Two strains of a dangerous fungus show strikingly different surface proteomes

The fungal pathogen Candida glabrata, now formally known as Nakaseomyces glabratus, has long been studied through the lens of a single reference strain. Yet new research from Jagiellonian University in Kraków suggests that this approach may be hiding crucial biology. In a preliminary comparative proteomic study published in Acta Biochimica Polonica, a team led by Aneta Bednarek and Dorota Satala compared the surface-exposed proteins of the reference strain CBS138 with the widely used clinical isolate BG2, revealing pronounced divergence in what each strain actually presents to its host environment.

C. glabrata is an opportunistic pathogen of considerable medical concern, particularly for immunocompromised individuals. The World Health Organization placed it on its 2022 Fungal Priority Pathogens List as a high-priority target for intensified research. Despite its close phylogenetic relationship with the benign baker’s yeast Saccharomyces cerevisiae, C. glabrata exhibits intrinsic and acquired antifungal resistance, a compact haploid genome with notable plasticity, and the ability to persist within host cells. These traits are increasingly attributed to rapid adaptation through chromosomal rearrangements, point mutations and transcriptional rewiring under antifungal or environmental pressure.

Accumulating evidence indicates substantial intraspecies diversity that may shape virulence, immune interactions and stress adaptation. CBS138 and BG2 differ substantially at the genomic, transcriptomic and phenotypic levels, including chromosomal organization, adhesin gene expression such as EPA1, metabolic flexibility in tryptophan and methionine utilization, and stress response profiles. BG2 also displays enhanced replication within macrophages and increased virulence in the insect model Galleria mellonella relative to CBS138. The two strains further differ in cell wall architecture and carbohydrate exposure, including mannan, β-1,3-glucan and chitin, according to prior electron microscopy and flow cytometry work.

To investigate whether these differences are reflected at the fungal cell surface, the researchers cultured both strains under three biologically relevant growth conditions: as budding yeast-like cells in rich YPD medium, as free-floating aggregates in RPMI 1640 medium that mimics host-like environments, and as mature biofilms formed in dynamic RPMI cultures using a roller-bottle system. Cells were harvested at 24 and 48 hours in each condition. Surface-exposed proteins were then selectively digested using a technique called trypsin shaving, in which the enzyme trypsin is applied briefly to intact cells so that it cleaves only the protein regions accessible from outside the cell wall, releasing peptides that can be identified by liquid chromatography coupled to tandem mass spectrometry.

The mass spectrometry analysis was performed on an ion trap instrument with electrospray ionization, and protein abundance was estimated semi-quantitatively using the spectral abundance factor, which accounts for both spectral counts and protein length. To ensure a conservative comparative dataset, only proteins supported by at least two distinct peptides and detected in at least two out of three biological replicates were included. The results revealed substantial variation in surfaceome composition across time points, culture models and genetic backgrounds. In YPD medium, CBS138 exposed far more detectable surface proteins than BG2, with 54 proteins identified at 24 hours and 65 at 48 hours, compared with only 9 and 12 respectively for the clinical isolate. Under RPMI planktonic conditions, the numbers were lower and more similar between strains, while in the biofilm model CBS138 produced more identifications at 24 hours but the difference narrowed considerably by 48 hours, indicating partial convergence during biofilm maturation.

Venn diagram comparisons generated with FunRich software showed that the overlap between the two strains was limited in all conditions, with the highest degree of overlap observed under RPMI planktonic conditions. Biofilm and YPD conditions showed more pronounced strain-specific differences in the detected surface protein sets. Because surface shaving captures only protease-accessible regions on intact cells, identification counts primarily reflect surface accessibility rather than total proteome coverage, meaning that a protein absent from the list may simply be buried within the wall rather than genuinely absent from the surface.

To place these proteomic differences in a structural context, the team examined cell wall ultrastructure using transmission electron microscopy. For both strains, representative images were acquired after 24 hours of growth under the same three culture conditions, and total cell-wall thickness was quantified by taking multiple independent measurements per cell. A two-way analysis of variance with strain and growth condition as fixed factors identified significant main effects of both condition and strain, as well as a significant interaction between them. Wall thickness differed strongly between growth models, with YPD-grown cells showing the thinnest walls, RPMI planktonic cells intermediate, and biofilm conditions the thickest. BG2 displayed a broader range of values and the most pronounced YPD-to-RPMI increase in thickness, providing a structural framework for interpreting strain- and condition-dependent differences in protease accessibility observed in the shaving experiments.

SAF-based analysis of relative abundance revealed distinct profiles depending on strain, medium and time point. In CBS138, RPMI planktonic cultures were characterized by the presence of canonical adhesins including Epa6 and Awp2, whereas biofilm conditions showed stronger representation of surface-associated proteins involved in host interaction and cell wall organization, such as the yapsin protease Yps3 and the wall protein Scw4. By contrast, BG2 biofilm cultures, particularly at 48 hours, showed increased representation of proteins linked to cell wall architecture and remodelling, including Ecm33, Cwp1.2 and Pir4. Several metabolic enzymes previously described as moonlighting proteins, meaning cytosolic enzymes or chaperones that acquire adhesive or immune-modulatory functions upon externalization, were detected across multiple conditions with variable relative abundance. These included enolase, glyceraldehyde-3-phosphate dehydrogenase, fructose-bisphosphate aldolase and phosphoglycerate kinase, all of which have been shown in prior antibody-based studies to be genuinely exposed at the candidal cell surface and capable of binding human extracellular matrix proteins.

