Dimitrios Sofras

Echinocandins are the first-line treatment for invasive fungal infections and work by inhibiting the fungal β-1,3-glucan synthase Fks1. Recently, a first-in-class triterpenoid Fks1 inhibitor, ibrexafungerp, was approved for clinical use. The extent to which the difference between these drug classes alters the landscape of resistance mutations in Fks1 is unknown. We used deep mutational scanning (DMS) to quantify the effects of all single-amino-acid variants spanning 159 residues across the three canonical Fks1 resistance hotspots and flanking regions in Saccharomyces cerevisiae. Drug selections revealed widespread position-specific cross-resistance and differential resistance, including an ibrexafungerp-specific resistance cluster downstream of hotspot 2 that defines a novel triterpenoid-specific resistance hotspot. We defined this as hotspot 4 and redefined the boundaries of the remaining three canonical echinocandin hotspots, based on the mutational landscape of our DMS analyses. Susceptibility testing of individual FKS1 variants of S. cerevisiae and Candidozyma auris strains confirmed conserved class-specific and cross-resistance patterns. Structure-guided ligand placement, molecular dynamics-based simulated annealing, and DMS-informed pose selection linked drug-specific phenotypes to distinct interaction networks, providing a molecular rationale for the observed differential resistance patterns. These data provide a framework for predicting antifungal resistance, informing treatment strategies and future drug design.

The emergence of multidrug-resistant fungal pathogens from urinary tract infections (UTIs) poses a growing challenge in clinical settings. Here, we report a case of a complicated UTI caused by Nakaseomyces glabratus ( Candida glabrata ) that progressed to urosepsis, leading to the emergence of an isolate carrying simultaneous loss-of-function mutations in ERG3 and ERG11, and abrogated ergosterol biosynthesis. Together with a missense mutation in FUR1 —likely responsible for 5-fluorocytosine resistance—this constellation confers resistance to all viable UTI antifungals: azoles, amphotericin B, and flucytosine. Engineered ERG3 Δ + ERG11 Δ strains recapitulated this multidrug resistance and revealed profound fitness costs that come with it, challenging the assumption that high-cost mutations are unlikely to persist during infection. Among fitness trade-offs, we detected collateral sensitivity to nitroxoline, a commonly used urinary tract antibiotic with potent antifungal activity and a unique mechanism of action. This study provides the first clinical evidence of an elusive mechanism of hyper-multidrug resistance in N. glabratus and highlights nitroxoline as a promising repurposing agent for treating multidrug-resistant fungal infections of the urinary tract. IMPORTANCE: Evolutionary theory states that fitness determines survival. In a drug-treatment environment, resistance increases fitness, but it often comes at a cost, such as slower growth or reduced stress tolerance. If these costs are too severe, they can undermine virulence, making resistance unlikely to persist. Our study challenges this assumption. We describe the first clinical case of Nakaseomyces glabratus evolving multidrug resistance through loss-of-function mutations in ERG3 and ERG11 , despite severe fitness trade-offs. This case suggests that certain infection niches, such as the urinary tract, can provide conditions where even highly impaired yet resistant strains persist under strong antifungal pressure. Importantly, we show that this extreme resistance induces collateral sensitivity to nitroxoline, a urinary tract infection antibiotic with potent antifungal activity and a unique mechanism of action. These findings open promising therapeutic avenues to counter multidrug-resistant fungal infections of the urinary tract.

Candidozyma auris is an emerging multidrug-resistant fungal pathogen that poses a major public-health challenge owing to high mortality and the limited efficacy of current therapies. Echinocandins, which inhibit β-glucan synthesis, are first-line therapy for invasive C. auris infection; however, resistance to this class is rising, underscoring the urgent need for new antifungal targets. Here we show that enzymes in the trehalose-biosynthetic pathway regulate stress responses, antifungal resistance/tolerance and virulence in C. auris. The tps2Δ strain displays heightened susceptibility to echinocandins, whereas tps1Δ and tps1Δ tps2Δ strains show resistance and tolerance comparable to wild type (WT). Mechanistically, the tps2Δ strain accumulates trehalose 6-phosphate (T6P), which inhibits hexokinase activity and reduces the flux of glucose 6-phosphate (G6P) into the chitin biosynthesis pathway, leading to substantially decreased cell wall chitin. During echinocandin exposure, the tps2Δ strain fails to compensate for reduced β-glucan with increased chitin, thereby rendering it highly susceptible to these drugs. In a systemic mouse infection model, deletion of the TPS2 gene results in lower tissue fungal burdens after treatment with caspofungin. Together, these findings identify Tps2 as a potential therapeutic target that can potentiate echinocandin efficacy in C. auris via a distinct mechanism of action.

