Introduction

Glycoproteins, destined for the secretory pathways, enter the endoplasmic reticulum (ER) lumen where protein folding occurs (Rapoport, 2007). The accumulation of misfolded proteins affects cell viability and homeostasis; therefore, eukaryotes evolved a conserved ER quality control (ERQC) system that recognizes folding defects, repairs them, or ensures the translocation of irreparable misfolded proteins into the cytosol for proteasome-mediated degradation via the ER-associated degradation (ERAD) system (Thibault and Ng, 2012; Xu and Ng, 2015; Balchin et al., 2016), a process that heavily relies on N-glycosylation to determine protein folding conformation (Aebi, 2013; Varki, 2017).

Many aspects of N-glycosylation are highly conserved, and most eukaryotes initiate this process by synthesizing a Dol-PP-linked Glc₃Man₉GlcNAc₂ oligosaccharide as a common core N-glycan, which attaches to nascent polypeptides in the ER via an asparagine residue. Glucosidases I and II (Gls1 and Gls2) remove two glucose residues from the core oligosaccharide. Proteins containing monoglucosylated N-glycans (Glc₁Man₉GlcNAc₂) enter a calnexin and/or calreticulin (CNX/CRT in mammals; Cne in yeast) chaperone-mediated folding cycle. Finally, Gls2 cleaves the remaining glucose residue. If proteins are misfolded, they are recognized by the ERQC checkpoint enzyme UDP-glucose:glycoprotein glucosyltransferase (UGGT), which reglucosylates them for re-entry into the folding cycle (Fig. 1A). Then, the N-glycans of the accurately folded proteins are further processed by Gls2 and α1,2-mannosidase I (Mns1) and moved to the Golgi apparatus. However, irreparably misfolded glycoproteins are targeted for ERAD, where they undergo demannosylation and retro-translocation for proteasomal degradation in the cytosol. Recently, N-glycan precursors in the ER of some protists and fungi were found to be shorter than the typical 14-sugar N-glycan precursors in most eukaryote organisms. The length of these N-glycan precursors significantly impacts N-glycan-dependent QC of glycoprotein folding and ERAD (Banerjee et al., 2007; Samuelson et al., 2015).

Figure 1.

The basidiomycetous fungus Cryptococcus neoformans is an opportunistic encapsulated human pathogen that primarily affects immunocompromised individuals, causing fatal meningoencephalitis (Gottfredsson and Perfect, 2000; Kwon-Chung et al., 2000). The N-glycosylation pathway of C. neoformans is evolutionarily conserved; nevertheless, the structure and biosynthesis of its N-glycans includes several unique features (Park et al., 2012). C. neoformans contains serotype-specific high-mannose-type N-glycans with or without a β-1,2-xylose residue attached to the trimannosyl core. Additionally, acidic N-glycans of C. neoformans contain xylose phosphates attached to the mannose residues both within the N-glycan core and outer mannose chains. The intact core N-glycan structure is crucial for C. neoformans pathogenicity (Thak et al., 2020); hence, alterations in the N-glycan structure modulates the interaction between the cell surface mannoproteins and host cells (Lee et al., 2023). Additionally, C. neoformans lacks homologous genes to the Asn-Linked Glycosylation (ALG) genes ALG6, ALG8, and ALG10, which are evolutionary conserved across most eukaryotic organisms and encode the glucosyltransferases that add the glucose residues to the core N-glycan before its attachment to proteins (Park et al., 2012). Man₇GlcNAc₂ and Man₈GlcNAc₂ without glucose residues are primarily detected in Dol-PP-linked glycans of C. neoformans (Samuelson et al., 2005). Moreover, the mature core N-glycan structures assembled on the cell surface mannoproteins of C. neoformans are primarily Man_6–7GlcNAc₂, which are shorter than the expected Man₈GlcNac₂ (Park et al., 2012). This observation led to the speculation that the terminal α-1,2-mannose residues of C. neoformans N-linked glycans may be more susceptible to trimming by ER α-1,2 mannosidases due to the lack of glucose residues compared to most eukaryotes. Alternatively, the presence of multiple α-1,2 mannosidases may generate more extensively trimmed core N-glycans in C. neoformans.

C. neoformans employs several virulence factors, including an extensive polysaccharide capsule composed of glucuronoxylomannan (GXM) and galactoxylomannan (GalXM) (Doering et al., 2009), melanin (Qiu et al., 2012), and various extracellular enzymes such as phosphatase and urease (Singh et al., 2013), all of which contribute to immune evasion and enhance fungal pathogenicity. To facilitate the export of these virulence-associated molecules, C. neoformans employs both conventional (Yoneda and Doering, 2006; Panepinto et al., 2009) and unconventional secretion pathways (Rodrigues et al., 2007; Casadevall et al., 2019; Rizzo et al., 2021). Classical secretion relies on the ER-Golgi network, while non-classical mechanisms, such as extracellular vesicle (EV)-mediated transport, provide an alternative route for cargo delivery beyond the conventional secretory pathway. C. neoformans EVs have been described as a heterogeneous population of “virulence bags” containing numerous fungal survival and pathogenicity-associated cargo. Proteomic analyses of EV cargo have identified several cell surface glycoproteins, including members of the CDA family, well-known immunomodulators (Specht et al., 2017; Rizzo et al., 2021). The ERQC system is essential for maintaining protein homeostasis by ensuring proper folding, glycosylation, and directed trafficking of secreted proteins to downstream compartments within the conventional secretory pathway. Defects in ERQC affect the fidelity of protein secretion not only by accumulation or degradation of misfolded proteins, leading to the impaired secretion of functional proteins, but also by the escape of misfolded forms, thus resulting in increased secretion of non-properly processed proteins (Marcus and Perlmutter, 2000; Chen et al., 2024). However, the potential of ERQC to modulate EV-mediated protein trafficking is not yet systematically investigated.

This study presents the first systematic analysis of N-glycan-dependent ERQC in C. neoformans using mutant strains lacking ERQC gene homologs. Our findings highlight the critical roles of ERQC in maintaining cellular fitness and facilitating EV-mediated transport of virulence factors.

Results

Evolutionary unique features of C. neoformans ERQC components

We performed BLAST analysis of the C. neoformans H99 genome to identify the ERQC pathway-associated homologous genes of C. neoformans, followed by comparison with the ERQC components in other eukaryotic organisms (Supplementary Fig. S1, A). Unlike most eukaryotes, several yeast species within the Ascomycota phylum, such as Saccharomyces cerevisiae and Candida albicans, do not possess a functional UGGT. In contrast, UGGT homologs were identified in most Basidiomycota fungal species, including C. neoformans (Supplementary Fig. S1, A and B). The C. neoformans UGGT (CNAG_03648), which has been named as Ugg1, consists of 1,582 amino acids (aa) and features a signal peptide (1–20 aa) along with four tandem-like thioredoxin-like (TRXL) domains: TRXL12 (33–320 aa), TRXL13 (28–416 aa), TRXL14 (432–618 aa), and TRXL15 (712–950 aa). Ugg1 also contains a glucosyltransferase (GT) 24 domain with a DXD motif (1,369–1,371 aa), and a KDEL-like ER retention signal (1579–1582 aa), which facilitates its retrieval from the Golgi apparatus (Fig. 1B, Supplementary Fig. S2, A).

