Phenotypic plasticity and the evolution of azole resistance in Aspergillus fumigatus; an expression profile of clinical isolates upon exposure to itraconazole
The prevalence of azole resistance in clinical and environmental Aspergillus fumigatus isolates is rising over the past decades, but the molecular basis of the development of antifungal drug resistance is not well understood. This study focuses on the role of phenotypic plasticity in the evolution of azole resistance in A. fumigatus. When A. fumigatus is challenged with a new stressful environment, phenotypic plasticity may allow A. fumigatus to adjust their physiology to still enable growth and reproduction, therefore allowing the establishment of genetic adaptations through natural selection on the available variation in the mutational and recombinational gene pool. To investigate these short-term physiological adaptations, we conducted time series transcriptome analyses on three clinical A. fumigatus isolates, during incubation with itraconazole.
After analysis of expression patterns, we identified 3955, 3430, 1207, and 1101 differentially expressed genes (DEGs), after 30, 60, 120 and 240 min of incubation with itraconazole, respectively. We explored the general functions in these gene groups and we identified 186 genes that were differentially expressed during the whole time series. Additionally, we investigated expression patterns of potential novel drug-efflux transporters, genes involved in ergosterol and phospholipid biosynthesis, and the known MAPK proteins of A. fumigatus.
Our data suggests that A. fumigatus adjusts its transcriptome quickly within 60 min of exposure to itraconazole. Further investigation of these short-term adaptive phenotypic plasticity mechanisms might enable us to understand how the direct response of A. fumigatus to itraconazole promotes survival of the fungus in the patient, before any “hard-wired” genetic mutations arise.
KeywordsAspergillus fumigatus Azole resistance Phenotypic plasticity Itraconazole RNA-seq
Differentially expressed genes
High osmolarity glycerol
Log2 fold change
Mitogen activated protein kinase
Principal component analysis
Quantitative real time PCR
Aspergillus fumigatus is an ubiquitous fungus that plays an important role in carbon and nitrogen recycling in nature by degradation of organic biomass. It can be found in many terrestrial habitats, including soil and decaying organic matter . A well-known example is the compost pile, which is considered a harsh environment as large numbers of microbes compete for nutrients, oxygen and space . Various biological characteristics allow A. fumigatus to thrive in this environment, including rapid and efficient germination, tolerance for elevated temperatures, and a metabolism that is responsive to variation in nutrient sources [3, 4]. Furthermore, A. fumigatus propagates via asexual conidia that can be dispersed over wide geographic distances by air currents, and that can germinate to grow under a broad range of environmental conditions . To survive and thrive, Aspergillus needs to rapidly adapt to these often challenging environments.
The remarkable ability of A. fumigatus to grow in diverse environments, contributes to its role as the predominant fungal pathogen of immunocompromised patients. As a result, Aspergillus infections are increasingly recognized as a global health problem. In humans, several diseases may be caused by A. fumigatus, ranging from allergic conditions to acute invasive aspergillosis [5, 6]. Azole antifungals are the primary choice for prevention and treatment of Aspergillus infections. Azoles inhibit the 14-α demethylase Cyp51A, as they function as a ligand of the heme group and interact with the protein structure through the azole drug side chains . The inhibition of Cyp51A results in accumulation of toxic sterols, destabilisation of the plasma membrane and subsequent cell death . However, A. fumigatus is able to adapt to the presence of azoles, which facilitates persistence of the species and compromises treatment options in the clinic .
Adaptation can be defined as the acquisition of adaptive traits through natural selection, which enables the organism to adjust to life in the prevailing or new environment . Before natural selection occurs, the organism quickly adapts to a new environment by exerting its phenotypic plasticity (PP). PP allows organisms to adjust their physiology to adapt to a new and challenging environment, to still be able to maintain growth and reproduction [11, 12]. These adaptations precede the development of genetic changes and the resulting newly acquired traits. In filamentous fungi, these traits often arise by spontaneous mutation and subsequent selection if the new genotype has achieved a higher overall fitness through its phenotypic traits. Crucially, in A. fumigatus, this process has led to the development of azole resistance [13, 14, 15, 16]. Azole resistant strains have emerged in patients with aspergilloma or chronic pulmonary aspergillosis during azole therapy, but also in for instance organic-waste that contained azole residues. Azole compounds are widely used for various applications including food production, plant protection, and material preservation . Thus, any environment that contains azole compounds can result in azole resistance development in A. fumigatus, provided that the fungus can complete its life-cycle [15, 18, 19]. As the dominant mechanism of resistance (TR34/L98H) is suggested to have originated in the environment, various studies have focused on the role of azole use in agriculture in the development of azole-resistant isolates [20, 21, 22].
