Candida albicans infection of Caenorhabditis elegans induces antifungal immune defenses

Abstract

Candida albicans yeast cells are found in the intestine of most humans, yet this opportunist can invade host tissues and cause life-threatening infections in susceptible individuals. To better understand the host factors that underlie susceptibility to candidiasis, we developed a new model to study antifungal innate immunity. We demonstrate that the yeast form of C. albicans establishes an intestinal infection in Caenorhabditis elegans, whereas heat-killed yeast are avirulent. Genome-wide, transcription-profiling analysis of C. elegans infected with C. albicans yeast showed that exposure to C. albicans stimulated a rapid host response involving 313 genes (124 upregulated and 189 downregulated, ~1.6% of the genome) many of which encode antimicrobial, secreted or detoxification proteins. Interestingly, the host genes affected by C. albicans exposure overlapped only to a small extent with the distinct transcriptional responses to the pathogenic bacteria Pseudomonas aeruginosa or Staphylococcus aureus, indicating that there is a high degree of immune specificity toward different bacterial species and C. albicans. Furthermore, genes induced by P. aeruginosa and S. aureus were strongly over-represented among the genes downregulated during C. albicans infection, suggesting that in response to fungal pathogens, nematodes selectively repress the transcription of antibacterial immune effectors. A similar phenomenon is well known in the plant immune response, but has not been described previously in metazoans. Finally, 56% of the genes induced by live C. albicans were also upregulated by heat-killed yeast. These data suggest that a large part of the transcriptional response to C. albicans is mediated through "pattern recognition," an ancient immune surveillance mechanism able to detect conserved microbial molecules (so-called pathogen-associated molecular patterns or PAMPs). This study provides new information on the evolution and regulation of the innate immune response to divergent pathogens and demonstrates that nematodes selectively mount specific antifungal defenses at the expense of antibacterial responses.

Figure 1. C. albicans yeast can kill C. elegans.

(A) Live C. albicans (closed diamonds) were pathogenic to nematodes on solid media, whereas heat-killed C. albicans (open circles) and E. coli (crosses) were not (P<0.001). The graph presents the average of three plates per strain, each with 30 to 40 animals per plate. Data are representative of two biological replicates. (B) Images of C. elegans animals exposed to heat-killed E. coli (HK E.c.), heat-killed C. albicans (HK C.a.) or live C. albicans (live C.a.) for 16 hours at 25°C are shown. Images of the proximal (left) and distal (right) intestine were obtained using Nomarski optics. Both live and heat-killed C. albicans accumulated within the intestine, but only live C. albicans caused marked distention of the proximal intestine. Arrows point to the pharyngeal grinder and arrowheads outline the lumen of the intestine. The scale bar represents 20 µm.

Figure 2. A C. albicans double mutant strain that is attenuated for pathogenicity in mammals is also unable to efficiently kill C. elegans.

The C. albicans efg1Δ/efg1Δ cph1Δ/cph1Δ double mutant strain (efg1/cph1) exhibited a reduced ability to kill C. elegans compared to its isogenic wild-type parent strain SC5314 (P<0.001). The graph presents the average of three plates per strain, each with 30 to 40 animals per plate. Data are representative of two biological replicates.

Figure 3. Infection with C. albicans yeast induces a rapid host response.

(A) C. elegans genes that were differentially regulated in C. albicans-exposed versus heat-killed E. coli-exposed young adult animals at 4 hours after infection are depicted on a genome-wide intensity plot of 22,548 sequences. Genes colored red were upregulated by C. albicans (P<0.01), those colored green were downregulated (P<0.01) and those colored blue were unchanged. Diagonal lines represent 2-fold change and the numbers of genes differentially regulated greater than 2-fold are indicated (P<0.01)(124 genes were upregulated and 189 genes were downregulated). (B) qRT-PCR was used to confirm the results of the microarray analysis. 11 genes with varying degrees of differential regulation were selected and studied under each condition in which they were differentially regulated in the microarray analysis (see Table S2 for gene identities). Correlation of microarray and qRT-PCR data was determined by plotting the average fold difference observed in the microarray analysis (three biological replicates) versus the average fold difference for the same gene obtained by qRT-PCR (three biological replicates). Linear regression analysis revealed strong correlation between the datasets (R² of 0.90).

Figure 4. The virulence of the infecting C. albicans strain affects the induction of putative antifungal immune effectors.

The induction of abf-2, fipr-22/23, cnc-4 and cnc-7 is reduced in wild-type C. elegans animals during infection with the virulence-attenuated C. albicans efg1Δ/efg1Δ cph1Δ/cph1Δ double mutant strain [vs. heat-killed (HK) E. coli] compared to its isogenic wild-type parent strain SC5314 (vs. heat-killed E. coli). Data are presented as the average of three biological replicates, each conducted in duplicate and normalized to a control gene with error bars representing SEM. *P = 0.06, **P<0.01 and ***P<0.025 for the comparison of gene induction on SC5314 versus efg1Δ/efg1Δ cph1Δ/cph1Δ.

Figure 5. The p38 MAP Kinase PMK-1 is Required for the response to C. albicans infection.

(A) A C. albicans infection assay with wild-type (N2) and pmk-1(km25) animals shows that pmk-1(km25) mutants were more susceptible to C. albicans infection (P<0.01). Each time point represents the average of three plates per strain, each with 30 to 40 animals per plate. Data are representative of two independent experiments. (B) N2 and pmk-1(km25) young adult animals were exposed to the indicated food source and the indicated genes were studied using qRT-PCR (HK equals heat-killed). Expression is relative to N2 on heat-killed E. coli and the data are presented as the average of three biological replicates each normalized to a control gene with error bars representing SEM. *P<0.001 and **P equals 0.05 for the comparison of relative expression of the indicated gene in wild-type animals on C. albicans versus pmk-1(km25) animals on C. albicans.

Figure 6. The transcriptional responses to C. albicans and bacteria comprise specific and overlapping gene sets.

A Venn diagram illustrates the overlap of genes induced 2-fold or greater (P<0.01) by C. albicans (this study), P. aeruginosa and S. aureus . All microarrays were conducted using the Affymetrix platform. Animals were exposed to C. albicans and P. aeruginosa for 4 hours and to S. aureus for 8 hours. See Table S3A for gene identities.

Figure 7. Heat-killed C. albicans yeast cells elicit a transcriptional response in C. elegans that overlaps with the response to live C. albicans.

Venn diagrams give the overlap of C. elegans genes upregulated (A) and downregulated (B) at least 2-fold (P<0.01) in response to C. albicans and heat-killed C. albicans, each compared to heat-killed E. coli. See Table S3B for gene identities.

Figure 8. The C. elegans response to C. albicans involves the downregulation of antibacterial effectors.

(A) A Venn diagram illustrates that a subset of C. albicans downregulated genes were upregulated after infection of C. elegans by pathogenic bacteria. See Table S3C for gene identities. (B) Transgenic C. elegans animals in which GFP expression was driven by the promoter for the C-type lectin clec-60, a secreted S. aureus immune effector that was downregulated by C. albicans in the microarray analysis, are shown. Worms were exposed to heat-killed (HK) E. coli, heat-killed C. albicans or live C. albicans for 20 hours at 25°C and then imaged. Green is clec-60::GFP. Red is the myo-2::mCherry co-injection marker used to identify transgenic animals.