Additionally, we identified two C. neoformans ORFs, Mns1 (CNAG_02081) and Mns101 (CNAG_03240), as homologs of the eukaryote α1,2-mannosidase I, which processes N-glycans before exporting them to the Golgi. The C. neoformans Mns1 and Mns101 show 42.4% and 28.8% amino acid identities to S. cerevisiae ER α1,2-mannosidase I, respectively, and share 30.9% identity between them. Furthermore, C. neoformans Mnl1 (CNAG_01987) and Mnl2 (CNAG_04498) were identified as putative components of ERQC in C. neoformans (Supplementary Fig. S1, A and C). C. neoformans Mnl1 is a homolog of the yeast α1,2-mannosidase-like protein Htm1, which processes N-glycans targeted for ERAD. In contrast, C. neoformans Mnl2 encodes a mannosidase that does not share significant similarity with mannosidases from other eukaryotes. Mns1, Mns101, and Mnl1 possess glucosyl hydrolase (GH) 47 domains, essential for mannosidase activity, whereas Mnl2 contains a GH92 domain, also associated with mannosidase activity (Fig. 1B). The Mns1 and Mnl1 families are characterized by conserved cysteine (Cys) and alanine (Ala) residues within their activity domains, as previously described (Jakob et al., 2001). These conserved residues are present in C. neoformans Mns1, Mns101, and Mnl1 but are absent in Mnl2 (Supplementary Fig. S2, B). Notably, Mns101 and Mnl2 are Basidiomycota-specific proteins (Supplementary Fig. S1, A) and appear to have diverged early from other fungal mannosidase clades based on phylogenetic analysis (Supplementary Fig. S1, C). Interestingly, neither Mnl1 nor Mnl2 contains canonical KDEL/HDEL-like ER retention signals. In S. cerevisiae, the ER retention of Mnl1/Htm1 is mediated through its interaction with protein disulfide isomerase Pdi1, which carries an HDEL sequence (Gauss et al., 2011). Similarly, C. neoformans Mnl1 and Mnl2 may employ a non-canonical retention mechanism, likely facilitated by interactions with other ER-resident proteins, to achieve ER localization.

Loss of UGG1, MNS1, and MNS101 causes alteration of N-glycan profiles in C. neoformans

We performed high-performance liquid chromatography (HPLC) analysis of the cell wall mannoproteins (cwMPs) from both the wild-type (WT) and ERQC mutant strains to investigate the ERQC malfunction-induced structural differences in N-glycans (Fig. 2). The HPLC profiles of cwMPs from the WT strain showed an M8 peak as the major species. This peak corresponded to N-glycans with 8 mannose residues (Fig. 2A, top). The glycan structure at the M8 peak (Man₈GlcNAc₂) in the WT primarily corresponded to the Man₇GlcNAc₂ core N-glycan with an additional mannose residue that is linked via an α1,6-linkage and added in the Golgi apparatus (Park et al., 2012). In the ugg1Δ mutant, M8 was also the main N-glycan species, but the pools of hypermannosylated N-glycans (larger than the M11 peak) were markedly reduced (Fig. 2A, middle). The altered N-glycan profile of the ugg1Δ mutant was restored to that of the WT after complementation with the WT UGG1 gene (Fig. 2A, bottom). The lectin blotting analysis using Galanthus nivalis agglutinin (GNA), which specifically binds to terminal α1,2-, α1,3-, and α1,6-linked mannose residues, showed a distinctive increase of glycoproteins with lower molecular weight in the secretory proteins in ugg1Δ compared to those in WT (Fig. 2B). This observation aligns with the decrease of hypermannosylated N-glycans in the ugg1Δ mutant. Overall, these results suggest that Ugg1 is involved in mediating the hypermannosylation of N-glycans in the Golgi apparatus.

Figure 2.

We also assessed putative α1,2-mannosidase I genes for roles in protein mannosylation. The mns1Δ mutant N-glycan profile showed a peak shift from an M8 to an M9 form, which strongly indicates that C. neoformans Mns1 could be the primary ER α1,2-mannosidase I (Fig. 2C, mns1Δ). Notably, the loss of Mns101, which is present only in Basidiomycota, increased the fractions containing hypermannosylated glycans (> M10) while maintaining M8 as a primary core N-glycan form. This suggests that the basidiomycete-specific Mns101 may potentially be a novel α1,2-mannosidase that functions to remove mannose residues from hypermannosylated N-glycans in the Golgi apparatus or further trims the M8 glycan in the ER before the glycoproteins are transported to the Golgi (Fig. 2C, mns101Δ). The matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometer analysis of neutral N-glycans (Fig. 2C), further confirmed the shift of the major M8 peak in the WT strain to M9 in the mns1Δ mutant and the increase in the hypermannosylation profile in the mns101Δ mutant (Fig. 2D). The N-glycan profiles of the mns1Δ101Δ double mutant showed the combined effect of each null mutation, and it showed both increased hypermannosylated glycans and a shift from the M8 to the M9 peak in both the HPLC (Fig. 2C, mns1Δ101Δ) and MALDI-TOF analyses. This indicates that Mns1 and Mns101 serve as mannosidases and play independent roles at different stages of N-glycan processing in C. neoformans.

In contrast, the loss of both MNL1 and MNL2 did not show notable differences in the HPLC profile of N-glycans from cell surface mannoproteins (Supplementary Fig. S3, A). Considering the expected function of Mnl1 and Mnl2 in ERAD, their substrates are likely misfolded proteins that have not been transported to the Golgi apparatus. Thus, we speculate that only the normally processed N-glycan profiles of cell surface mannoproteins in the mnl1Δ mnl2Δ mutant strain were observed.

Loss of ERQC components results in defective growth fitness and increased stress sensitivity

To investigate the changes in ERQC-related gene expression under stress conditions, we conducted quantitative reverse transcription PCR (qRT-PCR) analysis of C. neoformans following treatment with tunicamycin (TM, 5 µg/ml), a well-established inhibitor of N-glycosylation, or dithiothreitol (DTT, 20 mM), which disrupts disulfide bond formation and induces the accumulation of misfolded proteins. Additionally, we examined gene expression after growth at 37 °C (Fig. 3A). When misfolded proteins accumulate in the ER, the unfolded protein response (UPR) system is activated to restore homeostasis (Chakraborty et al., 2016). As expected, KAR2, encoding a molecular chaperone associated with the UPR system, was upregulated in response to TM treatment, DTT treatment, and heat stress at 37 °C. The ERQC components UGG1, MNS1, MNS101, MNL1, and MNL2 were upregulated by DTT treatment; however, only UGG1 was slightly induced following TM treatment. The high-temperature conditions also induced the expression of KAR2, UGG1 and MNL2. Notably, DTT treatment induced a 10-fold higher expression of MNS1 and MNL1. These results strongly suggest that the C. neoformans genes UGG1, MNS1, MNS101, MNL1, and MNL2 are crucial components of the ERQC and ERAD systems, although each respond uniquely to various cell/ER stress conditions.

Figure 3.

