The CCAAT-Binding Complex Controls Respiratory Gene Expression and Iron Homeostasis in Candida Glabrata

Abstract

The CCAAT-binding complex (CBC) is a heterotrimeric transcription factor which is widely conserved in eukaryotes. In the model yeast S. cerevisiae, CBC positively controls the expression of respiratory pathway genes. This role involves interactions with the regulatory subunit Hap4. In many pathogenic fungi, CBC interacts with the HapX regulatory subunit to control iron homeostasis. HapX is a bZIP protein which only shares with Hap4 the Hap4Like domain (Hap4L) required for its interaction with CBC. Here, we show that CBC has a dual role in the pathogenic yeast C. glabrata. It is required, along with Hap4, for the constitutive expression of respiratory genes and it is also essential for the iron stress response, which is mediated by the Yap5 bZIP transcription factor. Interestingly, Yap5 contains a vestigial Hap4L domain. The mutagenesis of this domain severely reduced Yap5 binding to its targets and compromised its interaction with Hap5. Hence, Yap5, like HapX in other species, acts as a CBC regulatory subunit in the regulation of iron stress response. This work reveals new aspects of iron homeostasis in C. glabrata and of the evolution of the role of CBC and Hap4L-bZIP proteins in this process.

Figure 1

The Hap5 network in Candida glabrata. An arrow indicates a potential regulatory interaction based on ChIP-seq. The color of the targets indicates their belonging to respiratory pathways (yellow) or not (white). The most enriched DNA motif in ChIP peaks is represented at the bottom right. The Yap5 data are from ref. . The gene names indicated are those of the S. cerevisiae orthologs (according to the CGD database, www.candidagenome.org). For the sake of clarity, only the names of the genes which are discussed in the main text are indicated.

Figure 2

Transcriptome analyses of Hap5 impact on gene expression. The wild type and hap5Δ strains were grown in three different conditions (rich media, iron excess or iron starvation) and their transcriptomes were compared using microarrays. (A) Venn diagram representing the overlaps between the lists of Hap5 ChIP targets being significantly down regulated compared with wild type in YPD (blue line), BPS (green line) or iron excess (black line). The red line includes the only Hap5 ChIP target (GRX4) which was significantly up-regulated upon BPS treatment. The gene names are from the S. cerevisiae orthologs, when available. (B) Eisengram of the expression profiles of the genes from the Venn diagram. The values used are log2 of hap5Δ/wild type expression ratios. The color scale is indicated. The conditions used are 1: YPD; 2: iron excess (2 mM FeSO4 for 30 minutes); 3: iron starvation (0.5 mM BPS for 30 minutes).

Figure 3

Analyses of the impact of Hap5, Hap4 and Yap5 on ATP2 and GRX4. The relative expression of ATP2 (A) and GRX4 (B) was measured by Q-RTPCR in wild type, hap5Δ, hap4Δ and yap5Δ strains grown in glucose, glycerol or iron excess. The values represent the expression levels of the ATP2 or GRX4 genes relative to ACT1 (used as an internal control) and to the wild type grown in glucose. The experiments were performed three times on biologically independent samples. Error bars represent the pearson standard deviation. A t-test was performed to compare, for each growth condition, the mutants to the corresponding wild type. The results of the test are indicated by the stars as follows *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 4

Molecular basis of the Yap5-Hap5 interaction. (A) Multiple alignments of the Hap4L domains of Hap4 (from S. cerevisiae), wild type C. glabrata Yap5, Yap5-Hap4LΔ and Yap5-mut2. For the latter, the substitutions are highlighted in red. (B) ChIP-QPCR was performed on strains expressing a myc-tagged Yap5 in presence (wild type) or absence (hap5Δ) of HAP5 and on strains expressing the two different Yap5 mutant versions. All strains were grown in YPD. The values represent the IP/Input ratios of the GRX4 promoter relative to the enrichment of the YHB1 promoter (used as an internal control), expressed as a percentage of the enrichment obtained for the wild type Yap5. The experiments were performed twice on biologically independent samples. Error bars hence represent the standard error of the mean. (C) Western blot analyses of the co-immunoprecipitation experiments using Hap5-Protein A as bait and wild type or mutated versions of Yap5-myc as prey. Upper panel: input samples (INPUT), lower panel: immunoprecipitated samples (IP). Immunoblotting was performed with a mouse anti-myc antibody (Roche). The Yap5 protein is fused to 13 c-Myc epitopes and the corresponding band is expected at 65 kDa. The Hap5-Protein A fusion is expected at 45 kDa and is also detected by the anti-myc antibody (although with a low affinity), because Protein A non-specifically interacts with IgG. Note the similar intensity of the Hap5-ProtA bands in the IP, which indicates that the IP efficiency was equivalent from one lane to another. The star indicates the 50 kDa band corresponding to the large chain of the anti-Mouse antibodies used for the IP. The co-immunoprecipitation experiment was performed twice on biologically independent samples and gave consistent results. The ladder on the right was copied and pasted from the white light image of the membrane. Immunoblotting of the same membranes with rabbit IgG-HRP polyclonal antibody (PAP; code Z0113; Dako), which has a high affinity for Protein A, can be found in Supplementary File S7.

Figure 5

A dual role for CBC in Candida glabrata. CBC plays a dual role in the control of cellular respiration (together with the regulatory subunit Hap4) and of the iron stress response mediated by the Yap5 transcription factor.

Figure 6

The evolution of the roles of CBC and its regulatory subunits in the control of fungal iron homeostasis. In Schizosaccharomyces pombe, Php4 plays an important role in the iron starvation response by repressing the iron consuming genes through its interaction with the CCAAT Binding Complex (CBC) which is mediated by its Hap4Like domain (Hap4L) (reviewed in ref. 15). In C. glabrata, Yap5 is a major regulator of the iron stress response which activates iron consuming genes^, . Yap5 binding to its targets requires CBC, probably by direct interaction with Hap5 through its vestigial Hap4L domain (this work). However, Yap5 also interacts directly with DNA through its bZIP domain and this interaction is essential for its regulatory activity^{, ,} . Interestingly enough, the situation in filamentous ascomycetes (e.g. Aspergillus or Fusarium species) and in C. albicans is an intermediate between S. pombe and S. cerevisiae. HapX plays an important dual role in activating the iron stress response and in repressing the same genes in iron starvation^{, , ,} . HapX interacts with CBC through its conserved Hap4L domain, but it also directly contributes to the binding of the CBC-HapX complex to a bipartite DNA motif, probably through its bZIP sequence^{, , ,} .