Previously, a lot of research focused on elucidating the genetic basis of azole resistance. The most common mechanisms of resistance in A. fumigatus are modifications in the cyp51A gene. Various non-synonymous mutations in this gene have been correlated to azole resistance [13, 23, 24, 25]. These mutations cause structural changes in the conserved active site, preventing azoles from binding to the enzyme . It is not well understood how the fungus maintains fitness before these mutations arise, and knowledge of the molecular basis of this development is needed in order to assess the impact of this phenomenon.
To understand the development of resistance, we previously investigated whether genetic substitutions could be induced during exposure to azoles under laboratory conditions [9, 26]. Wild type A. fumigatus isolates were exposed to both medical and environmental azoles, or a mixture of these compounds. The induction experiments generally resulted in a resistant phenotype already within three passages. However, genetic alterations were only found after at least five passages. In three out of 12 clones of A. fumigatus cultured under itraconazole pressure, cyp51A-substitutions G138C or P216L were detected. These substitutions have been reported in patients who developed azole-resistant Aspergillus disease during itraconazole therapy [24, 25].
It is not well understood how the fungus maintains fitness before these mutations arise, and thus current knowledge suggests that adaptation follows a two-step process. Firstly, patient- and environmentally acquired resistance both require an actively reproducing fungus even when exposed to azole compounds. Therefore, PP adjusts the physiology of A. fumigatus to maintain growth and reproduction at least to some extent . Secondly, after these physiological adjustments, the process of natural selection maintains the mutations from the gene pool that confer genetic resistance in the azole-containing environment. Clearly, it is important to understand the adaptation strategies of A. fumigatus in relation to azole resistance selection and the clinical implications thereof for management of Aspergillus diseases.
The capability of A. fumigatus to adapt in the human host or in the outside environment is an important fungal property, which contributes to treatment failure and to the increased numbers of resistant A. fumigatus found worldwide. Therefore, this report focuses on the role of PP in the evolution of azole resistance. For A. fumigatus, this plasticity may involve, for instance, increased efflux pump activity, the activation of several stress-response pathways, or altering its lipid metabolism as ergosterol depletion affects membrane homeostasis [28, 29, 30, 31]. However, research is lacking on the cellular events that take place in the fungus directly after introduction of an antifungal compound to the environment. Furthermore, it is not known how the plasticity of the fungal physiology facilitates the selection of mutations. The early response is crucial, as the fungus needs to retain its metabolism and preferably its reproductive capability. In this study, the differential expression of the A. fumigatus genome was measured at various time points, directly after addition of itraconazole to the medium. The results provided important details on the initial transcriptional dynamics, and gave insights into membrane transporter activity, lipid metabolism, and MAPK cascade proteins, implicated to be involved in the process of phenotypic plasticity. These insights will allow us to find leads for targeted interventions.
Three clinical A. fumigatus isolates used for transcriptomic analysis in this study. The minimum inhibitory concentration (MIC) was determined for itraconazole (ICZ), voriconazole (VCZ) and posaconazole (PCZ)
Date of isolation
Analysis of differentially expressed genes
Verification of RNA-seq data by qRT-PCR
To check the quality of our RNA-seq dataset, we verified the expression of eight differentially expressed genes (DEGs) over a range of fold changes by conducting quantitative real time PCR (qRT-PCR) experiments. Genes were selected from different categories: up-regulated, down-regulated or non-regulated. Patterns of the RNA-seq analysis were highly similar to the patterns obtained from the qRT-PCR assays, confirming that our sequencing results were reliable (Additional file 4: Figure S3).
Functional characterization of DEGs
Of the 30 most enriched GO-terms in the Biological process category, the terms ‘ribosome biogenesis’, ‘ncRNA metabolic process’, ‘gene expression’ and ‘RNA processing’ were most significant in the up-regulated genes after 30 min, and in the down-regulated genes after 240 min. In addition, most enriched processes in the up-regulated genes were related to amino acid metabolism, nucleic acid metabolism, transcription, translation and general transport of RNA and proteins. The only processes enriched in the down-regulated genes after 30 and 60 min were ‘fatty acid biosynthetic process’, ‘interspecies interaction between organisms’ and ‘oxidation-reduction process’. These results indicate a decreased production of fatty acids during the first 60 min. Furthermore, the enrichment of ‘interspecies interaction between organisms’ indicates a decreased expression of genes involved in, for instance, host interaction.