We next investigated the growth of the ugg1Δ mutant under various stress-inducing conditions to elucidate the roles of the ERQC components in ER stress response and adaptation, which are closely associated with C. neoformans virulence. The ugg1Δ mutant showed impaired growth even under normal growth conditions (Fig. 3B) and increased sensitivity to ER stress-inducing agents such as DTT, TM and cell wall stressors, including as calcofluor white (CFW), congo red (CR), sodium dodecyl sulfate (SDS), and caffeine. Additionally, ugg1Δ showed hindered growth in the presence of antifungal drugs such as azole agents (fluconazole and ketoconazole) and the glucose transport inhibitor fludioxonil. These findings suggest that Ugg1 is crucial for the robust growth and survival of C. neoformans under various stress conditions. In contrast, the single deletion (mns1Δ and mns101Δ) and double deletion (mns1Δ101Δ) strains did not exhibit noticeable phenotypic changes except for a slight increase in sensitivity of mns101Δ and mns1Δ101Δ to higher fludioxonil concentrations (Supplementary Fig. S3, B). Similarly, neither the single nor double disruption of MNL1 and MNL2 produced detectable changes in the tested growth conditions (Supplementary Fig. S3, C).

Amino acid analogues have been used for evaluating the functionality of protein folding, as its incorporation into newly synthesized proteins interrupts proper protein folding (Trotter et al., 2002; Cowie et al., 1959). We hypothesized that, in the context of a defective ERQC pathway, misfolded proteins cannot be adequately repaired and accumulate, thus triggering ER stress, which may ultimately inhibit cell growth in the presence of amino acid analogues. In the presence of the leucine analogue 5’,5’,5’-trifluoroleucine (TFL), ugg1Δ showed noticeably inhibited growth, whereas the WT strain showed no growth inhibition (Fig. 3C). The mns1Δ101Δ strain also exhibited slightly increased sensitivity to TFL compared with that of the WT strain.

As a defense mechanism against misfolded protein accumulation-mediated ER stress, UPR is induced by the unconventional splicing of the HXL1 transcription factor in C. neoformans (Cheon et al., 2011). RT-PCR analysis revealed a significantly elevated level of spliced HXL1 in the ugg1Δ strain compared to the WT, even under normal culture conditions (Fig. 3D), strongly indicating that the loss of UGG1 leads to the accumulation of misfolded proteins, triggering ER stress under standard growth conditions. Furthermore, green fluorescence protein (GFP)-tagged Ugg1, Mns1, and Mns101 proteins colocalized with the ER marker, supporting their role as confirming them to be functional ERQC components based on their subcellular localization in the ER (Supplementary Fig. S4).

ERQC defects lead to virulence attenuation in C. neoformans

We investigated the effects of a defective ERQC on in vitro virulence phenotypes of C. neoformans by analyzing capsule and melanin production, as these are two of the major virulence factors of C. neoformans. Culturing on L-DOPA-containing plates showed that melanin production was reduced in the ugg1Δ strain compared with that in the WT (Fig. 4A, top), whereas the mns1Δ101Δ mutant cells did not show detectable defects at either 30 or 37 °C (Fig. 4A, bottom). Considering the decreased growth of the ugg1Δ strain, the melanin production of C. neoformans cells cultivated in liquid L-DOPA was measured and normalized by cell density, which further confirmed the decreased melanin production activity of the ugg1Δ strain (Fig. 4B). India ink staining of the capsule showed significantly reduced capsule thickness in the ugg1Δ strain and a moderate defect in the mns1Δ101Δ strain (Fig. 4C).

Figure 4.

Defects in the ERQC system led to a more apparent decrease in pathogenicity in a murine model of systemic cryptococcosis. Notably, the ugg1Δ strain was almost avirulent with no detectable signs of illness in the infected animals (Fig 4D). Although MNS1 and MNS101 single disruptions did not cause a detectable decrease in C. neoformans pathogenicity, their double disruption resulted in decreased virulence. Histopathological analysis of infected mice lungs suggested significantly reduced lung colonization by both ugg1Δ and mns1Δ101Δ cells (Supplementary Fig. S5, A). Additionally, the fungal burden of ugg1Δ-infected animals showed a notable decrease in organ colonization at 60 days post-infection (dpi) (Supplementary Fig. S5, B). Analysis of the fungal burden at 7 dpi showed significantly reduced organ colonization of the mns1Δ101Δ mutant compared with that of the WT (Supplementary Fig. S5, C). Consistent with the in vivo survival data, the number of surviving ugg1Δ and mns1Δ101Δ cells within the macrophage-like cell line J447A.1 was significantly lower compared with that of the WT strain (Fig. 4E). These data suggest that a functional ERQC system is critical for maintaining full in vivo pathogenicity and ensuring robust survival within host immune cells.

Loss of Ugg1 function manifests defects in extracellular transport of GXM

Cryptococcal capsule materials are synthesized intracellularly and secreted to the cell surface. Subsequently, they are assembled and bound to the cell wall (Yoneda and Doering, 2006). The observed capsule defects in the ugg1Δ and mns1Δ101Δ mutant strains may be attributed to issues in the capsule polysaccharide synthesis or transport steps. Alternatively, defective capsule formation could arise from an increased shedding of capsule polymers, preventing their proper attachment to the cell surface. To determine the cause of the capsule defects, we examined the amounts of capsular polysaccharides synthesized intracellularly and those shed into the culture supernatant by a capsule blotting assay, using the 18B7 antibody directed against GXM, the major capsule component (Casadevall et al., 1998). As controls, we also tested capsule production and secretion in the cap59Δ strain with a defect in GXM synthesis (Grijpstra et al., 2009) and the rim101Δ strain with a capsule attachment defect despite normal polysaccharide synthesis and secretion (O’Meara et al., 2010). Intracellular GXM quantities did not notably differ between the WT, ugg1Δ, mns1Δ101Δ, and rim101Δ strains, although ugg1Δ showed slightly reduced GXM level (Fig. 5A, left). In contrast, no intracellular GXM was detected in cap59Δ, which was consistent with its defect in GXM synthesis. Moreover, the ugg1Δ strain was defective in capsule polymer secretion/shedding, whereas the mns1Δ101Δ strain displayed enhanced shedding of GXM-containing polymers that appeared to be shorter in length than the WT-type polymers, suggestive of poor polysaccharide polymerization (Fig. 5A, right). Therefore, these results suggest that ugg1Δ contains a defect in polysaccharide trafficking to the extracellular space, whereas mns1Δ101Δ has a defect in polysaccharide polymerization, resulting in defective capsule elaboration, albeit with differential degrees of severity in both ERQC mutants.

Figure 5.

Furthermore, to determine whether the ERQC mutants are also impaired in capsule attachment, we performed a capsule transfer assay, to assess whether exogenously shed capsule polysaccharides can bind to acapsular mutants (Reese and Doering, 2003). The hypocapsular rim101Δ mutant did not show reassociation of capsular polysaccharides; however, the acapsular cap59Δ mutant showed recovery of capsule formation when co-incubated with exogenous GXM (Fig. 5B). Similarly, ugg1Δ reverted its acapsular phenotype on incubation with the WT-shed polysaccharides, strongly suggesting that the defective capsule phenotype of ugg1Δ is primarily due to defective GXM secretion rather than impaired attachment of capsule polymers. We further conducted a capsule transfer assay using capsular polysaccharides shed from mns1Δ101Δ. The cap59Δ mutant reverted to the capsular phenotype by attaching to the mns1Δ101Δ-shed polymers. However, the fluorescence intensity of the capsule generated was significantly lower than that of the cells attached with polysaccharides obtained from the WT. Although the decreased length of capsule polysaccharides should be validated by techniques specifically measuring GXM size (De Jesus et al., 2010), this result suggests that the mns1Δ101Δ double mutant secretes incomplete capsule polysaccharides (Fig. 5C), leading to a hypocapsular phenotype. These results collectively indicate that disrupting the ERQC results in impaired capsule formation by defective GXM trafficking to the extracellular space.

