Phagocytic cells form the first line of defense against infections by the human fungal pathogen Candida albicans. Recent in vitro gene expression data suggest that upon phagocytosis by macrophages, C. albicans reprograms its metabolism to convert fatty acids into glucose by inducing the enzymes of the glyoxylate cycle and fatty acid ␤-oxidation pathway. Here, we asked whether fatty acid ␤-oxidation, a metabolic pathway localized to peroxisomes, is essential for fungal virulence by constructing two C. albicans double deletion strains: a pex5⌬/pex5⌬ mutant, which is disturbed in the import of most peroxisomal enzymes, and a fox2⌬/ fox2⌬ mutant, which lacks the second enzyme of the ␤-oxidation pathway. Both mutant strains had strongly reduced ␤-oxidation activity and, accordingly, were unable to grow on media with fatty acids as a sole carbon source. Surprisingly, only the fox2⌬/fox2⌬ mutant, and not the pex5⌬/pex5⌬ mutant, displayed strong growth defects on nonfermentable carbon sources other than fatty acids (e.g., acetate, ethanol, or lactate) and showed attenuated virulence in a mouse model for systemic candidiasis. The degree of virulence attenuation of the fox2⌬/fox2⌬ mutant was comparable to that of the icl1⌬/icl1⌬ mutant, which lacks a functional glyoxylate cycle and also fails to grow on nonfermentable carbon sources. Together, our data suggest that peroxisomal fatty acid ␤-oxidation is not essential for virulence of C. albicans, implying that the attenuated virulence of the fox2⌬/fox2⌬ mutant is largely due to a dysfunctional glyoxylate cycle.

The glyoxylate cycle, a metabolic pathway required for generating C 4 units from C 2 compounds, is an important factor in virulence, in both animal and plant pathogens. Here, we report the localization of the key enzymes of this cycle, isocitrate lyase (Icl1; EC 4.1.3.1) and malate synthase (Mls1; EC 2.3.3.9), in the human fungal pathogen Candida albicans. Immunocytochemistry in combination with subcellular fractionation showed that both Icl1 and Mls1 are localized to peroxisomes, independent of the carbon source used. Although Icl1 and Mls1 lack a consensus type I peroxisomal targeting signal (PTS1), their import into peroxisomes was dependent on the PTS1 receptor Pex5p, suggesting the presence of non-canonical targeting signals in both proteins. Peroxisomal compartmentalization of the glyoxylate cycle is not essential for proper functioning of this metabolic pathway because a pex5D/D strain, in which Icl1 and Mls1 were localized to the cytosol, grew equally as well as the wild-type strain on acetate and ethanol. Previously, we reported that a fox2D/D strain that is completely deficient in fatty acid boxidation, but has no peroxisomal protein import defect, displayed strongly reduced growth on non-fermentable carbon sources such as acetate and ethanol. Here, we show that growth of the fox2D/D strain on these carbon compounds can be restored when Icl1 and Mls1 are relocated to the cytosol by deleting the PEX5 gene. We hypothesize that the fox2D/D strain is disturbed in the transport of glyoxylate cycle products and/or acetyl-CoA across the peroxisomal membrane and discuss the possible relationship between such a transport defect and the presence of giant peroxisomes in the fox2D/D mutant.

In eukaryotes, acetyl coenzyme A (acetyl-CoA) produced during peroxisomal fatty acid ␤-oxidation needs to be transported to mitochondria for further metabolism. Two parallel pathways for acetyl-CoA transport have been identified in Saccharomyces cerevisiae; one is dependent on peroxisomal citrate synthase (Cit), while the other requires peroxisomal and mitochondrial carnitine acetyltransferase (Cat) activities. Here we show that the human fungal pathogen Candida albicans lacks peroxisomal Cit, relying exclusively on Cat activity for transport of acetyl units. Deletion of the CAT2 gene encoding the major Cat enzyme in C. albicans resulted in a strain that had lost both peroxisomal and mitochondrion-associated Cat activities, could not grow on fatty acids or C 2 carbon sources (acetate or ethanol), accumulated intracellular acetyl-CoA, and showed greatly reduced fatty acid ␤-oxidation activity. The cat2 null mutant was, however, not attenuated in virulence in a mouse model of systemic candidiasis. These observations support our previous results showing that peroxisomal fatty acid ␤-oxidation activity is not essential for C. albicans virulence. Biofilm formation by the cat2 mutant on glucose was slightly reduced compared to that by the wild type, although both strains grew at the same rate on this carbon source. Our data show that C. albicans has diverged considerably from S. cerevisiae with respect to the mechanism of intracellular acetyl-CoA transport and imply that carnitine dependence may be an important trait of this human fungal pathogen.