After 240 min, many terms were enriched in the down-regulated genes, including ‘response to drug’ and ‘response to chemical’, indicating that transcription of genes that was altered as a response to the addition of itraconazole, is especially important during the first 2 h of incubation.
In the category Molecular Function, the term ‘drug binding’ was enriched in the up-regulated genes after 30 min and enriched in the down-regulated genes after 240 min, indicating the increase of gene products that bind specifically to a drug during the first 30 min. Furthermore, ‘unfolded protein binding’, was enriched in the down-regulated genes after 240 min. This term mainly refers to binding unfolded ER proteins, indicating a possible increase of incorrectly folded proteins that activate an ER-stress response.
In the Component category, various terms specifically referred to adjustments to the exterior of the cell. As itraconazole disturbs ergosterol biosynthesis, it is likely that membrane integrity is impaired. Interestingly, the cell adjusts various transcriptional processes related to the plasma membrane and the fungal cell wall, as a response to itraconazole. The GO term ‘plasma membrane’ was the only enriched term after 30 min in the down-regulated genes. After 60 min, the term ‘cell periphery’ was also enriched in the down-regulated genes, and the terms ‘hyphal cell wall’ and ‘cell surface’ were enriched in the down-regulated genes after 240 min.
DEGs during all time points
As the majority of all genes is differentially expressed only transiently, it would be valuable to assess whether certain genes would be differentially expressed during the whole time course, indicating an important role for these genes in the transcriptional response to itraconazole.
Furthermore, nine transcription factors, three RING finger proteins, five heat shock proteins, six MFS transporters and 26 hypothetical proteins with no current known function were identified.
A complete overview of all differentially expressed core group genes is shown in Additional file 5: Table S2.
Analysis of the top DEGs
Expression profile of genes encoding putative membrane transporters
The effect of itraconazole on membrane homeostasis
Azole compounds interfere with ergosterol biosynthesis, which is suggested to have a significant impact on membrane rigidity . Therefore, we analyzed all genes involved in the ergosterol biosynthesis pathway. Additionally, we analyzed the key pathways involved in phospholipid biosynthesis; the CDP-DAG pathway and the Kennedy pathway, which are postulated to be very similar to the Saccharomyces cerevisiae pathways . These pathways are involved in the production of phospholipids from phosphatidic acid de novo, and the production of phosphatidylethanolamine (PE) and phosphatidylcholine (PC) from exogenous ethanolamine and choline, respectively . Therefore, they play an important role in maintaining membrane integrity and DEGs in these pathways could provide leads on how A. fumigatus exerts its phenotypic plasticity.
Differentially expressed genes in the ergosterol biosynthesis pathway
In the steroid biosynthesis pathway, an overall down-regulation of genes within 30 min was observed, including down-regulation of cyp51A and cyp51B (Fig. 7b). Out of the 20 genes, 15 genes were found to be down-regulated. The strongest up-regulation at all time points was measured at the erg6 gene, which functions directly upstream of Cyp51A and Cyp51B. The other genes that were measured to be up-regulated, were the C-8 sterol isomerase gene erg2 and the C-24 sterol reductase erg4.
Phosphatidylethanolamine biosynthetic genes show fast up-regulation after 30 min
Differential expression of MAP kinases
Comparison with transcriptome data of several stresses
The goal of this study was to investigate differential gene expression in azole-susceptible clinical strains after the addition of sub-lethal concentrations of itraconazole, to gain valuable insights into the way A. fumigatus exerts its phenotypic plasticity and manages to survive. We gained insight in the general expression patterns which take place as A. fumigatus adapts to the presence of an antifungal compound, and identified 25 potential drug-efflux transporters which potentially aid the fungus to survive by exporting itraconazole from the cell. Furthermore, we demonstrated that the erg6 gene was up-regulated during the whole time course and that the genes facilitating PE production from extracellular ethanolamine were up-regulated, indicating an increased production of this phospholipid. Finally, we observed an increased expression of several MAPK cascade proteins which are also involved in HOG signaling. This study is, to our knowledge, the first study to use RNA-seq to measure gene expression of clinical isolates in response to itraconazole.