Transmission electron microscopy (TEM) imaging of the WT and mutant cells distinctly showed a diminished capsule structure and loss of capsule shedding in the ugg1Δ strain compared with that of the WT strain (Fig. 5D). Additionally, we observed considerable thinning of the cell walls in both ERQC mutant cells. Notable changes in intracellular structures along with an increase in the number of pigmented vesicles were observed in the mutant cells even under YPD culture conditions. Under capsule-inducing conditions, a significant accumulation of vesicular structures (electron-lucent structures) was observed intracellularly, particularly in ugg1Δ. We performed lipid droplet (LD) staining to determine whether these vesicular structures might be LDs. We observed a notable increase in LDs in ugg1Δ under capsule-inducing culture conditions (Supplementary Fig. S6, A). LDs impact proteostasis by sequestering misfolded proteins intended for degradation and providing an “escape hatch” when the ERQC is overloaded (Ploegh, 2007; Vevea et al., 2015). Additionally, FM4-64 dye staining showed increased number of vacuoles in the ugg1Δ cells compared with that in the WT (Supplementary Fig. S6, B). Moreover, all the observed abnormal phenotypes were less pronounced in the mns1Δ101Δ mutant than in the ugg1Δ strain. Collectively, the altered vesicular structures observed in the ugg1Δ and mns1Δ101Δ strains highlight the critical role of ERQC in maintaining proper vesicular architecture and cell surface organization. These functions are tightly linked to extracellular trafficking of virulence factors and cell wall remodeling.

ugg1Δ transcriptomic profiling shows induced ER and cell wall integrity stress responses

To elucidate the mechanisms underlying UGG1 deletion-induced physiological changes in C. neoformans, we performed RNA sequencing (RNAseq)-based transcriptome analysis of the WT and ugg1Δ strains under standard growth conditions (YPD medium, 30 °C). Comparative transcriptome analysis showed statistically significant alterations exceeding 2-fold in the expression patterns of 146 genes, of which 85 were upregulated and 61 were downregulated compared with that of the WT strain (Fig. 6A, B, Supplementary Tables S2, A and B).

Figure 6.

Notably, the ugg1Δ mutant showed upregulation of genes such as SKN1 and KRE6 (transmembrane glucosidases involved in the sphingolipid biosynthesis and β-glucan biosynthesis), CAT2 (a putative peroxisomal catalase), CHS7 (chitin synthase export chaperone), and CNAG_05458 (a putative endo-1,3(4)-β-glucanase). In contrast, genes encoding ERG3 and ERG25 (associated with the ergosterol biosynthesis pathway) and FHB1 (flavohemoglobin denitrosylase associated with counteracting nitric oxide stress) were significantly downregulated. Gene Ontology (GO) analysis in ugg1Δ showed pronounced induction of genes implicated in various cellular processes, including the hydrolysis of O-glycosyl compounds for cell wall remodeling, DNA replication, proteolysis, ribosome biogenesis, protein folding, and serine/threonine kinase activity. In contrast, genes associated with carbohydrate and lipid metabolism, chaperone binding, iron homeostasis, and mitochondrial intermembrane space were considerably downregulated (Fig. 6C). The transcriptome profile strongly suggests that loss of Ugg1 function induces the expression of several genes involved in maintaining cell wall integrity to compensate for cell wall defects, particularly those associated with chitin (CDA2, CHS7, QRI1, and CNAG_06898) and glucan biosynthesis (SKN1, KRE6, EBG1, LPI9, CNAG_05458, and BLG2). Furthermore, genes coding for ER chaperones that aid in protein folding (ERO1, KAR2, LHS1, and PDI1) and chaperone regulator (SCJ1), and genes involved in proteolysis in the ER (CNAG_04635, CNAG_06658) were induced, likely as part of the ER stress response induced by the accumulation of misfolded proteins in the presence of a defective ERQC.

Further investigation of the effects of UGG1 disruption in the expression of genes involved in capsule biosynthesis, cell wall remodeling, and the secretion pathways of protein virulence factors using qRT-PCR analysis (Fig. 6D) showed no noticeable changes in the expression of genes related to capsule biosynthesis and conventional or unconventional secretion pathways at mRNA level, despite the aberrantly displayed defective phenotypes.

EV-mediated protein trafficking and non-conventional secretion is impaired in the ugg1Δ mutant strain

Many enzymes contribute to the composite cryptococcal virulence phenotype. Some of these virulence-associated enzymes are secreted through traditional secretion pathways, whereas others are packaged into EVs and released into the extracellular milieu via non-conventional secretion mechanisms (Almeida et al., 2015). The defects observed in the ERQC mutant strains, particularly in the production of melanin and polysaccharide capsules, both of which are transported by EVs, suggest a possible impairment in the EV-mediated extracellular transport of virulence-associated enzymes in these mutants. Therefore, we examined the in vitro activity of urease, and laccase and acid phosphatase in both intracellular and extracellular fractions.

Comparative analysis of the WT and ugg1Δ strains showed minimal changes in the intracellular activities of urease, laccase, and acid phosphatase. However, their activities were significantly reduced in the extracellular fractions of the ugg1Δ cells (Fig. 7A, 7B, 7C, and 7D). Importantly, urease activity was performed in both solid and liquid media, demonstrating a notable loss of secreted urease activity in ugg1Δ cells, with a slight decrease in the mns1Δ101Δ double mutant cells (Fig. 7A and 7B). As urease lacks a signal peptide required for conventional protein secretion, reduced urease secretion strongly indicates a compromised EV-mediated non-conventional secretory pathway in the ugg1Δ mutant strain. The decrease in extracellular laccase activity (Fig. 7C) corroborates our melanization assay results, in which ugg1Δ showed defective melanin production (Fig. 4A).

Figure 7.

We next examined cellulase and α-amylase activities to evaluate the impact of ERQC disruption on the conventional secretion pathway of non-virulence-related enzymes (Fig. 7E and 7F). The extracellular activities of these carbohydrate polymer-degrading enzymes in the ugg1Δ mutant were slightly lower than those observed in the WT, likely reflecting reduced intracellular enzyme production rather than a substantial impairment of secretion efficiency. Additionally, we analyzed the localization of chitin deacetylase I (Cda1), a glycosylphosphatidylinositol (GPI)-anchored protein involved in converting chitin to chitosan, essential for maintaining cell wall integrity (Baker et al., 2007; Baker et al., 2011; Upadhya et al., 2018; Upadhya et al., 2021). Cda1 was primarily detected in the insoluble cellular protein fraction, encompassing the cell wall and membrane compartments (Fig. 7G, left), consistent with its cell surface localization via a GPI anchor. Cda1 is secreted extracellularly upon cleavage of the GPI anchor. Notably, the ugg1Δ mutant exhibited significantly higher intracellular and extracellular Cda1 levels compared to the WT and mns1Δ101Δ strains (Fig. 7G, right), aligning with a 1.87-fold increase in CDA1 mRNA expression observed in the RNA-seq data. These findings indicate that the surface localization and secretion of Cda1 remain efficient, suggesting that the conventional secretion pathway is largely unperturbed in ugg1Δ. Collectively, these results suggest that while Ugg1-mediated ERQC defects have a pronounced negative effect on EV-mediated protein transport, their impact on the conventional secretion pathway is minimal.

