Pathway analysis of Candida albicans survival and virulence determinants in a murine infection model

Abstract

One potentially rich source of possible targets for antifungal therapy are those Candida albicans genes deemed essential for growth under the standard culture (i.e., in vitro) conditions; however, these genes are largely unexplored as drug targets because essential genes are not experimentally amenable to conventional gene deletion and virulence studies. Using tetracycline-regulatable promoter-based conditional mutants, we investigated a murine model of candidiasis in which repressing essential genes in the host was achieved. By adding doxycycline to the drinking water starting 3 days prior to (dox - 3D) or 2 days post (dox + 2D) infection, the phenotypic consequences of temporal gene inactivation were assessed by monitoring animal survival and fungal burden in prophylaxis and acute infection settings. Of 177 selected conditional shut-off strains tested, the virulence of 102 was blocked under both repressing conditions, suggesting that the corresponding genes are essential for growth and survival in a murine host across early and established infection periods. Among these genes were those previously identified as antifungal drug targets (i.e., FKS1, ERG1, and ERG11), verifying that this methodology can be used to validate potential new targets. We also identify genes either conditionally essential or dispensable for in vitro growth but required for survival and virulence, including those in late stage ergosterol synthesis, or early steps in fatty acid or riboflavin biosynthesis. This study evaluates the role of essential genes with respect to pathogen virulence in a large-scale, systems biology context, and provides a general method for gene target validation and for uncovering unexpected antimicrobial targets.

Fig. 1.

Identification of the essential role of ERG11 and FKS1 in a murine model of candidiasis using conditional shut-off strains. (A) The Tet-off system used to construct conditional shut-off strains. The active Tet-transactivator is indicated by a filled diamond, inactive open symbol, and tetracycline red dot. (B) The terminal phenotypes of indicated strains under the repressing conditions (100 μg/mL tetracycline) on media indicated. S = 20% mouse serum. (C) Standard survival curves of animals infected with the pTET-ERG11 strains under different conditions, sugar, nonrepressing; dox - 3D, repressing conditions established 3 days prior to infection; dox + 2D, established 2 days p.i.; dox - 3D/W, nonrepressing conditions from days 22–35 for surviving animals from the dox - 3D group; dox + 2D/W, for animals from the dox + 2D group. Arrows indicate necropsy and fungal burden determination on days 21 and 35 p.i. (D) Summary of independent animal experiments with the conditional shut-off strains of ERG11 and FKS1, with each experiment consisting of three conditions indicated on the left. The survival rates are represented by red color with different shades, with scale shown on the top. The survival rates on day 21 are on the right. Green bars indicate the day on which ≥60% of the animals expired due to lethal infection. (E) Effects of doxycycline on the in vivo virulence of the parental strain used in this study. Animals were fed with sugar or doxycycline immediately after infection and thereafter. (F) Kidney fungal burdens (days 21 and 35 p.i.) of animals infected with the pTET-ERG11 and pTET-FKS1 strains under the two repressing conditions, dox - 3D and dox + 2D. Dashed horizontal line indicates limit of detection in enumerating fungal burden.

Fig. 2.

Virulence phenotypes of 177 C. albicans genes. A. Graphic summary of animal survival rates. Genes are divided into three classes, essential (S_[21,dox-3D] and S_[21,dox+2D]≥80%); with virulence attenuated (40% ≤ S_[21,dox-3D] and S_[21,dox+2D] < 80%), and nonessential (S_[21,dox-3D] and S_[21,dox+2D] ≤ 20%). For each class, the survival rates under the dox - 3D conditions are presented proportionally (indicated by numbers of strains) in the inner donut with the corresponding dox + 2D survival presented in the outer donut. (B) Distribution of kidney fungal burdens (CFU/g, on day 21 p.i.) of animals infected with strains of essential genes (red) and those with attenuated virulence (green) under the repressing conditions of dox - 3D (open) and dox + 2D (filled). Data used in the analysis are indicated in Table S3. (C) Concordance of in vivo virulence and in vitro growth phenotype. Genes are divided into four classes, and the corresponding in vitro phenotypes in each class are presented proportionally (indicated by number of strains) by E (essential), NE (nonessential), and CE (conditional-essential, defined as essential on YNBD but not on YPD). A summary of all animal survival curves and p-values for C. albicans strains corresponding to genes with attenuated virulence is provided (Fig. S1).

Fig. 3.

Summary of survivals of animals infected with the C. albicans conditional shut-off strains of genes involved in the ergosterol pathway. Genes are listed according to their roles in the pathway (Left). The growth phenotypes of the corresponding strains on YPD (1% ethanol, the nonrepressing conditions) and YPD + TET (100 μg/mL tetracycline and 1% ethanol, repressing) at 30 °C (2 days) are shown in the middle. Survival data are summarized on the right. Each strain was used to infect animals under three conditions (5 animals each): S, fed with sugar in the drinking water (nonrepressing); -3D and +2D, fed with doxycycline in the drinking water (repressing) 3 days prior to and 2 days post infection, respectively. Animal survivals (red with different shades, with the scale shown on top) were monitored for 3 weeks (separated by white bars). The statistic significance of survival under the each repressing condition was compared with that in the corresponding sugar group, ***, p < 0.001; and **, p < 0.01. Those under two repressing conditions were also compared, with ††† indicating p < 0.001; †† p < 0.01; and † p < 0.05.

Fig. 4.

Summary of the 177 C. albicans genes characterized in this study in the context of their protein sequence similarity to A. fumigatus and human homologs. The similarity is expressed by - lg(BLAST E-value); i.e., the higher this value, the higher the similarity, with 200 including those that are great than 200. The absence of homolog is indicated by 0. The in vivo virulence phenotype is represented by different colors; with known targets of antifungal drugs as indicated.