The Heat Shock Transcription Factor HsfA Plays a Role in Membrane Lipids Biosynthesis Connecting Thermotolerance and Unsaturated Fatty Acid Metabolism in Aspergillus fumigatus

ABSTRACT Thermotolerance is a remarkable virulence attribute of Aspergillus fumigatus, but the consequences of heat shock (HS) to the cell membrane of this fungus are unknown, although this structure is one of the first to detect changes in ambient temperature that imposes on the cell a prompt adaptative response. Under high-temperature stress, fungi trigger the HS response controlled by heat shock transcription factors, such as HsfA, which regulates the expression of heat shock proteins. In yeast, smaller amounts of phospholipids with unsaturated fatty acid (FA) chains are synthesized in response to HS, directly affecting plasma membrane composition. The addition of double bonds in saturated FA is catalyzed by Δ9-fatty acid desaturases, whose expression is temperature-modulated. However, the relationship between HS and saturated/unsaturated FA balance in membrane lipids of A. fumigatus in response to HS has not been investigated. Here, we found that HsfA responds to plasma membrane stress and has a role in sphingolipid and phospholipid unsaturated biosynthesis. In addition, we studied the A. fumigatus Δ9-fatty acid desaturase sdeA and discovered that this gene is essential and required for unsaturated FA biosynthesis, although it did not directly affect the total levels of phospholipids and sphingolipids. sdeA depletion significantly sensitizes mature A. fumigatus biofilms to caspofungin. Also, we demonstrate that hsfA controls sdeA expression, while SdeA and Hsp90 physically interact. Our results suggest that HsfA is required for the adaptation of the fungal plasma membrane to HS and point out a sharp relationship between thermotolerance and FA metabolism in A. fumigatus. IMPORTANCE Aspergillus fumigatus causes invasive pulmonary aspergillosis, a life-threatening infection accounting for high mortality rates in immunocompromised patients. The ability of this organism to grow at elevated temperatures is long recognized as an essential attribute for this mold to cause disease. A. fumigatus responds to heat stress by activating heat shock transcription factors and chaperones to orchestrate cellular responses that protect the fungus against damage caused by heat. Concomitantly, the cell membrane must adapt to heat and maintain physical and chemical properties such as the balance between saturated/unsaturated fatty acids. However, how A. fumigatus connects these two physiological responses is unclear. Here, we explain that HsfA affects the synthesis of complex membrane lipids such as phospholipids and sphingolipids and controls the enzyme SdeA, which produces monounsaturated fatty acids, raw material for membrane lipids. These findings suggest that forced dysregulation of saturated/unsaturated fatty acid balance might represent novel strategies for antifungal therapy.

ABSTRACT Thermotolerance is a remarkable virulence attribute of Aspergillus fumigatus, but the consequences of heat shock (HS) to the cell membrane of this fungus are unknown, although this structure is one of the first to detect changes in ambient temperature that imposes on the cell a prompt adaptative response. Under hightemperature stress, fungi trigger the HS response controlled by heat shock transcription factors, such as HsfA, which regulates the expression of heat shock proteins. In yeast, smaller amounts of phospholipids with unsaturated fatty acid (FA) chains are synthesized in response to HS, directly affecting plasma membrane composition. The addition of double bonds in saturated FA is catalyzed by D9-fatty acid desaturases, whose expression is temperature-modulated. However, the relationship between HS and saturated/unsaturated FA balance in membrane lipids of A. fumigatus in response to HS has not been investigated. Here, we found that HsfA responds to plasma membrane stress and has a role in sphingolipid and phospholipid unsaturated biosynthesis. In addition, we studied the A. fumigatus D9-fatty acid desaturase sdeA and discovered that this gene is essential and required for unsaturated FA biosynthesis, although it did not directly affect the total levels of phospholipids and sphingolipids. sdeA depletion significantly sensitizes mature A. fumigatus biofilms to caspofungin. Also, we demonstrate that hsfA controls sdeA expression, while SdeA and Hsp90 physically interact. Our results suggest that HsfA is required for the adaptation of the fungal plasma membrane to HS and point out a sharp relationship between thermotolerance and FA metabolism in A. fumigatus. IMPORTANCE Aspergillus fumigatus causes invasive pulmonary aspergillosis, a life-threatening infection accounting for high mortality rates in immunocompromised patients. The ability of this organism to grow at elevated temperatures is long recognized as an essential attribute for this mold to cause disease. A. fumigatus responds to heat stress by activating heat shock transcription factors and chaperones to orchestrate cellular responses that protect the fungus against damage caused by heat. Concomitantly, the cell membrane must adapt to heat and maintain physical and chemical properties such as the balance between saturated/unsaturated fatty acids. However, how A. fumigatus connects these two physiological responses is unclear. Here, we explain that HsfA affects the synthesis of complex membrane lipids such as phospholipids and sphingolipids and controls the enzyme SdeA, which produces monounsaturated fatty acids, raw material for membrane lipids. These findings suggest that forced dysregulation of saturated/unsaturated fatty acid balance might represent novel strategies for antifungal therapy. reduction in the phospholipid concentration in the wild-type strain. Despite that, these results suggest a possible inhibitory role of HsfA on phospholipid biosynthesis, which is more evident during the HS.
