Characterization and Differentiation of Candida auris on Dixon's Agar Using Raman Spectroscopy

Abstract

Candida auris, an emerging multidrug-resistant fungal pathogen, poses significant challenges in healthcare settings due to its high misidentification rate and resilience to treatments. Despite advancements in diagnostic tools, a gap remains in rapid, cost-effective identification methods that can differentiate C. auris from other Candida species, particularly on non-standard culture media. We used Raman spectroscopy to characterize C. auris grown on modified Dixon's agar (mDixon) and differentiated it from Candida albicans and Candida parapsilosis. Key Raman spectral markers at 1171 cm^-1 and 1452 cm^-1, linked to mannan and β-glucan composition, differentiated C. auris into two subgroups, A and B. Despite the spectral similarities of groups A and B with C. albicans and C. parapsilosis, respectively, all Candida species were distinguishable through principal component analysis (PCA). Additionally, this study is the first to demonstrate the distinct spectral signature of mDixon agar, achieved through spatially offset Raman spectroscopy (SORS), which enables accurate discrimination between the culture medium and fungal samples. The observed inter-individual variability within C. auris, coupled with the spectral overlap between C. auris subgroups and other Candida species, highlights a major challenge in differentiating closely related fungi due to their similar molecular composition. Enhancements in spectral resolution and further fluorescence minimization from the culture medium are needed to reliably detect the subtle biochemical differences within these species. Despite these challenges, the results underscore the potential of Raman spectroscopy as a real-time, non-destructive, and complementary tool for fungal pathogen identification.

Keywords: Raman spectroscopy; SORS; candida; candida auris; modified Dixon’s agar; spatially offset Raman spectroscopy.

Figure 1

Average Raman spectra of the three Candida species and the culture medium (mDixon). The shaded regions in the spectrum highlight the three Raman bands that are key to distinguishing the characteristic differences between the species. Dotted lines mark other vibrational modes associated with the molecular components of the cell walls of the Candida species.

Figure 2

(a): PCA scores plot of the first two principal components (PC1 and PC2). The shaded regions represent the 95% confidence ellipses of the respective classes; (b,c): Additional PCA plots focusing on pairs of closely related clusters: (b) C. auris B vs. C. parapsilosis, and (c) C. albicans vs. C. auris A. In both cases, a linear discriminant analysis (LDA) decision boundary was calculated using the scores from the first two principal components. The decision boundary highlights the separation achieved by LDA between the groups based on their PC scores, further supporting the distinction between the Candida species.

Figure 3

PC1 loadings plot representing 78.72% of the variance in the Raman spectral data. The loadings plot indicates the contribution of each Raman shift to PC1, with positive and negative loadings highlighting the key spectral regions responsible for the separation of the samples in the PCA analysis.

Figure 4

Difference spectra of C. albicans vs. C. auris A and of C. auris B vs. C. parapsilosis.