In our experimental setup, we compared the isolates of each time point with the isolates directly harvested after 24 h. Although we do not distinguish between the transcriptional changes caused by itraconazole or other changes related to mycelial growth progression, previous research demonstrated that A. fumigatus resides in the exponential growth phase after 24 h, in various nutrient media . Interpreting transcriptomics results is limited by the fact that only 487 (4.95%) genes have been experimentally verified and characterized in A. fumigatus, making the interpretation of transcriptomics results challenging (Aspergillus fumigatus Af293 Genome Snapshot, AspGD, May 14th, 2018). Although many gene functions are predicted computationally through known functions of homologous genes in other closely related species, they are yet to be experimentally confirmed. It was notable that the number of enriched GO terms per time point was not a direct consequence of the number of genes in the group; the 2254 down-regulated genes after 30 min yielded only five enriched GO terms in the category process, as opposed to the 1701 up-regulated genes after 30 min, which yielded a total of 194 enriched GO terms in the category process before the removal of redundant terms. This suggests that many cellular processes are down-regulated after 30 min, but all to similar extent as no specific terms are enriched.
GO term enrichment analysis showed increased activity of various specific processes related to the exterior of the cell, consistent with previous research that states that itraconazole decreases membrane integrity by interfering with ergosterol biosynthesis [40, 57]. The cellular component category also showed the enriched term ‘extracellular matrix’ in the up-regulated genes after 60 min, which mainly comprises of genes which are part of the fungal biofilm matrix. These biofilms are often involved in host adherence and provide protection from extracellular stresses, suggesting an important role for biofilm formation in the physiological response of A. fumigatus [58, 59].
The involvement of elastase in fungal virulence is discussed in several studies, and its contribution to pathogenicity, by breaking down elastin in lung tissue, is not well established for A. fumigatus [60, 61, 62, 63]. We found that the putative elastase inhibitor aeiA (Afu3g14940) was the most down-regulated gene in our dataset (log2FC 8.27). There is a possibility that these clinical strains have specifically adapted to the lung environment, and that there are other beneficial effects that do not influence virulence, although these results suggest that A. fumigatus greatly adjusts its elastase activity upon itraconazole stress.
Amongst the strongest up-regulated genes were two putative RTA1 domain proteins. Not much is known about this gene family, that was initially found to be involved in the export of long sphingoid bases in S. cerevisiae . Interestingly, overexpression of ScRTA1 results in 7-aminocholesterol resistance, an antifungal compound that inhibits sterol biosynthesis. Heme-deficiency was also shown to induce ScRTA1 expression. It is likely that disturbance of ergosterol biosynthesis influences iron homeostasis in the fungal cell, as these processes share mevalonate as a precursor and are both regulated by the sterol regulatory element binding protein SrbA [65, 66]. When heme or iron sources are deficient, siderophore mediated iron-uptake is increased, and hmg1 expression is up-regulated. This is consistent with our finding that hmg1 expression is increased [64, 66]. The role of RTA1 proteins in the physical response to azoles has not yet been established and could provide an interesting lead.
A large number of transporters have been described in A. fumigatus, and the fungus can defend itself against a large number of toxic compounds via for instance drug-efflux pump activity [67, 68]. The genome of A. fumigatus is predicted to contain at least 49 ATP binding cassette (ABC) family transporter genes and 278 major facilitator superfamily (MFS) genes. From the AspGD website, only genes encoding transporters were extracted that were predicted to play a role in drug efflux specifically. This might leave out potential other transporter genes, which are not yet computationally predicted to be involved in drug efflux in general. Nevertheless, 25 transporters showed comparable expression patterns to the verified drug-efflux transporters within the first. These transporters could play an important role in physical resilience induced by PP, and their involvement should be addressed experimentally in future research.
We investigated to what extent itraconazole influences the expression of ergosterol biosynthetic genes. Alterations in the ergosterol biosynthesis pathway have been shown to confer resistance in a previous study with Candida albicans . Strikingly, erg6 showed a significant up-regulation of log2FC 1.78–3.54 during the whole time course, as opposed to the rest of the ergosterol biosynthetic pathway, which showed a general down-regulation, or no-regulation. Recent studies suggested that not Cyp51A, but Erg6 is the 14α-lanosterol demethylase in A. fumigatus, as opposed to S. cerevisiae, where ScERG6 encodes a sterol 24-C-methyltransferase further downstream [8, 70, 71, 72]. Erg6 functions directly upstream of Cyp51A, where it first converts lanosterol to eburicol. Inhibition of Cyp51A by azoles and up-regulation of erg6 would then lead to a strong accumulation of eburicol. Why the increased production of eburicol would be a favorable trait when this compound possibly already accumulates as the Cyp51A enzyme is inhibited by azoles , is not clear. It is possible that a feedback regulation mechanism senses the absence of the Cyp51A/B product, 4,4,24-trimethylcholesta-8,24(28)-dien-3β-ol, when they are inhibited by azoles with the increased production of their substrate as a direct result.
