Transcription factor FfmA interacts both physically and genetically with AtrR to properly regulate gene expression in the fungus Aspergillus fumigatus

Abstract

Transcriptional regulation of azole resistance in the filamentous fungus Aspergillus fumigatus is a key step in development of this problematic clinical phenotype. We and others have previously described a C2H2-containing transcription factor called FfmA that is required for normal levels of voriconazole susceptibility and expression of an ATP-binding cassette transporter gene called abcG1 . Null alleles of ffmA exhibit a strongly compromised growth rate even in the absence of any external stress. Here we employ an acutely repressible doxycycline-off form of ffmA to rapidly deplete FfmA protein from the cell. Using this approach, we carried out RNA-seq analyses to probe the transcriptome of A. fumigatus cells that have been deprived of normal FfmA levels. We found that 2000 genes were differentially expressed upon depletion of FfmA, consistent with the wide-ranging effect of this factor on gene regulation. Chromatin immunoprecipitation coupled with high throughput DNA sequencing analysis (ChIP-seq) identified 530 genes that were bound by FfmA using two different antibodies for immunoprecipitation. More than 300 of these genes were also bound by AtrR demonstrating the striking regulatory overlap with FfmA. However, while AtrR is clearly an upstream activation protein with clear sequence specificity, our data suggest that FfmA is a chromatin-associated factor that may bind to DNA in a manner dependent on other factors. We provide evidence that AtrR and FfmA interact in the cell and can influence one another's expression. This interaction of AtrR and FfmA is required for normal azole resistance in A. fumigatus .

Figure 1.. Structure of FfmA.

A. A cartoon depiction of the structure of FfmA is shown with the two C2H2 domains indicated as boxes 1 and 2. These two domains are located between amino acid residues 244 and 290. B. Sequence alignment of FfmA with S. cerevisiae Mot3. Identical residues are boxed in black, conservative substitutions are boxed in gray and gaps are indicated by the dashes.

Figure 2.. Doxycycline-depletion of doxycycline-regulated ffmA transcription.

A. A doxycycline-repressible (doxycycline off: DO) promoter-driven ffmA gene (DO-ffmA)-containing strain was grown overnight in the absence of doxycycline. A sample of this culture was processed to measure FfmA expression in the absence of doxycycline (0 hours) and then doxycycline was added for the numbers of hours indicated. Cultures were processed at each indicated time point and whole cell protein extracts prepared to determine the effect of doxycycline addition on expression of FfmA protein by western blotting with anti-FfmA. Tubulin was also analyzed as a loading control along with Ponceau S staining of the membranes after protein transfer. The numbers refer to relative expression normalized to the 0 hour time point. B. Samples from the strains grown as in A were processed for RNA and levels of either ffmA or act1 mRNA were determined by qPCR.

Figure 3.. RNA-seq analysis of doxycycline-regulated ffmA gene expression.

A. The numbers of genes significantly up- or down-regulated by at least two-fold in the presence of the doxycycline-repressed (dox off: DO) DO-ffmA allele are quantitated under conditions in which FfmA expression is high (Dox Minus) or after doxycycline addition to lower FfmA levels (Dox Plus). B. Significant categories of genes either induced (Up Dox+) or repressed (Down Dox+) by FfmA in the presence of doxycycline. C. Number of genes that are commonly regulated between the doxycycline-inhibited DO-ffmA strain and earlier experiments using a ffmAΔ allele (Paul et al. 2022a).

Figure 4.. Chromatin immunoprecipitation-high throughput sequencing (ChIP-seq) analysis of FfmA.

A. Integrative Genetics Viewer (IGV) plots of ChIP-seq data for AtrR, FfmA and RNA-seq data for FfmA are shown. ChIP-seq data of AtrR was taken from (Paul et al. 2019) and HA antibody was used to recover chromatin from isogenic cells containing or lacking an atrR-3X HA fusion gene. FfmA ChIP-seq data were generated using anti-FLAG antibody to isolate chromatin from a strain containing or lacking a FLAG-ffmA fusion gene (FLAG-ffmA). Additionally, chromatin was isolated from a wild-type strain using either an anti-FfmA polyclonal antibody or with no primary antibody (No Ab). RNA-seq data are shown for RNA isolated from a wild-type (wt) strain or from the DO-ffmA strain grown in the absence (−Dox) or presence (+Dox) of 10 mg/l doxycycline. B. Overlap in genes with ATREs and exhibiting ChIP-seq peaks for immunoprecipitated chromatin performed with either anti-FLAG or anti-FfmA as detailed above. C. RECOUP plots of ChIP-seq data from either AtrR (left hand panel) or FfmA (right hand panel). These plots were generated using ChIP-seq data for all genes from −2 kb upstream and downstream of the transcription start site (TSS) and represent the average occupancy for either AtrR or FfmA across all genes in A. fumigatus larger than 1 kb (to aid in peak separation between 5’ and 3’ ends of genes). These data correspond to ChIP-seq analysis of strains expressing two different forms of atrR-3X HA (either driven by the wild-type promoter or the strong hspA promoter) as well as a strain lacking any HA tag. The two different means of detecting FfmA-bound chromatin described above as well as their controls are plotted on the right. Note that AtrR shows the expected enrichment for a factor that binds to the 5’ end of genes while FfmA shows no such enrichment and is even more likely to bind to the region downstream of the TSS.

Figure 5.. Interaction at the level of expression between AtrR and FfmA.

A. Isogenic wild-type and atrRΔ strains were grown and whole cell protein extracts prepared. Equal amounts of these extracts were analyzed by SDS-PAGE using either anti-AtrR or anti-FfmA antisera. B. The DO-ffmA-containing strain was grown either in the presence (+) or absence (−) of doxycycline for 12 hours. Whole cell protein extracts were analyzed as described above using the same antibodies. Molecular mass standards are indicated in kD on the left.

Figure 6.. AtrR interacts with FfmA.

The DO-FLAG-ffmA strain was grown and whole cell extracts prepared under nondenaturing conditions. An aliquot of this original extract was retained and used as an input control (Input). Immunoprecipitations were conducted with anti-FLAG antibody or with no primary antibody (No Ab). Immunoprecipitates were washed and then denatured. Aliquots of immunoprecipitated proteins were analyzed by western blotting using either anti-AtrR or anti-FLAG antibody.

Figure 7.. Pigmentation caused by depletion of FfmA is dependent on expression of hasB.

A. ChIP- and RNA-seq data at the hasA/B locus. An IGV plot as shown in Figure 3A summarizes the location of the AtrR binding region as well as the absence of significant FfmA binding under these conditions. When FfmA is depleted by doxycycline treatment of the DO-ffmA strain, hasB transcription was seen to dramatically increase. B. Loss of hasB reduces the pigmentation seen in DO-ffmA. The hasB gene was disrupted with insertion of a hygromycin resistance marker and two isolates were tested for their degree of pigmentation after repression of FfmA expression by doxycycline. Note the deeper orange color of the hasB isolate compared to the lighter yellow color of the two hasBΔ strains in the presence of doxycycline.