Loss of a conserved C-terminal region of the Aspergillus fumigatus AtrR transcriptional regulator leads to a gene-specific defect in target gene expression

Abstract

Treatment of fungal infections associated with the filamentous fungus Aspergillus fumigatus is becoming more problematic as this organism is developing resistance to the main chemotherapeutic drug at an increasing rate. Azole drugs represent the current standard-of-care in treatment of aspergillosis with this drug class acting by inhibiting a key step in biosynthesis of the fungal sterol ergosterol. Azole compounds block the activity of the lanosterol α-14 demethylase, encoded by the cyp51A gene. A common route of azole resistance involves an increase in transcription of cyp51A. This transcriptional increase requires the function of a Zn2Cys6 DNA-binding domain-containing transcription activator protein called AtrR. AtrR was identified through its action as a positive regulator of expression of an ATP-binding cassette transporter (abcC/cdr1B here called abcG1). Using both deletion and alanine scanning mutagenesis, we demonstrate that a conserved C-terminal domain in A. fumigatus is required for expression of abcG1 but dispensable for cyp51A transcription. This domain is also found in several other fungal pathogen AtrR homologues consistent with a conserved gene-selective function of this protein segment being conserved. Using RNA-seq, we find that this gene-specific transcriptional defect extends to several other membrane transporter-encoding genes including a second ABC transporter locus. Our data reveal that AtrR uses at least two distinct mechanisms to induce gene expression and that normal susceptibility to azole drugs cannot be provided by maintenance of wild-type expression of the ergosterol biosynthetic pathway when ABC transporter expression is reduced.

Figure 1.. Identification and functional characterization of a conserved C-terminal domain in fungal AtrR proteins.

A. Alignments of AtrR proteins from selected fungal species are shown. AtrR homologues from the indicated fungi along with several abbreviated as follows: A. fumigatus (Afu), Paracoccidioides brasiliensis (Pbrasiliensis), Histoplasma capsulatum (Hcapsulatum), Aspergillus flavus (Aflavus), Aspergillus niger (Aniger), Penicillium marneffei (Pm), Aspergillus nidulans (An), Neurospora crassa (Nc). A plot generated by Multiple Sequence Alignment at NCBI of the conserved region in the AtrR C-termini is shown with the most conserved resides indicated in Rasmol notation. B. Loss of the conserved 855–879 region of A. fumigatus AtrR increased voriconazole susceptibility. Isogenic strains containing the indicated forms of AtrR were tested for radial growth in the absence (Minimal Medium) or the presence of different voriconazole concentrations 0.1 or 0.2 μg/ml) C. Loss of the 855–879 region of AtrR does not lead to a large reduction in the steady-state level of the resulting mutant protein. The strains containing the different atrR alleles were grown in the presence or absence of voriconazole and whole cell protein extracts prepared. Equal amounts of protein were analyzed by western blotting using an anti-AtrR antiserum (Top panel) and equal loading confirmed by Ponceau S staining of the membrane after protein transfer.

Figure 2.. Selective transcriptional defect in the Δ855–879 AtrR protein.

Isogenic strains varying at their atrR locus were grown 24 hours in the absence (−) or for 16 hours in the absence and then 8 hours in the presence (+) of voriconazole and total RNA prepared. Steady-state levels of mRNA for abcG1 (top panel) or cyp51A (bottom panel) were measured using appropriate primers. Values presented are normalized to the level of expression determined in wild-type cells grown in the absence of voriconazole.

Figure 3.. Transactivation domain of AtrR extends to its extreme C-terminus.

A. Isogenic strains expressing wild-type (wt) or the indicated forms of AtrR were tested for growth on minimal medium lacking or containing voriconazole at the concentrations listed using the radial growth assay as described above. B. The strains in A were grown for 24 hours in minimal medium and whole cell protein extracts prepared. Equal amounts of protein from each strain were analyzed by western blotting using the anti-AtrR antiserum and equivalent loading confirmed by staining with Ponceau S. The lanes below each panel are identical and correspond to wild-type (Lane 1), Δ855–879 (Lane 2), 1–654 (Lane 3), 1–754 (Lane 4) and 1–854 (Lane 5). M indicates a lane containing molecular mass standards. The arrows denote the major species of AtrR detected. C. Expression levels of abcG1 and cyp51A were determined for each indicated form of AtrR (listed at the bottom for both panels) using RT-qPCR as above.

Figure 4.. Internal deletions in AtrR identify a region important in expression of the factor.

A. The indicated strains were tested in a radial growth assay on either minimal medium alone or containing 0.05 μg/ml voriconazole. B. Expression of abcG1 and cyp51A was assessed by RT-qPCR as above in the presence and absence of voriconazole. C. The strains above along with a control strain expressing the 1–654 C-terminal truncation mutant of AtrR were analyzed by western blotting as above. The estimated location of each AtrR polypeptide chain is indicated by the arrow.

Figure 5.. Alanine scanning mutagenesis of the conserved 855–879 region of AtrR.

Adjacent three amino acid positions in the conserved 855–879 region of AtrR were replaced with three alanine residues. In each case the 3 amino acids changed are shown along with a number referring to the first position to be changed to alanine. A. The indicated strains were evaluated by the radial growth assay for their ability to grow on minimal medium with or without the indicated concentration of voriconazole. These assays were repeated at least 3 times and diameter of each colony measured. These values are shown in the plot on the right-hand side of the panel. B. Expression of abcG1 and cyp51A was measured by RT-qPCR in the absence and presence of voriconazole treatment.

Figure 6.. RNA-seq analysis of Δ855–879 AtrR.

Venn diagrams of the numbers of overlapping genes from different analyses are shown. All reductions in gene expression are limited to those that are changed significantly (padj<0.05) and a log2≥1. A. Comparison of genes reduced in expression in the atrRΔ strain in the absence (NulldownYG) or presence (Nulldownvori) of voriconazole to genes with reduced expression in the Δ855–879 AtrR strain with voriconazole (Mutdownvori). These expression profiles are then assessed for their genes in common with those bound by AtrR via ChIP-seq (ChIP). B. Genes that are reduced in expression in the strain containing the Δ855–879 AtrR protein in the absence (MutdownYG) or presence (Mutdownvori) of voriconazole and that contain AtrR binding sites as measured by ChIP-seq (ChIP).

Figure 7.. Functional domains in the AtrR transcription factor.

The numbers refer to positions along the protein chain from A. fumigatus. The abbreviations for the identified protein regions are DNA-binding domain (DBD) and central regulatory domain (CRD). The minimal transcriptional activation domains required for normal expression of either cyp51A or abcG1 are shown as gray bars.