The genes involved in biosynthesis of phosphatidylethanolamine (PE) from extracellular ethanolamine showed a predominant up-regulation. A recent study on mammalian cells suggests that PE is the key phospholipid in maintaining proper membrane rigidity, and they found that a decrease in cholesterol resulted in a strong increase of PE relative to the other phospholipids . These genes were identified as the most up-regulated of all lipid biosynthesis genes investigated in this study and could imply an increase of PE in the membranes to maintain proper rigidity, when ergosterol levels are depleted [45, 74].
The important role of cell signaling in fungal adaptation to stress factors and antifungals has been extensively demonstrated [31, 75, 76, 77]. We have investigated the involvement of MAPK cascade proteins upon azole stress. The mpkB gene is still uncharacterized in A. fumigatus, however, the mpkB gene in A. nidulans and the FUS3 gene in S. cerevisiae are well-described homologs. Their role in the cellular pheromone response and sexual development is established [78, 79], indicating a possible role of these processes in the response to itraconazole stress. Furthermore, we measured a significant up-regulation in the sakA MAPK gene, which is homologous to the HOG1 gene of S. cerevisiae. This gene is the key MAP kinase involved in the regulatory processes in response to osmotic, oxidative, heat, and hypoxia mediated stresses . In A. fumigatus, sakA has been shown to play an important role in the osmotic and oxidative stress responses. Deletion mutants showed an increased sensitivity to itraconazole and amphotericin B , which is consistent with our finding that sakA expression is increased as a response of A. fumigatus to itraconazole. Our results suggest that upon exposure to itraconazole, A. fumigatus differentially regulates genes which are related to the osmotic stress response and increases its glycerol production. This is supported by our findings that at least three genes involved in glycerol formation were up-regulated within 60 min. Transcriptional changes of MAP kinases could give insight into the stress-response mechanisms , although MAP kinases are activated by phosphorylation . Therefore, analysis of the phosphoproteome could provide additional insights into their direct involvement in adaptation to itraconazole stress.
Recently, the mitochondrial complex I has been implicated to play a role in azole resistance. Interestingly, our results show only a slight down-regulation of the 29.9 KD subunit of mitochondrial complex I (Afu2g10600: Log2FC -0.81, P-value 0.01, adjusted P-value 0.07) (Additional file 7: Table S4) after 240 min, although deletion of this gene leads to partial azole resistance (80). The authors suggest a resistance mechanism based on restoring an unbalanced oxidative stress response when itraconazole (1 mg/L) is added to the medium. However, the DEGs in our dataset do not indicate an oxidative response in our isolates exposed to itraconazole, as we did not measure up-regulation of genes that are induced upon hypoxic conditions . Therefore, a down-regulation of the 29.9 KD subunit would not be necessary to rebalance the oxidative stress response. It is also possible that this gene is not differentially regulated and does not contribute to PP to gain azole resilience, although further research is necessary to elucidate if higher concentrations of itraconazole could result in a different expression pattern.
Many mechanisms that cope with cell stress respond to various stressful conditions, and are not limited to a specific stress condition or a specific antifungal compound [47, 83, 84]. To investigate if A. fumigatus differentially expresses genes solely in response to itraconazole, we compared our results to the transcriptomics data of previously conducted studies [47, 54]. We found that 251 genes were differentially regulated specifically in response to itraconazole. Although genetic and physiological differences between isolates compared across different studies may influence this outcome, these results suggest that transcription of these genes is specifically altered as a response to itraconazole stress. Amongst these genes is the down-regulated insA (Afu4g07680) gene, which is a putative sterol sensing factor. The mammalian homolog encodes an ER-anchored protein which prevents the transcriptional activation of genes involved in the synthesis and uptake of cholesterol, by binding to sterol regulatory element binding proteins (SREBPs) . It is possible that A. fumigatus anticipates on ergosterol deficiency by down-regulating this gene to prevent inhibition of sterol synthesis and uptake, although little is known about the sterol sensing mechanisms in A. fumigatus.