Extracellular vesicle biogenesis and cargo loading are defective in ugg1Δ strain

We thus analyzed the number and size distribution of EVs to determine possible abnormal EV-mediated trafficking of virulence factors in the ugg1Δ mutant. Nanoparticle tracking analysis (NTA) showed a major peak in size distribution at approximately 150 (134±28) nm in the WT strain, which was similar in size to mammalian exosomes and typical microbial EVs. Additionally, a minor peak ranging from 300–500 nm was observed and corresponded to microvesicles (Fig. 8A). Notably, the size distribution of ugg1Δ EVs was more heterogeneous with smaller EVs ranging from 50–150 nm (82±16; 124±14) compared with that of the WT. The cap59Δ EVs displayed a major distribution of approximately 150 nm (120±21). Microvesicles were barely detected in either cap59Δ or ugg1Δ strains.

Figure 8.

We observed a significant reduction in the total number of secreted EVs in the ugg1Δ mutant (approximately 40%), suggesting defective EV biogenesis and/or stability in the absence of functional Ugg1 (Fig. 8B). Cryo-TEM analysis of EV morphology confirmed that the sizes of EVs released by ugg1Δ were significantly smaller and more diverse than those of the WT or the acapsular mutant (Fig. 8C), further supporting the NTA results. Examining EV size distribution by measuring their diameter (Fig. 8D) further confirmed that although the EVs in the WT strain ranged from approximately 50–550 nm in size with distribution primarily concentrated at approximately 150 nm, the EVs in the ugg1Δ mutant were smaller (< 100 nm) with heterogeneous sizes. Notably, cap59Δ EVs were uniformly distributed at approximately 150 nm and were present at a higher concentration than those of EVs from the WT. An increase in EV release and virulence factors has been also observed in acapsular mutant strains of C. neoformans and C. gattii (Rodrigues et al., 2007; Reis et al., 2019), supporting the notion that the capsule serves as a barrier in EV release in capsule forming species.

To characterize the protein composition of EVs from WT and ugg1Δ strains, EV-associated and whole-cell lysate (WCL) proteins were extracted and subjected to proteomic analysis (Fig. 8E). In the WCL, 4,678 proteins were identified in the WT strain, with 333 exhibiting differential expression (>2-fold) in ugg1Δ versus the WT strain, including 123 upregulated and 210 downregulated proteins (Supplementary Fig. S7, A, left). In EVs, 2,075 proteins were detected in the WT strain, of which 693 were differentially expressed in EVs from ugg1Δ relative to EVs from the WT strain (>2-fold), comprising 273 upregulated and 420 downregulated proteins (Supplementary Fig. S7, A, right; Supplementary Table S3). Additionally, comparative proteomic analysis revealed 180 proteins uniquely enriched in EVs (Supplementary Fig. S7, B). The GO enrichment analysis of differentially expressed proteins (> 2-fold) showed distinctive patterns in biological process (BP) and cellular components (CC) between EVs and the WCL (Supplementary Fig. S7, C). Three proteomic analyses of C. neoformans EVs have been reported previously (Rodrigues et al., 2008; Wolf et al., 2014; Rizzo et al., 2021), identifying 76, 202, and 1,847 proteins associated with EVs, respectively. We compared the EV protein content between WT and ugg1Δ identified in our study, focusing on overlapped proteins from the three EV proteomic data sets reported. Among 88 commonly reported EV-associated proteins, 29 were downregulated while 16 were overexpressed in ugg1Δ EVs compared to WT EVs (Fig. 8E; Supplementary Table S4). The differential expression pattern of the EV-associated proteins is quite distinctive from that of WCL, revealing significantly detectable changes in EVs. Thus, the absence of functional Ugg1 not only caused morphological alterations and reduced EV numbers but also altered EV protein abundance in C. neoformans.

To further examine possible cargo loading defects, we investigated the presence of GXM in the ugg1Δ EVs, comparing GXM quantities in the total cell extracts from WT, ugg1Δ, and cap59Δ strains (Fig. 8F). Cda1 was used as the representative EV-associated protein as its presence was reported within the C. neoformans EV membrane, despite a substantial portion of Cda1 being secreted through the conventional secretion pathway (Rizzo et al., 2021). GXM polysaccharides were not detected in the EVs released from ugg1Δ, although it was present in the whole-cell extract. Taken together with the reduced quantity of EV-associated proteins, the absence of GXM polysaccharides in ugg1Δ EVs strongly indicates that EV cargo loading is defective when the ERQC is dysfunctional in C. neoformans.

Discussion

As glycoproteins pass through the ER-Golgi secretory pathway, their N-glycans undergo extensive modifications in association with protein quality control. The ERQC system is highly conserved among eukaryotes but also diverged with distinct and species-specific features. The ERQC composition of C. neoformans is unique in that it possesses UGGT but lacks the three glucosyltransferases necessary for the addition of glucose residues to the core precursor N-glycans (Park et al., 2012). Additionally, it lacks CRT, and carries multiple α1,2-mannosidases. In this study, we investigated the molecular features and functions of the N-glycan-dependent ERQC in C. neoformans, which generates unique N-glycan precursors without glucose addition, and shorter in length than those of most eukaryotes. Our data strongly suggest that despite the incomplete composition of the ERQC components, the UGGT-centered ERQC plays pivotal roles in cellular fitness, and particularly in the EV-mediated extracellular transport, which is crucial for pathogenicity.

The virulence of the C. neoformans ugg1Δ mutant was almost abolished in mice (Fig. 4D), a phenotype consistent with the observed defects in key virulence determinants, such as the capsule and melanin, as well as poor growth at 37 °C (Fig. 3B; Fig. 4A and 4B). The mns1Δ101Δ mutant displayed intermediate phenotypes between those of the WT and ugg1Δ. Our data from the UPR induction analysis and comparative transcriptomic analysis strongly indicate that the ERQC mutation generates ER stress; this accounts for significant similarity of the defective phenotypes of the C. neoformans ERQC mutants to those reported in other mutants that exhibit defective protein folding in the ER, particularly in terms of increased stress sensitivity and decreased virulence. Connections between ER stress and thermotolerance have previously been established in C. neoformans, as growth at 37 °C requires key ER protein chaperones and protein processing machinery, including components of the UPR signaling pathway Ire1 and the ER stress-responsive transcription factor Hxl1 (Cheon et al., 2011; Havel et al., 2011; Jung et al., 2013). Other mutations causing defects in ER function in C. neoformans, such as mutants lacking DNJ1 (an ER J-domain containing co-chaperone) and CNE1 (an ER chaperone), resulted in growth inhibition. The dnj1Δ mutant strain displayed impaired elaboration of virulence factors, such as the exopolysaccharide capsule and extracellular urease activity, when cultured at human body temperature (Horianopoulos et al., 2021). Altogether with our data from the ERQC mutants, these findings strongly support the notion that maintaining ER homeostasis is crucial for survival and virulence factor production at elevated temperatures.

