A core filamentation response network in Candida albicans is restricted to eight genes

Abstract

Although morphological plasticity is a central virulence trait of Candida albicans, the number of filament-associated genes and the interplay of mechanisms regulating their expression remain unknown. By correlation-based network modeling of the transcriptional response to different defined external stimuli for morphogenesis we identified a set of eight genes with highly correlated expression patterns, forming a core filamentation response. This group of genes included ALS3, ECE1, HGT2, HWP1, IHD1 and RBT1 which are known or supposed to encode for cell- wall associated proteins as well as the Rac1 guanine nucleotide exchange factor encoding gene DCK1 and the unknown function open reading frame orf19.2457. The validity of network modeling was confirmed using a dataset of advanced complexity that describes the transcriptional response of C. albicans during epithelial invasion as well as comparing our results with other previously published transcriptome studies. Although the set of core filamentation response genes was quite small, several transcriptional regulators are involved in the control of their expression, depending on the environmental condition.

Figure 1. Dynamics of filamentous growth in Candida albicans.

Cells of C. albicans SC5314 strain were incubated under filament- inducing conditions in time course experiments for up to 12 h. For pH shift, cells were transferred from a M199 pH4 preculture to M199 pH8. For the other shifts, cells were transferred from a preculture in SDG medium into either SDG with 10% human serum (serum shift) or into SDN medium with 20 g/l N- acetylglucosamine as exclusive carbon source (GlcNAc shift). Note, that precultures were already grown at 37°C over night, so there was no additionally temperature shift effect to stimulate germ tube formation. For later time points, individual hyphae are shown instead of hyphal conglomerates but were identical in average length. Precipitates visible in serum induction regulary occur with human serum, baccterial contaminations were excluded. Scalebar: 20 µm.

Figure 3. Transcriptional landscape modeling of Candida albicans hyphae.

(A) Based on data from whole genome DNA microarrays, correlation- based networks of gene expression were modeled using the Cytoscape software visualizing the overall transcriptomic response to the three stimuli in the time-course of filamentation. For germ tube formation, only genes which were differentially expressed during 1 h and/or 2 h in all three shifts were used for modeling and only those which were linked directly to each other with high correlation (≥0,75) were integrated into the networks, leading to the identification of an upregulated set of early filamentation genes (early filamentation network EFN, B) and a group of genes downregulated during induction of filamentation (germ-tube formation network GFN, B). The same modeling was performed for hyphal elongation with genes which were differentially expressed at 3 h and 6 h in all three shifts, resulting in the description of a single network of upregulated genes (late filamentation network, LFN, B). (C) Integration of the time-point networks defined the core filamentation response consisting of a network of eight genes showing highly correlated expression patterns.

Figure 2. Shift- specific gene expression patterns in Candida albicans hyphae.

A summary of the data from the whole genome DNA microarrays used for this study. Genes showing fold changes of at least 1.5 were evaluated for significance (p≤0.05) and illustrated in blue for down- regulation and red for up- regulation. Genes were marked in yellow as not differentially expressed. (A). Differentially expressed genes for all three shifts at 1 h. (B–D) The expression dynamics of genes closely linked to the pH shift(B), the serum shift (C) or the change of the carbon source from glucose to N- acetylglucosamine (D) are shown. The presented data were taken from the whole genome DNA microarrays used for this study. The fold changes of at least 1.5 were evaluated for significance (p≤0.05) and illustrated in blue for down- regulation and red for up- regulation. Genes were marked in yellow as not differentially expressed.

Figure 4. Deletion of IHD1 and orf19.2457 did not affect filamentous growth.

Cells of the C. albicans wild type strain SC5314 and mutants ihd1Δ and orf19.2457Δ were transferred from an overnight preculture in M199 pH4 grown at 37°C to M199 pH8 and incubated for 2 h or 6 h prior to microscopy. For hyphal growth on solid media in either plates with 20 g/l N-acetylglucosamine as sole carbon source or chocolate agar, 1×10⁶ cells/ml from a preculture grown in SDG minimal medium at 37°C overnight, were dropped onto plates and incubated for 2 or 3 days prior to photography. Scalebar: 20 µm.

Figure 5. Expression pattern of the core filamentation response genes in selected transcriptome studies.

Transcriptional data from selected studies were analyzed for information about the eight core filamentation response network genes. As different technologies and normalization pattern were used, it is only possible to provide information about up (red)- or down- regulation (blue) or a no change of expression (yellow) of the indicated genes. The open reading frame orf19.2457 was partially not part of the microarray design in some studies and were marked absent (black). (A) Expression dynamics for the genes of the early and late filamentation networks as well as of the germ tube formation network during the invasion of human oral epithelial cells (TR146 cell line) with C. albicans wild type SC5314. The dataset were taken from the study of Wächtler and coworkers . (B) Expression dynamics for the core filamentation response genes in different transcriptome studies covering more hyphae- inducing conditions thant those which were used in this study.

Figure 6. Regulation of core filamentation response gene expression.

Mutants lacking the transcriptional regulators Efg1, Cph1, Rim101 and Nrg1 were grown for 3 h at 37°C (pH shift: from pH4 to pH8, serum shift: addition of 10% human serum, GlcNAc shift: N- acetylglucosamine as carbon source). Expression of the eight core filamentation response genes was analyzed by quantitative RT PCR in three independent experiments and gene expression was normalized against the housekeeping gene ACT1 (actin) and a common reference (wild type, 6 h in YPD at 37°C). To calculate the fold change of expression, relative gene expression of hyphae- inducing conditions (e.g. pH8) were compared to yeast promoting conditions (e.g. pH4). The illustrated fold changes were evaluated for significance (p≤0.05, student’s t test).