The Hxk1 protein of Candida albicans phosphorylates N-acetylglucosamine (GlcNAc) which is necessary for various cellular functions, including energy production and chitin synthesis. Further, this protein also regulates morphogenesis independently of its role in GlcNAc catabolism. When HXK1 is deleted, cells are hyperfilamentous on serum-containing medium. Furthermore, Hxk1 translocates to the nucleus in the presence of glucose. To gain a broad understanding of the effect of Hxk1 on gene expression in C. albicans , we performed genome-wide transcriptional profiling of the hxk1 mutant strain by RNA-Seq. The analysis of these RNA-Seq data showed that Hxk1 affects gene expression in both a carbon source-dependent and -independent manner. However, the effect on gene expression occurs via an indirect mechanism, as genome-wide CUT&RUN binding experiments demonstrated that Hxk1 does not bind to the upstream intergenic regions of the differentially expressed genes. Deletion of HXK1 not only resulted in differences in gene expression of genes present in the GlcNAc and galactose regulons, but also in glucose transporter genes, including HGT13 . Hxk1 also negatively influences the expression of virulence-associated genes, including HWP1 , BRG1 , and UME6 . Consequently, an hxk1 mutant strain showed higher toxicity toward gut epithelial cells compared to the WT strain. Furthermore, the hxk1 mutant strain had higher expression levels of SOD4 and SOD5 and showed higher resistance toward H ₂ O ₂ . These findings highlight the multiple functions of Hxk1 in different cellular processes. IMPORTANCE: Candida albicans is a fungus that lives in the human body but does not cause any harm in healthy individuals. However, when the immune system is weakened, C. albicans can spread via the bloodstream all over the body and can lead to severe illness and even death. To infect the human body, multiple proteins hold distinct functions. Hxk1 is one of these proteins. This protein is involved in N-acetylglucosamine (GlcNAc) phosphorylation, as well as hyphae formation and glucose transport. To obtain a complete view of the processes regulated by Hxk1, we performed an RNA-Seq experiment. These data revealed that Hxk1 influences the regulation of genes involved in metabolic and virulence-related processes, such as GlcNAc metabolism, sterol metabolism, and oxidative stress resistance. These findings are important to better understand how C. albicans adapted itself to infect the host.

Echinocandins, which target the fungal β-1,3-glucan synthase (Fks), are essential for treating invasive fungal infections, yet resistance is increasingly reported. While resistance typically arises through mutations in Fks hotspots, emerging evidence suggests a contributing role of changes in membrane sterol composition due to ERG3 mutations. Here, we present a clinical case of Nakaseomyces glabratus ( Candida glabrata ) in which combined mutations in ERG3 and FKS2 , but not FKS2 alone, appear to confer echinocandin resistance. Integrated analyses reveal a recurrent association between Erg3 loss-of-function and echinocandin resistance mediated by Fks variation across Candida species, but exclude Erg3 loss-of-function as an independent resistance mechanism. Advances in Fks structural biology and insights into echinocandin-Fks interactions support a model of epistatic cross-talk between membrane sterols and Fks function. Understanding this interaction is crucial, as it may underlie not only acquired echinocandin resistance but also the broader development of multidrug resistance across major antifungal drug classes. IMPORTANCE: A clinical case in which the combination of variation in a β‑1,3‑glucan synthase-encoding gene ( FKS2 ) and the sterol desaturase-encoding gene ERG3 seems to underlie echinocandin resistance, prompted us to hypothesize that membrane sterol changes may modulate, rather than independently cause, Fks‑linked resistance. We were able to explore this hypothesis due to recent developments in the field, such as the release of the FungAMR database, which enables global co‑occurrence analyses; AI‑driven variant effect predictors such as Evolutionary Scale Modeling (ESM) that can explore the impact of thousands of ERG3 alleles; the cryo‑EM resolution of the Fks1 protein; and the first mechanistic model of echinocandin‑Fks1 binding. Together, these advances provide the structural and computational framework needed to delineate our hypothesis that specific sterol variants might influence β‑1,3‑glucan synthase function and drug binding. Further surveillance of this potentially epistatic interaction can be of significant clinical importance amid rising multidrug‑resistant infections, as overlooking such interactions could lead to under‑calling resistance and misguided therapy.

Echinocandins, which target the fungal β-1,3-glucan synthase (Fks), are essential for treating invasive fungal infections, yet resistance is increasingly reported. While resistance typically arises through mutations in Fks hotspots, emerging evidence suggests a contributing role of changes in membrane sterol composition due to ERG3 mutations. Here, we present a clinical case of Nakaseomyces glabratus ( Candida glabrata ) in which combined mutations in ERG3 and FKS2 , but not FKS2 alone, appear to confer echinocandin resistance. Integrated analyses reveal a recurrent association between Erg3 loss-of-function and echinocandin resistance mediated by Fks variation across Candida species, but exclude ERG3 loss-of-function as an independent resistance mechanism. Advances in Fks structural biology and insights into echinocandin-Fks interactions support a model of epistatic crosstalk between membrane sterols and Fks function. Understanding this interaction is crucial, as it may underlie not only acquired echinocandin resistance but also the broader development of multidrug resistance across major antifungal drug classes.