When the temperature rises, one of the first responses associated with the cell membrane is the decrease in unsaturated FA synthesis to maintain lower fluidity and sustain plasma membrane properties (44). To gather information about the chemical diversity of the phospholipids that accumulated in our experimental conditions, the quantification results were categorized based on the saturation level of the FA comprising each phospholipid ( Fig. 1B to D). The accumulation pattern of saturated phospholipids in both strains was similar pre-and post-HS. However, while saturated PC and PS had a nonsignificant trend toward increasing, saturated PE significantly decreased throughout the HS in both strains (#, Fig. 1C). When the levels of unsaturated phospholipids were compared between wild-type and xylP::hsfA, there was a significant increase in PC and PS concentration, especially after 15  . Overall, the results show that, differently from PC and PS, the concentration of FA that builds up PE is significantly reduced during the HS in both strains, regardless of the saturation number of the FA. Altogether, these results suggest that hsfA depletion affects the levels of mono-and polyunsaturated FA phospholipids, pointing out that HsfA repression exerts a negative global role in unsaturated phospholipid biosynthesis exclusively during HS.
Next, we investigated the role of HsfA during plasma membrane stress. The wild-type and xylP::hsfA strains were exposed to drugs that disrupt the plasma membrane homeostasis by interfering with sphingolipids, ergosterol, and FA biosynthesis. Given that xylP::hsfA strains do not grow in the complete absence of xylose, for these experiments, solid MM (1% glucose) was supplemented with 0.06% xylose, a concentration at which the radial growth of the mutant was equivalent to the wild-type strain in the absence of stress, as described elsewhere (36). Interestingly, the xylP::hsfA mutant was slightly more susceptible to fluconazole  . Together, these results confirm our lipidomic, suggesting that HsfA is necessary for plasma membrane homeostasis, which is likely associated with the abnormalities in the balance of lipids with different unsaturation levels that are ultimately raw materials to produce complex lipids such as phospholipids and sphingolipids. Identification of the A. fumigatus SdeA homolog. To better understand the increased accumulation of unsaturated phospholipids during HS in the xylP::hsfA strain ( Fig. 1), we investigated the enzymes that desaturate FA molecules in A. fumigatus. In our RNA-seq analysis, we found that FA desaturases (Afu7g05350 and Afu7g05920; Table 1) were downregulated in response to the hsfA overexpression, indicating the requirement of this HSF in the biosynthesis of unsaturated FA and the balance of saturated/unsaturated lipids. Enzymes that desaturate the C9 of FA are known as D9-fatty acid desaturases, which are highly distributed and conserved from bacteria to mammals. Analysis of A. fumigatus A1163 genome using sequences of S. cerevisiae, Schizosaccharomyces pombe, C. albicans Ole1, A. nidulans, Aspergillus niger, Aspergillus oryzae SdeA, and human stearoyl-CoA desaturases (SCD1 and SCD5) as queries revealed that A. fumigatus A1163 strain possesses two putative D9-fatty acid desaturase-encoding genes: AFUB_091500 and AFUB_090930. Consistent with recently reported data (45), the predicted A. fumigatus SdeA protein encoded by the gene AFUB_091500 shows the highest homology with yeast Ole1, the human SCD1, and aspergilli SdeA (Fig. S3 (45). Moreover, SdeA harbors the conserved histidine-box motifs (HXXXXH and HXXHH) inside the FA desaturase domain, which coordinate iron atoms necessary for catalysis (46). S. cerevisiae Ole1 possesses six lysine residues ubiquitinated for proteasome degradation (25,47), while A. fumigatus SdeA harbors three of these residues (K296, K313, and K379). Interestingly, the human and yeast enzyme sequences have extended N-terminal regions compared to all filamentous fungi sequences. In contrast, fungi uniquely present an extended C-terminal region. Similarly to human SCD1 (46), SdeA has four transmembrane helices that presumably span the endoplasmic reticulum membrane, as determined through the use of DeepTMHMM prediction software ( SdeA is essential and localizes to the endoplasmic reticulum. To characterize the relevance of SdeA to A. fumigatus biology, we attempted to delete sdeA to obtain the DsdeA strain. Despite many attempts, we obtained no positive transformants (data not shown), suggesting that A. fumigatus sdeA is essential for viability. Recently, Wang et al. (45) reported that sdeA in A. fumigatus is an essential gene and demonstrated that sdeA overexpression affected the vegetative growth and decreased the sensitivity to itraconazole.