It is quite notable that the C. neoformans ugg1Δ EVs exhibited defective loading of GXM and many protein cargo, alongside a decrease in the number and changes in the size distribution (Fig. 8). Combined with compromised cell fitness, these defects in EV biogenesis and cargo loading, even under normal growth conditions, likely contribute to the complete loss of virulence owing to the defective export of virulence factors in the ugg1Δ mutant. Interestingly, our comparative proteome analysis of culture supernatants (secretome) revealed that the ugg1Δ strain exhibited more pronounced defects in the secretion of proteins lacking signal peptides, which are often linked to unconventional secretion mechanisms, compared to those with canonical secretion signals (Supplementary Fig. S8; Supplementary Table S5). These findings further support the role of ERQC in modulating non-conventional secretion pathways.

It is intriguing how the absence of Ugg1, leading to ERQC defects, results in defective EV biogenesis and cargo loading in C. neoformans. In fungi, EV production and release occur through the maturation of endosomes into multivesicular bodies (MVBs), which are directed to the cell surface. Upon fusion with the plasma membrane, MVBs release vesicles into the extracellular environment, functioning as an unconventional secretion pathway (Oliveira et al., 2010). MVBs formation relies on the functionality of the endosomal sorting complex required for transport (ESCRT), a highly intricate pathway involving a series of finely regulated events (Henne et al., 2011). Deletion of several ESCRT complex-related genes involved in EV transport directly impacts the cryptococcal capsule, notably resulting in reduced capsule size. Specifically, the C. neoformans vps27Δ strain showed altered EV size distribution, reduced capsule dimensions, defects in laccase export to the cell wall, and poor extracellular export of urease (Park et al., 2020). Other regulators of unconventional secretion are also linked to EV biogenesis. For example, the Golgi reassembly and stacking proteins (GRASPs) regulate EV cargo and dimensions in C. neoformans (Peres de Silva et al., 2018). A graspΔ mutant strain produced EVs with dimensions that significantly differed from those produced by WT cells, along with attenuated virulence and abnormal RNA composition. Moreover, the GRASP protein Grasp homology 1 (Grh1) serves as a chaperone that directly influences EV cargo (Malhotra, 2013; Peres da Silva et al., 2018). Autophagy regulators, which participate in EV formation in other eukaryotes, also play a role in cryptococcal EV formation. An atg7Δ strain manifests hypovirulence, and EVs produced by this strain show slightly different RNA composition compared with that of the WT cells (Oliveira et al., 2016). Notably, our data of the transcriptome and proteome analysis did not show significant changes in the ESCRT complex and GRASPs.

In addition to protein folding and secretion, the ER is crucial for regulating lipid metabolism (Moncan et al., 2021). Consequently, ER stress significantly impacts lipid and sterol synthesis, although some of these mechanisms are yet to be clarified. LDs aid the UPR and ERAD in degrading misfolded proteins during ER stress in S. cerevisiae (Garcia et al., 2021). In the present study, we observed a drastic increase in LDs in the ugg1Δ mutant (Supplementary Fig. S6, A), consistent with the previous hypothesis that LDs help maintaining ER homeostasis. Furthermore, we observed increased number of vacuoles in ugg1Δ (Supplementary Fig. S6, B). This may thus indicate abnormal lipid homeostasis caused by the ERQC defects, which could, in turn, affect EV biogenesis. The importance of lipids and membrane regulators in proper EV formation and GXM export has been suggested in a previous study on the Apt1 flippase in C. neoformans (Rizzo et al., 2018). Additionally, ergosterol is essential for membrane fluidity, permeability, and protein transport (Ermakova and Zuev, 2017). Erg6 is involved in the ergosterol biosynthesis pathway and was identified as essential for the trans-Golgi network transport of proteins (Proszynski et al., 2005; Nes et al., 2009). In C. neoformans, the erg6Δ mutant released EVs with a significantly larger diameter than those of the WT, carrying increased levels of proteins and sterols, highlighting the role of ergosterol in cryptococcal EV biogenesis (Oliveira et al., 2020). Notably, the ugg1Δ mutant exhibited loss of microvesicles, which are derived from the plasma membrane (Fig. 8A). Our preliminary analysis data on the surface lipid fraction of the ugg1Δ mutant indicated detectable alteration of sphingolipids and sterol profiles. Altogether, these findings suggest that ER stress caused by misfolded glycoprotein accumulation in ERQC-defective mutants may alter lipid composition, which could affect EV biogenesis.

A recent study using kidney cells presented strong evidence for a key role of ER stress in modulating EV biogenesis, demonstrating that ER stress decreases exosome production in mice (Fukuoka et al., 2023). They showed that T-cadherin is downregulated by ER stress through IRE1α activation at mRNA and protein levels and that ER stress decreases EV production through adiponectin/T-cadherin-independent way, which may involve interferon pathway activation in mice. We similarly observed induced activity of the ER stress sensor Ire1 in ugg1Δ, even under normal growth conditions, suggesting a possible association between the Ire1-mediated UPR pathway and EV biogenesis (Fig. 9). Additionally, glycosylation regulates the biogenesis of small EVs and affects protein cargo loading efficiency in melanoma cells (Harada et al., 2020). This suggests that the altered N- glycosylation observed in C. neoformans ugg1Δ may influence cargo loading of certain EV-targeting glycoproteins. Despite their findings not being conducted in fungi, they provide valuable guides on the investigation of conserved mechanisms shared with C. neoformans. Further studies to investigate the mechanisms underlying the ERQC-mediated modulation of EV biogenesis and cargo loading in C. neoformans will provide insights into understanding the regulation, production, composition, and diversity of fungal EVs, enabling a better understanding of their biological function. Expanding our knowledge on pathogenic fungal EVs would pave the way for utilizing native or engineered EVs as promising candidates for therapeutic applications, including fungal infection diagnosis and vaccine development.

Figure 9.

Materials and Methods

Strains, culture conditions, plasmids, and primers

The C. neoformans strains that were constructed and used in this study are listed in Supplementary Table S1A. The plasmids and primers used are listed in Supplementary Tables S1B and S1C, respectively. The construction of the deletion mutants is described in the Supplementary Information. The strains were typically cultured in YPD medium (1% yeast extract, 2% bacto peptone, and 2% glucose) at 30 °C with shaking (220 rpm). C. neoformans transformants were selected by culturing on YPD solid medium containing 100 μg/ml nouseothricin (Jena Bioscience, Germany; indicated as YPD_NAT), YPD solid medium with 100 μg/ml hygromycin B (Sigma-Aldrich, USA; indicated as YPD_HyB), or YPD solid medium with 100 μg/ml G418 (Duchefa, Netherlands; indicated as YPD_NEO). For capsule induction, C. neoformans cells were cultured in liquid Sabouraud dextrose medium (Difco) at 30 °C for 16 h and incubated in 10% Sabouraud dextrose medium (pH 7.3) supplemented with 50 mM morpholinepropanesulfonic acid (MOPS) at 30 °C for 2 days. For melanin production analysis, the cells were spotted onto an L-DOPA agar plate (7.6 mM L-asparagine monohydrate, 5.6 mM glucose, 22 mM KH₂PO₄, 1 mM MgSO₄·7H₂O, 0.5 mM L-DOPA, 0.3 mM thiamine-HCl, and 20 nM biotin) and incubated at 30 °C or 37 °C for 2 days.