Background: Nakaseomyces glabratus ( Candida gabrata ) poses a significant clinical challenge due to common drug resistance. We report a case of a complicated urinary tract infection (UTI) progressing to prostatitis and urosepsis, with the emergence of a hyper-multidrug-resistant isolate with low stress tolerance, slow growth and a short life span. This study elucidates the genetic mechanisms and phenotypic characteristics underlying antifungal hyper-resistance with strong fitness trade-offs, and explores potential alternative therapies for resistant UTI’s. Methods: Whole-genome sequencing was performed to identify resistance-associated mutations and gene knock-out strains were generated to assess the relative impact of putative loss-of-function (LoF) mutations on antifungal resistance, fitness and membrane sterol composition. Drug susceptibility testing of the antibiotic nitroxoline and related compounds was conducted to evaluate it as a therapeutic alternative and study the mechanism of action. Findings: Loss-of-function mutations in ERG3 and ERG11 were identified and linked to the accumulation of 4,14-dimethylzymosterol and lanosterol instead of ergosterol. Engineered ERG3Δ+ERG11Δ strains recapitulated the clinical isolate’s hyper-multidrug resistance and associated fitness deficits. While ERG3Δ strains showed no resistance but enhanced thermotolerance, ERG11Δ and ERG3Δ+ERG11Δ strains exhibited multidrug resistance with severe fitness trade-offs. Interestingly, ERG3Δ+ERG11Δ strains showed mild resistance to flucytosine, but an additional FUR1 mutation in the clinical isolate most probably underlies hyper-resistance to flucytosine. The UTI antibiotic nitroxoline demonstrated high antifungal activity against all strains, and the LoF of ERG3 and/or ERG11 induced collateral sensitivity to this drug. Testing of related compounds suggest a mode of action beyond iron chelation. Interpretation: This case demonstrates that hyper-resistant strains of N. glabratus can emerge despite significant fitness costs and persist under prolonged antifungal therapy in specific clinical settings. These findings underscore the importance of vigilant antifungal resistance monitoring and highlight nitroxoline as a promising alternative treatment for complicated fungal UTIs. These results challenge the notion that strains with fitness deficits are clinically irrelevant and emphasize the need for novel therapeutic strategies including repurposed agents.

Candida auris is a growing concern due to its resistance to antifungal drugs, particularly amphotericin B (AMB), detected in 30 to 60% of clinical isolates. However, the mechanisms of AMB resistance remain poorly understood. Here we investigated 441 in vitro- and in vivo-evolved C. auris lineages from 4 AMB-susceptible clinical strains of different clades. Genetic and sterol analyses revealed four major types of sterol alterations as a result of clinically rare variations in sterol biosynthesis genes ERG6, NCP1, ERG11, ERG3, HMG1, ERG10 and ERG12. In addition, aneuploidies in chromosomes 4 and 6 emerged during resistance evolution. Fitness trade-off phenotyping and mathematical modelling identified diverse strain- and mechanism-dependent fitness trade-offs. Variation in CDC25 rescued fitness trade-offs, thereby increasing the infection capacity. This possibly contributed to therapy-induced acquired AMB resistance in the clinic. Our findings highlight sterol-modulating mechanisms and fitness trade-off compensation as risks for AMB treatment failure in clinical settings.

Antifungal drug resistance represents a serious global health threat, necessitating new treatment strategies. Here we investigated collateral sensitivity (CS), in which resistance to one drug increases sensitivity to another, and cross-resistance (XR), in which one drug resistance mechanism reduces susceptibility to multiple drugs, since CS and XR dynamics can guide treatment design to impede resistance development, but have not been systematically explored in pathogenic fungi. We used experimental evolution and mathematical modelling of Candida auris population dynamics during cyclic and combined drug exposures and found that especially CS-based drug cycling can effectively prevent the emergence of drug resistance. In addition, we found that a CS-based treatment switch can actively select against or eradicate resistant sub-populations, highlighting the potential to consider CS in therapeutic decision-making upon resistance detection. Furthermore, we show that some CS trends are robust among different strains and resistance mechanisms. Overall, these findings provide a promising direction for improved antifungal treatment approaches.

Candida auris is an emerging nosocomial fungal pathogen associated with life-threatening invasive disease due to its persistent colonization, high level of transmissibility and multi-drug resistance. Aggregative and non-aggregative growth phenotypes for C. auris strains with different biofilm forming abilities, drug susceptibilities and virulence characteristics have been described. Using comprehensive transcriptional analysis we identified key cell surface adhesins that were highly upregulated in the aggregative phenotype during in vitro and in vivo grown biofilms using a mouse model of catheter infection. Phenotypic and functional evaluations of generated null mutants demonstrated crucial roles for the adhesins Als5 and Scf1 in mediating cell-cell adherence, coaggregation and biofilm formation. While individual mutants were largely non-aggregative, in combination cells were able to co-adhere and aggregate, as directly demonstrated by measuring cell adhesion forces using single-cell atomic force spectroscopy. This co-adherence indicates their role as complementary adhesins, which despite their limited similarity, may function redundantly to promote cell-cell interaction and biofilm formation. Functional diversity of cell wall proteins may be a form of regulation that provides the aggregative phenotype of C. auris with flexibility and rapid adaptation to the environment, potentially impacting persistence and virulence.