Remarkably, previous reported DEGs resulting from voriconazole treatment share but 72,6% with DEGs from our study, despite of their shared mechanism of action. These difference could be appointed to the treatment of the isolates with a higher concentration of voriconazole (0.5 μg/mL) and although the exact MIC of the tested isolates was not documented, this could explain the different DEGs found. Additionally, the isolates tested in this study demonstrated that even between different isolates that have the same MIC, the basal expression of the transcriptome can differ. Furthermore, the difference could be appointed to the comparison of microarray results with RNA-seq results, with the latter having higher resolution. Future studies could be improved by introduction of a conserved laboratory strain in every experiment, which would facilitate better comparison.
As demonstrated in this study, RNA-seq is a promising method to provide better insight in the physiological changes and cellular defense mechanisms [86, 87, 88]. Although many processes can be monitored this way, it should be taken into account that a decreased or increased transcription, does not necessarily lead to an increased translation. Additionally, post-translational processes could further impact on the suggested increased abundance of activated proteins. It is essential that potential genes or pathways involved in the cellular response to itraconazole are confirmed by the generation of gene-deletion mutants. Furthermore, it could be promising to look at the direct consequences of the various transcriptional changes related to ergosterol and lipid biosynthesis, by investigating the full lipidome of A. fumigatus mycelia under various conditions.
Our data strongly suggests that phenotypic plasticity allows A. fumigatus to substantially adapt its transcriptome within 60 min of exposure to itraconazole, after which the cells proposedly reach an improved homeostatic state. We showed that short-term adaptation through PP includes immediate up-regulation of several drug-efflux membrane transporters, activation of signaling cascades, and various alterations in gene expression in the ergosterol biosynthesis and phospholipid biosynthetic processes. Thus, our study supports the notion that transcriptomic analyses by RNA-seq can give major insights in the differentially expressed genes, enabling us to understand how the direct response of A. fumigatus to itraconazole promotes survival of the fungus in the patient and in the outside environment before any stable genetic mutations arise. Further research on PP and the short-term adaptation processes of A. fumigatus can improve the understanding on how the fungus survives during azole exposure. Increasing this fundamental knowledge is crucial in managing the spread of azole resistant A. fumigatus isolates and developing future antifungal therapies.
Isolates, media and culture conditions
Isolates were used with an MIC of 0.5 or 1 for itraconazole and the same MIC for voriconazole and posaconazole respectively, to exclude the possibility that great differences in outcome could be assigned to the differences in MICs for these isolates. The isolates were selected from three different patients from separate hospitals in the Netherlands. All isolates are listed in Table 1. Isolates were cultured on either Sabouraud Dextrose Agar (SDA) (Oxoid) (1% peptone, 4% glucose, 1.5% agar, pH 5.6) or Aspergillus Minimal Medium (AMM) as described in the supplemental material of reference  containing per liter: 10 g glucose, 5.95 g NaNO3, 0.522 g KCl, 1.5 g KH2PO4, 50 mg MgSO4·7H2O, 1 mL trace elements). Trace elements contained (per 200 mL): 10 g EDTA, 4.4 g ZnSO4·7H2O, 1.01 g MnCl2·4H2O, 0.315 g CuSO4·5H2O, 0.22 g (NH4)6Mo7O24·4H2O, 1.0 g Fe(II)SO4·7H2O, and 2.2 g H3BO3 . All compounds were produced by Merck (Darmstadt, Germany). Isolates were stored in 20% glycerol at − 80 °C, and subcultured on SDA at 37 °C for 4–5 days. Conidia were harvested with a wet cotton swab and resuspended in Milli-Q containing 0.1% Tween 20. Approximately 2,6 × 105 CFU/mL of conidia were used to inoculate 5 mL of AMM in 15 mL tubes. Conidia were inoculated in duplicate. After 24 h of growth at 37 °C while rotating, itraconazole (Sigma) was added at IC50 to each tube except the T = 0 tubes, which were immediately harvested through Miracloth (Merck). Remaining mycelia were harvested through Miracloth after 30, 60, 120 and 240 min. Harvested mycelia were directly frozen in liquid nitrogen.
For determining the itraconazole concentration where growth is estimated to be 50% (IC50), dose-response curves were made by growing each strain (n = 5) in a 96-wells plate containing AMM with itraconazole concentrations ranging from 0.016 μg/mL to 8 μg/mL. Plates were kept at 37 °C in a microtiter plate reader (Anthos Labtec Instruments GmbH, Salzburg, Austria) for 48 h, and optical density (OD) was measured every hour at 405 nm. After subtraction of the ODs of the empty wells, the percentage of growth from the ODs of the inoculated wells for each well was correlated with the relative OD estimated by the following equation: (OD405 of wells that contained itraconazole/OD405 of the itraconazole-free well) × 100. The relationship was determined by a non-linear regression analysis and the Hill equation with a variable slope fitted to the data. Goodness of fit was checked by the r2 values. All analyses were performed by using GraphPad Prism, version 6.0, for OS X (GraphPad Software, San Diego, USA).
