fide ScHrd1/ScUbc7 substrate in S. cerevisiae (Jaenicke et al., 2011); enrichment in both Ca_hrd1/Ca_hrd1 and Ca_ubc7/Ca_ubc7 mutants indicates CaErg3 is likely to be a physiological target of the C. albicans enzymes as well. Overall, our results suggest C. albicans ERAD pathway homologs play important roles in protein quality control and regulated protein degradation.