Metabolic specialization associated with phenotypic switching in Candidaalbicans

Abstract

Phase and antigenic variation are mechanisms used by microbial pathogens to stochastically change their cell surface composition. A related property, referred to as phenotypic switching, has been described for some pathogenic fungi. This phenomenon is best studied in Candida albicans, where switch phenotypes vary in morphology, physiology, and pathogenicity in experimental models. In this study, we report an application of a custom Affymetrix GeneChip representative of the entire C. albicans genome and assay the global expression profiles of white and opaque switch phenotypes of the WO-1 strain. Of 13,025 probe sets examined, 373 ORFs demonstrated a greater than twofold difference in expression level between switch phenotypes. Among these, 221 were expressed at a level higher in opaque cells than in white cells; conversely, 152 were more highly expressed in white cells. Affected genes represent functions as diverse as metabolism, adhesion, cell surface composition, stress response, signaling, mating type, and virulence. Approximately one-third of the differences between cell types are related to metabolic pathways, opaque cells expressing a transcriptional profile consistent with oxidative metabolism and white cells expressing a fermentative one. This bias was obtained regardless of carbon source, suggesting a connection between phenotypic switching and metabolic flexibility, where metabolic specialization of switch phenotypes enhances selection in relation to the nutrients available at different anatomical sites. These results extend our understanding of strategies used in microbial phase variation and pathogenesis and further characterize the unanticipated diversity of genes expressed in phenotypic switching.

Fig 1.

Reproducibility of custom Candida microarrays. (A) Correlation between two independent hybridizations using the same cRNA sample. (B) Correlation between two independent cRNA samples prepared from duplicate experiments. Only probe sets of large ORFs with average differences ≥50 are shown.

Fig 2.

Expression of some known W and O specific genes. (A) Relative expression differences comparing O to W cells. The relative expression differences are represented by fold-change. Positive values represent higher expression in O cells as compared with W cells. Negative values represent higher expression in W cells as compared with O cells. (B) Northern blot of WH11 and OP4 gene expression in the W and O cell.

Fig 3.

Diagram of metabolic pathways and relative expression of specific genes by fold changes comparing O to W cells. The comparison for O(12h)/W(12h), O(18h)/W(18h), O(24h)/W(24h) and O(48h)/W(48h) are represented by a, b, c, and d, respectively. Genes: 1, HXT3; 2, HXT4; 3, ORF6.187; 4, ORF6.2377; 5, HXK1; 6, PFK2; 7, CDC19; 8, IDP2; 9, MDH1; 10, MLS1; 11, FAA2; 12, POX1; 13, ECI1; 14, FOX2; 15, FOX3; 16, ORF6.6337; 17, ADH3; 18, ORF6.4277; 19, ALD6; 20, GDB1; 21, TPS1; 22, TPS2; 23, GLG2.

Fig 4.

Functional classification of genes with higher expression in W or O cells. The total 152 W-up-regulated ORFs (A) and 221 O-up-regulated ORFs (B) are grouped by functional category. These categories are based on the S. cerevisiae Comprehensive Yeast Genome Database, described at http://mips.gsf.de/proj/yeast/CYGD/db/index.html.