After harvesting, mycelia were lysed by bead-bashing using a MagNa Lyser instrument, three times for 30 s at 7000 rpm (Roche, Basel, Switzerland). RNA was extracted using TRIzol reagent (Invitrogen, Breda, the Netherlands). Pellets were resuspended in 100 μL DEPC treated water. RNA integrity was verified by agarose gel electrophoresis. Total RNA was stained with Midori Green (Nippon Genetics, Dueren, Germany) and visualized with UV light. Purity of all samples (A260/A280 ≥ 1.5 and A260/A230 ≥ 2.0) was determined using a NanoDrop 1000 (Thermo Scientific, MA, US). Pooled cDNA libraries were constructed with the TruSeq RNA Sample Prep Kit v2 (Illumina, San Diego, USA) according to manufacturer instructions. Quality of the libraries was verified with an Agilent 4200 Tapestation System (Agilent Technologies). Paired-end reads of 2 × 75 bp were generated in High Output mode on a NextSeq 500 sequencer (Illumina, San Diego, USA).
Identification of differentially expressed genes
Reads were aligned with STAR v2.5.3A  against the reference genome sequence of A. fumigatus Ensemble CADRE 30 downloaded from EnsemblFungi. Differentially expressed genes (DEGs) were identified by DESeq2 for R by Bioconductor (BaseMean ≥30, P-value < 0.05, adjusted P-value < 0.1, log2FC > 1 or < − 1), and results were corrected for multiple testing by the Benjamini-Hochberg principle .
Functional categories of DEGs were grouped according to the Gene Ontology (GO) term, using the GO Term Finder on the AspGD website to identify significant terms (p ≤ 0.05) . The False Discovery Rate cutoff value was set to p ≤ 0.05. REVIGO was used to remove redundant GO terms  and allowed similarity was set to 0.5. Terms with over 30% frequency were removed to avoid general terms. Up-regulated and Down-regulated genes were screened separately for enriched categories, to improve statistical power .
Validation of RNA-seq results by quantitative real-time PCR
All primers are listed in Additional file 9: Table S6. Eight genes with different expression patterns were selected, and expression was confirmed in all three strains separately. RNA isolation and cDNA synthesis were performed as described above. Quantitative real-time PCR was performed with gene-specific primers, using the actin gene (act1) as reference gene. PCR was performed using Roche Lightcycler 480 SYBR Green I Master according to the protocol of the manufacturer. PCR conditions were as follows: Denaturation at 95 °C for 5 min, followed by 45 cycles of denaturation at 95 °C for 10 s, primer annealing at 60 °C for 10 s, and extension at 72 °C for 10 s. Expression levels were calculated by the ΔΔCT-method, and normalized for actin mRNA expression levels . All qRT-PCR experiments were per-formed at least in duplicates.
Statistical analysis and figure construction
Heatmaps and volcano plots were generated by Rstudio Version 1.0.143 for OS X. For generating heatmaps, gplots v3.0.1 was used. For generating volcano plots, ggplot2 V2.2.1 was used. Exported images were assembled with Adobe Illustrator CC 2015. Venn diagrams were created with Biovenn (http://www.biovenn.nl/index.php) and the Venn tool by (http://bioinformatics.psb.ugent.be/webtools/Venn). Graphs of qRT-PCR and dose-response curve analyses were generated by using GraphPad Prism, version 6.0, for OS X (GraphPad Software, San Diego, USA). Error bars represent standard error of the mean.
We would like to thank Hein van der Lee and Marlou Tehupeiory-Kooreman for their excellent experimental support and their fungal expertise. We would also like to thank Mike Renne for valuable discussions and critical reading of the manuscript.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Availability of data and materials
The sequences reported in this paper have been deposited in the NCBI Sequence Read Archive (SRA) database, in a BioProject under accession number PRJNA482512.