To further assess the functions of sdeA in the physiology of A. fumigatus and during the HS, we generate the xylP::sdeA conditional mutant to study sdeA loss-of-function phenotypes ( Fig. S4A to C [https://doi.org/10.6084/m9.figshare.22355971]). Similarly to xylP:: hsfA, the expression of sdeA in this mutant is under the control of the xylP promoter, which is repressed by glucose or induced by xylose (36,37). To evaluate whether the xylP promoter function sharply controls the sdeA transcription, the mRNA abundance of sdeA in the xylP::sdeA strain was analyzed by reverse transcription (RT)-qPCR ( Fig. 2A). While glucose drastically represses the sdeA accumulation in the conditional mutant, increasing xylose concentrations induces sdeA expression from 3.5 to 7.0 times. As expected, sdeA expression in the wild-type strain remained constant in all xylose concentrations. To evaluate whether sdeA repression causes any change in fungal growth, the wild-type and xylP::sdeA strains were cultured in solid MM supplemented with various xylose concentrations at different temperatures. The conditional mutant was unable to grow in the absence of xylose at all temperatures tested (Fig. 2B), confirming that the A. fumigatus sdeA gene is essential, similar to S. cerevisiae and C. albicans OLE1 (28,30). A small concentration of xylose (0.25%) was sufficient to induce the growth of the conditional lethal mutant at 37 and 48°C, but not at 30°C. However, the growth of xylP::sdeA at 37°C did not fully recapitulate that of the wild-type strain, even at higher xylose concentrations. Even at 48°C, a temperature at which xylP::sdeA mutant grew at equivalent rates to the wild-type strain at high xylose concentrations, the significant reduction in conidiation was not rescued, as noted by the whitish colony color at both 37 and 48°C (Fig. 2B).
Next, we examined the cellular distribution of SdeA in germlings, and for this purpose, a sdeA::green fluorescent protein (GFP) strain was constructed (  (Fig. 2C). This protein localization pattern is consistent with previous observations of A. fumigatus proteins located in the perinuclear endoplasmic reticulum (48,49). Moreover, it is also possible to observe intense fluorescence signals at the hyphal tips, as well as subapical and basal regions of the hyphae in an elongated network fashion throughout the cytoplasm and septum, suggesting that part of the SdeA::GFP protein presents peripheral endoplasmic reticulum and septum localization resembling that of Cyp51A (48), which was used to confirm the localization of SdeA to the endoplasmic reticulum (ER) compartments (Fig. 2C).
It has been demonstrated that Cyp51A and Cyp51B are redistributed in the cell during the treatment with antifungals or cell wall-damaging agents (48). Given that SdeA and Cyp51A colocalize under basal conditions, we assessed whether the localization of SdeA at the hyphal tip or in the septum would reallocate in the presence of antifungals. We observed that the echinocandin caspofungin (CASP) and the fatty acid synthase inhibitor trans-chalcone significantly caused the most evident reallocation of SdeA (Fig. 3). Quantification demonstrated a significant translocation of SdeA to the hyphal tip of cells treated with CASP, suggesting that monounsaturated FAs are required at the cell membrane to cope with the cell wall damage caused by b-glucan synthase inhibition at the vulnerable sites of apical growth. In the case of trans-chalcone, the ER morphology was dramatically altered, revealing a loss of the ER network associated with the presence of prominent granular fluorescent structures distributed throughout the hyphae, even though septum localization was minimally altered (Fig. 3). It is noteworthy that this abnormal organization of large punctate GFP signals within the cytoplasm is specific to SdeA and not a general response to trans-chalcone since no similar reallocation was observed for the ER-resident protein DapA::GFP challenged by 60 mg/mL of trans-chalcone ( Fig. S6 at https://doi.org/10.6084/m9.figshare.22356781). dapA encodes a cytochrome b 5 -like heme-binding damage resistance protein previously described as involved in ergosterol biosynthesis and azole susceptibility, kindly provided by Ling Liu (Nanjing Normal University, Nanjing, China) (49). In the presence of trans-chalcone, the normal distribution of DapA confirms that the ER structure is not dismantled. These results indicate that SdeA is an essential endoplasmic reticulum-resident protein in A. fumigatus, playing a role in maintaining cell membrane and cell wall organization upon different damaging compounds. Further experimentation is required to investigate how trans-chalcone affects SdeA function.
