Secretion of the fungal toxin candidalysin is dependent on conserved precursor peptide sequences.

The opportunistic fungal pathogen Candida albicans damages host cells via its peptide toxin, candidalysin. Before secretion, candidalysin is embedded in a precursor protein, Ece1, which consists of a signal peptide, the precursor of candidalysin and seven non-candidalysin Ece1 peptides (NCEPs), and is found to be conserved in clinical isolates. Here we show that the Ece1 polyprotein does not resemble the usual precursor structure of peptide toxins. C. albicans cells are not susceptible to their own toxin, and single NCEPs adjacent to candidalysin are sufficient to prevent host cell toxicity. Using a series of Ece1 mutants, mass spectrometry and anti-candidalysin nanobodies, we show that NCEPs play a role in intracellular Ece1 folding and candidalysin secretion. Removal of single NCEPs or modifications of peptide sequences cause an unfolded protein response (UPR), which in turn inhibits hypha formation and pathogenicity in vitro. Our data indicate that the Ece1 precursor is not required to block premature pore-forming toxicity, but rather to prevent intracellular auto-aggregation of candidalysin sequences.

A New Phenotype in Candida-Epithelial Cell Interaction Distinguishes Colonization- versus Vulvovaginal Candidiasis-Associated Strains.

Vulvovaginal candidiasis (VVC) affects nearly 3/4 of women during their lifetime, and its symptoms seriously reduce quality of life. Although Candida albicans is a common commensal, it is unknown if VVC results from a switch from a commensal to pathogenic state, if only some strains can cause VVC, and/or if there is displacement of commensal strains with more pathogenic strains. We studied a set of VVC and colonizing C. albicans strains to identify consistent in vitro phenotypes associated with one group or the other. We find that the strains do not differ in overall genetic profile or behavior in culture media (i.e., multilocus sequence type [MLST] profile, rate of growth, and filamentation), but they show strikingly different behaviors during their interactions with vaginal epithelial cells. Epithelial infections with VVC-derived strains yielded stronger fungal proliferation and shedding of fungi and epithelial cells. Transcriptome sequencing (RNA-seq) analysis of representative epithelial cell infections with selected pathogenic or commensal isolates identified several differentially activated epithelial signaling pathways, including the integrin, ferroptosis, and type I interferon pathways; the latter has been implicated in damage protection. Strikingly, inhibition of type I interferon signaling selectively increases fungal shedding of strains in the colonizing cohort, suggesting that increased shedding correlates with lower interferon pathway activation. These data suggest that VVC strains may intrinsically have enhanced pathogenic potential via differential elicitation of epithelial responses, including the type I interferon pathway. Therefore, it may eventually be possible to evaluate pathogenic potential in vitro to refine VVC diagnosis. IMPORTANCE Despite a high incidence of VVC, we still have a poor understanding of this female-specific disease whose negative impact on women's quality of life has become a public health issue. It is not yet possible to determine by genotype or laboratory phenotype if a given Candida albicans strain is more or less likely to cause VVC. Here, we show that Candida strains causing VVC induce more fungal shedding from epithelial cells than strains from healthy women. This effect is also accompanied by increased epithelial cell detachment and differential activation of the type I interferon pathway. These distinguishing phenotypes suggest it may be possible to evaluate the VVC pathogenic potential of fungal isolates. This would permit more targeted antifungal treatments to spare commensals and could allow for displacement of pathogenic strains with nonpathogenic colonizers. We expect these new assays to provide a more targeted tool for identifying fungal virulence factors and epithelial responses that control fungal vaginitis.

Augmented Enterocyte Damage During Candida albicans and Proteus mirabilis Coinfection.

The human gut acts as the main reservoir of microbes and a relevant source of life-threatening infections, especially in immunocompromised patients. There, the opportunistic fungal pathogen Candida albicans adapts to the host environment and additionally interacts with residing bacteria. We investigated fungal-bacterial interactions by coinfecting enterocytes with the yeast Candida albicans and the Gram-negative bacterium Proteus mirabilis resulting in enhanced host cell damage. This synergistic effect was conserved across different P. mirabilis isolates and occurred also with non-albicans Candida species and C. albicans mutants defective in filamentation or candidalysin production. Using bacterial deletion mutants, we identified the P. mirabilis hemolysin HpmA to be the key effector for host cell destruction. Spatially separated coinfections demonstrated that synergism between Candida and Proteus is induced by contact, but also by soluble factors. Specifically, we identified Candida-mediated glucose consumption and farnesol production as potential triggers for Proteus virulence. In summary, our study demonstrates that coinfection of enterocytes with C. albicans and P. mirabilis can result in increased host cell damage which is mediated by bacterial virulence factors as a result of fungal niche modification via nutrient consumption and production of soluble factors. This supports the notion that certain fungal-bacterial combinations have the potential to result in enhanced virulence in niches such as the gut and might therefore promote translocation and dissemination.