MWJH participated in designing the research, carried out the RNA-seq, analysed data and wrote the manuscript. JZ participated in designing the research, and carried out and analysed the qRT-PCR experiments. JPMC analysed RNA-seq data. BJZ conceived the study and wrote the manuscript. PEV participated in designing the research and coordinated the study. WJGM conceived and coordinated the study, participated in designing the research and wrote the manuscript. All of the authors read and approved the final manuscript.
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 11.Stern S, Dror T, Stolovicki E, Brenner N, Braun E. Genome-wide transcriptional plasticity underlies cellular adaptation to novel challenge. Mol Syst Biol. 2007;3(106):1–9.Google Scholar
- 12.Agrawal AA. Phenotypic plasticity in the interactions and evolution of species. Science. 2001;294(5541):321–6.Google Scholar
- 16.Zhang J, Debets AJM, Verweij PE, Melchers WJG, Zwaan BJ, Schoustra SE. Asexual sporulation facilitates adaptation: the emergence of azole resistance in Aspergillus fumigatus. Evolution (N Y). 2015;69(10):2573–86.Google Scholar
- 25.Mann PA, Parmegiani RM, Wei SQ, Mendrick CA, Li X, Loebenberg D, et al. Mutations in Aspergillus fumigatus resulting in reduced susceptibility to posaconazole appear to be restricted to a single amino acid in the cytochrome P450 14α-demethylase. Antimicrob Agents Chemother. 2003;47(2):577–81.PubMedPubMedCentralCrossRefGoogle Scholar
- 26.Camps SMT, Van Der Linden JWM, Li Y, Kuijper EJ, Van Dissel JT, Verweij PE, et al. Rapid induction of multiple resistance mechanisms in Aspergillus fumigatus during azole therapy: a case study and review of the literature. Antimicrob Agents Chemother. 2012;56(1):10–6.PubMedPubMedCentralCrossRefGoogle Scholar
- 37.Perlin MH, Andrews J, Toh SS. Essential letters in the fungal alphabet: ABC and MFS transporters and their roles in survival and pathogenicity. Adv Genet. 2014;85:201–53 Elsevier.Google Scholar
- 39.Paul S, Diekema D, Moye-Rowley WS. Contributions of both ATP-binding cassette transporter and Cyp51A proteins are essential for azole resistance in Aspergillus fumigatus. Antimicrob Agents Chemother. 2017;61(5):e02748-16.Google Scholar
- 42.Henry SA, Kohlwein SD, Carman GM. Metabolism and regulation of glycerolipids in the yeast Saccharomyces cerevisiae. Genetics. 2012;190(2):317–49.Google Scholar
- 57.Ghannoum MA, Rice LB. Antifungal agents: mode of action, mechanisms of resistance, and correlation of these mechanisms with bacterial resistance. ClinMicrobiolRev. 1999;12:501–17.Google Scholar
- 69.Vale-Silva LA, Coste AT, Ischer F, Parker JE, Kelly SL, Pinto E, et al. Azole resistance by loss of function of the sterol Δ5,6- desaturase gene (ERG3) in Candida albicans does not necessarily decrease virulence. Antimicrob Agents Chemother. 2012;56(4):1960–8.PubMedPubMedCentralCrossRefGoogle Scholar
- 71.Alcazar-Fuoli L, Mellado E. Ergosterol biosynthesis in Aspergillus fumigatus: its relevance as an antifungal target and role in antifungal drug resistance. Front Microbiol. 2012;3:1–6.Google Scholar
- 80.Kim JH, Chan KL, Faria NCG. Martins M de L, Campbell BC. Targeting the oxidative stress response system of fungi with redox-potent chemosensitizing agents. Front Microbiol. 2012;3(MAR):1–11.Google Scholar
- 82.Kroll K, Pähtz V, Hillmann F, Vaknin Y, Schmidt-Heck W, Roth M, et al. Identification of hypoxia-inducible target genes of Aspergillus fumigatus by transcriptome analysis reveals cellular respiration as an important contributor to hypoxic survival. Eukaryot Cell. 2014;13(9):1241–53.PubMedPubMedCentralCrossRefGoogle Scholar
- 83.Cowen LE. Hsp90 potentiates the rapid evolution of new traits: drug resistance in diverse fungi. Science. 2015;309(5744):1–6.Google Scholar
- 88.Daly P, van Munster JM, Blythe MJ, Ibbett R, Kokolski M, Gaddipati S, et al. Expression of Aspergillus niger CAZymes is determined by compositional changes in wheat straw generated by hydrothermal or ionic liquid pretreatments. Biotechnol Biofuels. 2017;10(1):35.PubMedPubMedCentralCrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.