Functional classification of the surface-exposed proteins revealed marked strain- and condition-dependent differences in the distribution of functional categories. In YPD-grown cells, CBS138 displayed a higher proportion of ribosomal, nuclear and metabolic moonlighting proteins, whereas BG2 was enriched in classical cell wall and secreted proteins such as Cwp1, Scw4 and Pir family members. Under RPMI planktonic conditions, both strains showed a substantial contribution of moonlighting proteins, with this feature being particularly evident in BG2. In the biofilm model, CBS138 retained a more mixed surfaceome composition, while BG2 was enriched in remodelling and cell wall integrity factors including Cwp1.2, Ecm33, Pir3 and Pir5, with reduced detection of adhesins. Interestingly, Yps-family aspartyl proteases were detected in the biofilm condition, with Yps3 present in CBS138 biofilms at both time points and in BG2 biofilm at 48 hours. Previous work has linked these proteases not only to cell wall integrity and stress adaptation but also to the degradation of host-derived antimicrobial peptides such as LL-37 and histatin-5, and to interference with epithelial immune signalling that limits neutrophil recruitment.

The apparent paradox that BG2 showed very low surface protein detectability in YPD despite a wall thickness comparable to CBS138 was resolved when wall porosity was considered independently of overall thickness. Prior electron microscopy work has shown that CBS138 is characterized by a more open and highly fibrillar outer mannan layer, with long, loosely organized fibrils and increased exposure of mannan, β-1,3-glucan and chitin, whereas BG2 displays a more compact and densely packed outer layer accompanied by two- to four-fold lower fluorescence signals for these carbohydrate components. This suggests that the relatively thick but porous outer wall of CBS138 may remain readily permeable to proteolytic enzymes, allowing efficient release of peptides derived from GPI-anchored proteins and embedded moonlighting factors, while the compact and tightly cross-linked mannan network of BG2 likely constitutes a physical barrier that restricts protease penetration. Consistent with this model, BG2 showed greater surface protein detectability under selected RPMI-based conditions coinciding with an increase in wall thickness, suggesting that host-mimicking environments can induce substantial remodelling of the BG2 cell wall matrix leading to altered permeability and surface exposure.

Adhesin expression in C. glabrata is known to be tightly regulated by subtelomeric silencing mechanisms mediated by Sir3. EPA1 expression has been shown to be highly heterogeneous in BG2 as a consequence of weakened Sir3-dependent repression, whereas CBS138 exhibits a more uniform adhesin expression profile. These regulatory differences provide a plausible molecular basis for the stronger representation of adhesins in CBS138 surfaceomes and the more condition-dependent adhesin exposure in BG2. Beyond wall architecture itself, strain-dependent immunological consequences of mannan backbone disruption have been reported, with BG2 mutants displaying reduced uptake and intracellular survival in macrophages yet triggering stronger GM-CSF secretion, while CBS138 mutants more strongly activated the EphA2 receptor through enhanced β-glucan exposure. These immunological differences highlight that wall remodelling and glycan masking do not simply affect surface accessibility but profoundly shape host recognition and inflammatory signalling.

Collectively, the findings support the view that CBS138 and BG2 adopt distinct surfaceome-level responses to host-like environments, with CBS138 maintaining a broader multifunctional profile enriched in adhesion- and interaction-related proteins, and BG2 shifting more strongly toward wall remodelling and reinforcement. These mechanistic differences align with previously reported phenotypic traits, including the enhanced macrophage replication and increased virulence of BG2, and emphasize that intraspecies variation in C. glabrata can translate into markedly different surface-mediated pathogenic strategies. The authors note that while shaving-based proteomics is highly informative for defining the protease-accessible surfaceome, it does not by itself provide definitive proof of cell wall localization for each identified protein, particularly for newly detected atypical or moonlighting factors, and that complementary orthogonal approaches on intact cells should verify their exact mode of association with the wall.

In conclusion, the study shows that C. glabrata surfaceome composition is highly plastic and strongly dependent on both strain identity and growth environment. The parallel variation in cell wall thickness provides a structural basis for differences in protease accessibility and surface protein detection. By extending surfaceome analysis beyond standard YPD conditions and incorporating ultrastructural evidence, this preliminary work highlights the critical impact of intraspecies diversity on surface architecture and underscores the necessity of considering strain-specific traits when interpreting virulence- or immunity-related phenotypes. As the authors emphasize, these findings reinforce the concept that CBS138 and BG2 differ not only in their structural and proteomic surface landscapes but also in the way these architectures are sensed by the immune system, ultimately dictating strain-specific trajectories of virulence and host adaptation. For a pathogen listed among the world’s most dangerous fungi, such strain-level nuance may prove essential for designing therapeutics that target the fungal cell surface effectively.

Subject of Research: Divergence in surface protein exposure between reference and clinical-derived Candida glabrata (Nakaseomyces glabratus) strains (CBS138 vs. BG2) – a preliminary proteomic perspective

Article Title: Divergence in surface protein exposure between reference and clinical-derived Candida glabrata (Nakaseomyces glabratus) strains (CBS138 vs. BG2) – a preliminary proteomic perspective

Article References: Bednarek, A., Barczyk-Woznicka, O., Karkowska-Kuleta, J., Pyza, E., Rapala-Kozik, M., & Satala, D. (2026). Divergence in surface protein exposure between reference and clinical-derived Candida glabrata (Nakaseomyces glabratus) strains (CBS138 vs. BG2) – a preliminary proteomic perspective. Acta Biochimica Polonica, 73, Article 16376. https://doi.org/10.3389/abp.2026.16376

Image Credits: AI Generated

DOI: 10.3389/abp.2026.16376

Keywords: Candida glabrata, Nakaseomyces glabratus, surfaceome, proteomics, cell wall, biofilm, Divergence, surface, protein, exposure, between, reference