Transmission electron microscopy

C. neoformans cells were cultured either at an initial optical density of 600 nm (OD₆₀₀) of 0.2 at 30 °C in YPD medium until OD₆₀₀ reached 0.8, or for 2 days in 10% Sabouraud media at 30 °C. The cell pellets were washed twice in PBS and fixed for 12 h in 2% glutaraldehyde-2% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). The cells were washed in 0.1 M phosphate buffer and post-fixed with 1% OsO₄ in 0.1 M phosphate buffer for 2 h. The cells were dehydrated using an ascending ethanol series (50, 60, 70, 80, 90, 95, 100%) for 10 min each and infiltrated with propylene oxide for 10 min. The specimens were embedded with a Poly/Bed 812 kit (Polysciences Inc., USA) and polymerized in an electron microscope oven (TD-700, DOSAKA, Japan) at 65 °C for 12 h. The block was cut into 200-nm semi-thin sections with a diamond knife in the ultramicrotome and stained with toluidine blue for observation using optical microscopy. The region of interest was further cut into 80 nm thin sections using the ultramicrotome, placed on copper grids, double stained with 3% uranyl acetate for 30 min and 3% lead citrate for 7 min, and imaged using a transmission electron microscope (JEM-1011, JEOL, Tokyo, Japan) equipped with a Megaview III CCD camera (Soft imaging system-Germany) at the acceleration voltage of 80 kV.

Capsule transfer and shedding analysis

Capsule transfer assays were performed as described previously (Reese and Doering, 2003). Briefly, conditioned medium (CM) was prepared as a source of GXM by culturing the respective strains for 5 days in YPD medium, followed by filtering and storing the culture supernatant at 4 °C. The acceptor strains were cultured overnight at 30 °C in YPD medium. In total, 2 × 10⁶ acceptor cells were incubated with 1 µl of CM for 1 h at room temperature under rotation (18 rpm) and washed twice with PBS. Capsule acquisition was visualized by incubating the cells with an anti-GXM (18B7) antibody conjugated with AlexaFluor 488 (ThermoFisher Scientific, USA) for 1 h at 37 °C, and observing under an Eclipse Ti-E fluorescence microscope (Nikon, Japan), equipped with a Nikon DS-Qi2 camera. The images were processed using the NIS-elements microscope imaging software (Nikon, Japan). Capsule shedding analysis was performed using a modified previously described protocol (Yoneda and Doering, 2008). The respective strains were cultured in 10% Sabouraud media for 2 days, and the culture supernatant was sterile filtered. Enzyme denaturation was performed by subjecting the filtrate to heating at 70 °C for 15 min, followed by centrifugation at 13,000 rpm for 3 min. Intracellular GXM analysis was performed by resuspending the pellets in TNE buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 5 mM EDTA) with the same volume of glass beads (425–600 µm in diameter, Sigma-Aldrich, USA). The cells were disrupted four times for 15 s at 5,000 rpm using a Precellys 24 Tissue Homogenizer (Bertin Technologies, France), followed by centrifugation at 16,000 rpm for 5 min at 4 °C. Next, 10 µl of the supernatant of either secreted or intracellular polysaccharide was mixed with 6X loading dye and run on a 0.6% certified megabase agarose (Bio-rad, USA) gel in 0.5X TBE (44.5 mM Trisma base, 44.5 mM boric acid, 1 mM EDTA [pH 8.0]) at 25 V for 16 h. The polysaccharides were transferred onto a nylon membrane using the southern blotting technique. The membrane was air-dried, blocked using 5% skim milk, and treated overnight with 2 µg/ml 18B7 antibody. After washing, the membrane was incubated with an anti-mouse peroxidase-conjugated secondary antibody and subjected to detection using chemiluminescence.

Detection of enzymatic activities in intracellular and secretary fractions

Biochemical enzymatic activities of acid phosphatase, urease, and laccase in the intracellular and secreted fractions were assayed spectrophotometrically, using previously described methods (Lev et al., 2014; Fu et al., 2018; de Sousa et al., 2022). Acid phosphatase activity was assayed by culturing cells in MM-KCL medium (0.5% KCl, 15 mM glucose, 10 mM MgSO₄·7H₂O, 13 mM glycine, and 3 µM thiamine) for 3 h at 30 °C. The culture supernatant and soluble cell lysate were allowed to react with 2.5 mM p-nitrophenyl phosphate (pNPP) for 30 min at 37 °C. Urease activity was determined after culturing the cells in Rapid Urea Broth (RUH broth) for 24 h at 30 °C and subjecting the culture supernatant and soluble cell lysate to incubation with phenol red. Laccase activity was assayed by culturing the cells in asparagine salts media (7.6 mM L-asparagine, 0.1% glucose, 22 mM KH₂PO₄, 1 mM MgSO₄.7H₂O, 0.3 mM thiamine-HCl, and 20 nM biotin) for 48 h at 30 °C and allowing the supernatant and soluble cell lysate to react overnight with 10 mM L-DOPA. The reactions were quantified by measuring OD₄₂₀ (acid phosphatase), OD₅₇₀ (urease), or OD₄₈₀ (laccase). Cellulase activity was determined by culturing the cells in cellulose media (0.1% NaNO₃, 0.1% K₂HPO₄, 0.1% KCl, 0.5% MgSO₄, 0.5% yeast extract, 0.1% glucose, and 0.5% low viscosity carboxymethyl cellulose) for 24 h at 30 °C. The culture supernatants were concentrated using an Amicon tube (30 kDa cutoff, Sigma-Aldrich, USA). The concentrated supernatants and soluble cell lysates were allowed to react at 40 °C for 10 min with the substrate 4,6-O-(3-ketobutylidene)-4-nitrophenyl-β-D-cellopentaoside (BPNPG5), which was provided in the cellulase assay kit (CellG5 Method, Megazyme, Ireland). The reaction was terminated by adding 2% [w/v] Tris buffer (pH 10), and the absorbance of 4-nitrophenol was measured at 400 nm. Cellulase activity (CellG5 Units/ml) was calculated as indicated in the cellulase assay kit. α-amylase activity was measured using the α-Amylase Activity Colorimetric Assay Kit (Biovision Technologies, USA) according to the manufacturer’s instructions. Yeast cells were cultured overnight in YPD medium at 30 °C, and the culture supernatants were concentrated using an Amicon tube (30 kDa cutoff, Sigma-Aldrich, USA). The concentrated supernatants and soluble cell lysates were incubated for 1 h at 25 °C in assay buffer with the substrate ethylidene-pNP-G7, which was provided in the kit. OD₄₀₅ was measured. All the activity analysis results were normalized according to cell density (OD₆₀₀).

