Investigation of the defective growth pattern and multidrug resistance in a clinical isolate of Candida glabrata using whole-genome sequencing and computational biology applications

Abstract

Candida glabrata is increasingly isolated from blood cultures, and multidrug-resistant isolates have important implications for therapy. This study describes a cholesterol-dependent clinical C. glabrata isolate (ML72254) that did not grow without blood (containing cholesterol) on routine mycological media and that showed azole and amphotericin B (AmB) resistance. Matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) and whole-genome sequencing (WGS) were used for species identification. A modified Etest method (Mueller-Hinton agar supplemented with 5% sheep blood) was used for antifungal susceptibility testing. WGS data were processed via the Galaxy platform, and the genomic variations of ML72254 were retrieved. A computational biology workflow utilizing web-based applications (PROVEAN, AlphaFold Colab, and Missense3D) was constructed to predict possible deleterious effects of these missense variations on protein functions. The predictive ability of this workflow was tested with previously reported missense variations in ergosterol synthesis genes of C. glabrata. ML72254 was identified as C. glabrata sensu stricto with MALDI-TOF, and WGS confirmed this identification. The MICs of fluconazole, voriconazole, and amphotericin B were.256,.32, and.32 mg/ mL, respectively. A novel frameshift mutation in the ERG1 gene (Pro314fs) and many missense variations were detected in the ergosterol synthesis genes. None of the missense variations in the ML72254 ergosterol synthesis genes were deleterious, and the Pro314fs mutation was identified as the causative molecular change for a cholesterol-dependent and multidrug-resistant phenotype. This study verified that web-based computational biology solutions can be powerful tools for examining the possible impacts of missense mutations in C. glabrata.IMPORTANCE In this study, a cholesterol-dependent C. glabrata clinical isolate that confers azole and AmB resistance was investigated using artificial intelligence (AI) technologies and cloud computing applications. This is the first of the known cholesterol-dependent C. glabrata isolate to be found in Turkey. Cholesterol-dependent C. glabrata isolates are rarely isolated in clinical samples; they can easily be overlooked during routine laboratory procedures. Microbiologists therefore need to be alert when discrepancies occur between microscopic examination and growth on routine media. In addition, because these isolates confer antifungal resistance, patient management requires extra care.