SdeA is required for unsaturated FA biosynthesis. To investigate the role of SdeA in FA biosynthesis and identify the potential substrates and products of this enzyme, a phenotypic assay for the xylP::sdeA strain was performed using different FA supplementations.
The FA biosynthesis pathway in fungi is shown in Fig. 4A, emphasizing the requirement of the D9-desaturase enzyme in the generation of the D9 monounsaturated FA palmitoleic and oleic acids, which are precursors of polyunsaturated fatty acids (PUFAs). As expected, the repressed xylP::sdeA mutant (glucose 1%; no xylose) could not grow when supplemented by saturated FA such as palmitic and stearic acids as the sole source of lipids FIG 2 SdeA is essential in A. fumigatus and localizes to the endoplasmic reticulum. (A) sdeA expression in the wild-type and xylP::sdeA strains. The strains were grown for 24 h at 37°C in liquid minimal medium (MM) (1% glucose) supplemented with xylose 1% to allow growth and transferred to fresh MM (1% glucose) or MM supplemented with different concentrations of xylose in the absence of glucose for 4 h to induce sdeA expression. The fold increase of each condition represents the normalized values of mRNA in each growth condition relative to the wild-type strain in the control condition (i.e., glucose 1%). Average 6 SD (n = 3) are shown. The bars indicate statistically significant differences (P # 0.05; two-way ANOVA and Sidak's post-test). (B) Radial growth of the wild-type and xylP:: sdeA strains at different temperatures. A total of 1 Â 10 4 conidia of wild-type and xylP::sdeA strains were inoculated into the center of solid MM plates supplemented with the indicated concentrations of xylose and incubated at 30, 37, or 48°C for 72 h. (C) Fluorescence microscopy was performed with the double-labeled sdeA::green fluorescent protein (GFP) cyp51A::monomeric red fluorescent protein (mRFP) strain grown in MM (1% glucose) for 8 h at 37°C. On the right, the merged signal of the GFP, mRFP, and Hoechst staining to visualize the nuclei shows that the SdeA::GFP protein accumulates perinuclearly (white arrows), in the septum (blue arrowheads) or distributed throughout the cytosol or subapically (white arrowheads). Magnification, Â100. Bars, 5 mm.
( Fig. 4B). In contrast, the growth of the conditional lethal mutant was restored in the presence of monounsaturated FAs, palmitoleic acids, and oleic acids under the same growth conditions (glucose 1%; no xylose). These results indicate that xylP::sdeA mutant is auxotrophic for monounsaturated FA since the addition of palmitoleic and oleic acids bypasses the sdeA requirement for the synthesis of D9-monounsaturated FA. It is noteworthy that the conidiation defect observed in this mutant was not fully rescued by the supplementation with such molecules, suggesting that endogenous production of oleic and palmitoleic acid is essential for the A. fumigatus conidiation. The supplementation of PUFAs such as linoleic, linolenic, and arachidonic acids accounted only for poor or no growth of the xylP::sdeA strain (Fig. 4B), reinforcing that the main products of the SdeA enzyme are monounsaturated FA. In addition, higher concentrations of all FA used here inhibited the growth of both strains (data not shown). These results confirm that SdeA is essential for synthesizing monounsaturated FA.