Purification, nanoparticle tracking analysis (NTA), and Cryo-TEM imaging of extracellular vesicles

EV purification was performed according to a previously published protocol (Reis et al., 2019) with modifications. One loop of cells was inoculated into 10 ml of liquid YPD and incubated at 30 °C for 24 h with shaking (220 rpm). The cells were washed twice with 10 ml of PBS, counted, and diluted in PBS to a density of 3.5 × 10⁷ cells/ml. Aliquots of cell suspension (300 μl) were spread onto synthetic dextrose (SD) solid medium plates and incubated for 24 h at 30 °C. The cells were carefully recovered from each plate using an inoculation loop, gently resuspended in 30 ml PBS, and pelleted through centrifugation at 4,000 rpm for 10 min at 4 °C. The supernatant was collected and centrifuged again at 15,000 ×g for 15 min at 4 °C. Then, the supernatant was filtered through 0.45-μm syringe filters and ultracentrifuged at 100,000 ×g for 1 h at 4 °C (MLA-50 fixed angle rotor, Beckman Coulter, Germany). The supernatant was discarded, and the EV pellets were collected and resuspended in 1 ml of PBS for immediate use or stored at −80 °C for further experiments. The samples were diluted 100-fold, and EV sizes were measured using an NTA instrument (NFEC-2024-03-295455, Nanosight Pro, Malvern Panalytical, Netherlands) coupled to a 532-nm laser (Malvern Panalytical, Netherlands), SCMOS camera (Hamamatsu Photonics. Japan), and syringe pump (Malvern Panalytical, Netherlands). The data were analyzed using the NS Xplorer software (v1.1.0.6, Malvern Panalytical, Netherlands).

Cryo-TEM imaging was performed by loading purified EVs (3 µl) onto a lacey carbon grid (Lacey Carbon, 300mesh Cu, Ted Pella Inc., USA), which was glow discharged at 15 mA for 60 s. The sample-loaded grid was blotted for 3 s at 15 °C and 100% humidity and immediately plunge-frozen in liquid ethane. The process was performed by Vitrobot Maek IV (Thermofisher Scientific, USA, SNU, CMCI). The frozen grids were imaged using TEM (JEM-2100F, JEOL, Japan); the temperature of the grid was maintained at approximately −180 °C at an acceleration voltage of 200 keV. The images were recorded using an ultrascan 1000 electron detector.

Proteomic analysis based on mass spectrometry and data processing

The samples for proteomic analysis were prepared as described in Supplementary Information. The labeled peptides (total, 100 µg proteins) were combined prior to offline basic reverse-phase liquid chromatographic (bRPLC) fractionation. Linear gradient was performed using buffer A (10 mM TEAB in water) and buffer B (10 mM TEAB in 90% acetonitrile), and 10 fractions were analyzed in total using am LC-MS/MS system. The samples were dissolved in 0.1% formic acid using an UltiMate 3000 RSLCnano system and analyzed using an Orbitrap Eclipse Tribrid mass spectrometer (Thermo Fisher Scientific, USA). All MS raw files were converted into mzML and ms2 file formats using the MSConvert (version 3.0.20033) software. The C. neoformans proteome was determined by downloading a protein FASTA file from Uniprot (http://uniprot.org), which included 7,492 reviewed (Swiss-Prot) and unreviewed (TrEMBL) proteins entries. A proteome search database with reversed sequence and contaminants in the Integrated Proteomics Pipeline (IP2, version 5.1.2, Integrated Proteomics Applications Inc., San Diego, CA) was generated. The proteome search from 20 ms2 files was performed with IP2 and its following parameters, followed by evaluation for a false discovery rate (FDR) using IP2 and Proteininferencer (version 1.0, Integrated Proteomics Applications Inc., USA). Protein quantification and statistical analysis for discovery of differentially expressed proteins (DEPs) was performed using the ms2 files with tandem mass tag (TMT) reporter ions using an in-house program coded using Python 3.8, where t-test and Pearson’s correlation analysis between comparison samples was performed using the scikit-learn (version 0.23.2), Scipy (version 1.6.0), and statsmodels (0.12.1) Python libraries.

Acknowledgements

We thank Arturo Casadevall for providing the 18B7 antibody, Jennifer K. Lodge for providing the anti-Cda1 antibody, and Ji-Yeon Kang for technical assistance with MALDI-TOF analysis.

Additional information

Data availability

The raw RNA sequencing data have been submitted to the NCBI GEO database under accession no. GSE254772.

Funding

This study was supported by the National Research Foundation of Korea (Grant nos. NRF-2022R1A2C1012699, NRF2018R1A5A1025077, and RS-2023-00212663) and by the Korea Institute of Marine Science & Technology Promotion (Grant no. RS-2024-00405273).

Author Contributions

Catia Mota: Investigation, Construction of mutants, Performing all virulence analysis, Data curation, and Writing—original draft; Kiseung Kim: Construction of mutants and Performing phenotype analysis; Ye Ji Son: Transcriptome analysis and Validation; Eun Jung Thak and Su-Bin Lee: Glycan analysis; Ju-El Kim and Jeong-Kee Yoon: NTA analysis; Min-Ho Kang: Cryo-TEM analysis; Heeyoun Hwang: Proteome analysis; Yong-Sun Bahn and J. Andrew Alspaugh: Supervision, Data curation, Writing—review, and editing; Hyun Ah Kang: Conceptualization, Project administration, Supervision, Data curation, Writing—original draft, review, and editing.

Additional files

Supplementary material and methods. (a) Construction of C. neoformans ugg1Δ and UGG1 complementation strains. (b) Construction of C. neoformans mns1Δ, mns101Δ, mns1Δ101Δ, and respective complementation strains. (c) Construction of C. neoformans mnl1Δ, mnl2Δ, and mnl1Δmnl2Δ strains. (d) Construction and localization analysis of C. neoformans GFP-Ugg1, Mns1-GFP, and Mns101-GFP fusion proteins. (e) HPLC and MALDI-TOF-based N-glycan structure analysis. (f) Animal study and in vitro survival analysis. (g) Subcellular fractionation and western blot analysis of Cda1. (h) RNA preparation, qRT-PCR, and RNA-sequencing. (i) Sample preparation for proteomic analysis.

Supplementary figures. (a) Figure S1. Evolutionary conservation and divergence of endoplasmic reticulum quality control (ERQC)-related gene homologues predicted in different eukaryote organisms. (b) Figure S2. Domain analysis of putative C. neoformans ERQC-associated proteins. (c) Figure S3. Phenotype characterization of C. neoformans α1,2-mannosidases. (d) Figure S4. Subcellular localization of GFP-tagged Ugg1, Mns1, and Mns101. (e) Figure S5. Fungal burden of mice infected with C. neoformans ERQC-related mutant cells. (f) Figure S6. Lipid and vacuole staining of WT and ugg1Δ strains. (g) Figure S7. Comparative analysis of whole-cell lysate (WCL) and EV proteomes between WT and ugg1Δ strains. (h) Figure S8. Comparative analysis of whole-cell lysate (WCL) proteome and secretome between WT and ugg1Δ strains.

Table S1. (A) List of strains used in this study. (B) List of plasmids used in this study. (C) List of primers used in this study.

Table S2. (A) Up-regulated (> 2-fold) Cryptococcus neoformans genes in ugg1Δ than that of wild type (WT) cultivated in yeast extract peptone dextrose (YPD) at 30 °C. (B) Down-regulated (> 2-fold) Cryptococcus neoformans genes in ugg1Δ than that of wild type (WT) cultivated in YPD at 30 °C.

Table S3. Excel file containing the list of proteins identified in the proteomic analysis of the wild type (WT) and ugg1Δ extracellular vesicles (EVs) and whole-cell lysates (WCL).

Table S4. Representative extracellular vesicle (EV)-associated proteins commonly detected in this study and previously reported Cryptococcus neoformans EV proteome data sets.

Table S5. Excel file containing the list of proteins identified in the whole-cell lysate (WCL) proteome and secretome analysis in the wild type (WT) and ugg1Δ strains.