To further determine whether sdeA had broad physiological impacts on A. fumigatus cell membrane homeostasis, we grew the xylP::sdeA lethal conditional mutant in the presence of cell membrane-disturbing agents and different inhibitors of lipid biosynthesis. To allow growth of the xylP::sdeA mutant in this experiment, the xylose concentration of 2.5% was chosen, since radial growth was nearly similar to the wild type (Fig. 2B). Interestingly, xylP::sdeA strain was as sensitive as the wild type to SDS and voriconazole or fluconazole (data not shown). In contrast, the conditional mutant exhibited increased sensitivity to AMB and myriocin (Fig. 4C). Moreover, the FA biosynthesis inhibitors cerulenin and trans-chalcone significantly inhibited the growth of the xylP:: sdeA mutant compared to the wild-type strain (Fig. 4C). These results highlight the  To probe whether the altered balance of saturated/unsaturated FA translated into possible cross talk between SdeA activity and CWI, we investigated the sensitivity of mature A. fumigatus biofilm of the wild-type and xylP::sdeA strains to CASP. Given the involvement of HsfA in the increased accumulation of PC and PS (Fig. 1) and the CWI maintenance (36), the xylP::hsfA strain was also included in this experiment. The 2,3bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide salt (XTT) assay was employed as previously described (19). The xylP::sdeA mutant grown under repressive (glucose 1%, no xylose) conditions exhibited a significant decrease in metabolic activity (55%) when the biofilm was treated with CASP compared to the wild type, regardless of the exposure time (Fig. 4D). Surprisingly, the xylP::sdeA mutant showed an enhanced biomass accumulation on the biofilms compared to the wild-type strain (Fig. 4E), yet the hyphae of such a strain displayed no significant alterations in cell wall thickness when inspected by transmission electron microscopy (data not shown). Notably, the xylP::sdeA strain susceptibility to CASP or other cell wall-disturbing agents (e.g., calcofluor white, Congo red, and caffeine) was equivalent to the wild-type strain in radial growth assays of conidia on agar plates (data not shown). Hence, while sdeA loss of function causes no impact on conidia tolerance to cell wall stress, depletion of sdeA significantly sensitizes A. fumigatus biofilm to CASP treatment. Our results highlight a novel promising synergistic effect of b-glucan inhibition via echinocandin antifungals and depletion of monounsaturated FA in the cell membrane of A. fumigatus biofilm.
Interestingly, despite the documented susceptibility of xylP::hsfA conidia to CASP and the abnormalities in the CWI upon hsfA depletion in fungal biofilms (36), this effect was much lower in the xylP::hsfA strain (35% reduction) and significant only after 30 min of CASP exposure (Fig. 4D). Aiming to understand further the membrane perturbations imposed by the repression of hsfA and sdeA, we measured the ergosterol content in the conditional mutants under control (30°C) and HS conditions as an additional approach to assessing the membrane composition of the mutants. We observed that under control conditions, there is a significant increase in the ergosterol levels in both repressed conditional mutants (2.5-fold) compared to the wild-type strain (Fig. 4F). Such an increase can be an initial explanation for the lack of increased susceptibility to azoles presented by these strains in the phenotypic tests (data not shown and A total of 1 Â 10 4 conidia of the wild-type and xylP::sdeA strains were inoculated in solid MM (1% glucose) supplemented with 2.5% xylose (a concentration at which the radial growth of the mutant was equivalent to the wild-type strain in the absence of stress) and the indicated concentrations of myriocin, amphotericin B, cerulenin, and trans-chalcone. The plates were incubated at 37°C for 72 h. (D) HsfA and SdeA contribute to A. fumigatus caspofungin resistance. The 2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5carboxanilide salt (XTT) assay was used to measure the metabolic activity of mature biofilm of wild-type, xylP::hsfA, and xylP::sdeA strains. The biofilms were obtained by growing each strain for 18 h in MM (1% glucose) supplemented with 1% xylose at 37°C in 24well plates. Biofilms were then washed with MM and incubated in MM (glucose 1%; no xylose) for 4 h at 37°C for xylP repression. Cell wall stress was induced by incubating repressed biofilms with caspofungin (0.5 mg/mL) for 15, 30, or 60 min. The results were expressed as means 6 SD (n = 3). ***, P # 0.001; ****, P # 0.0001 (two-way ANOVA and Sidak's post-test). (E) Biofilm biomass was evaluated by crystal violet absorbance at 570 nm and expressed as the percentage of biomass adhesion considering 100% for the wild-type strain. The biofilms were obtained by growing each strain for 20 h in MM (1% glucose) supplemented with 1% xylose at 37°C. The biofilms were then washed with prewarmed MM and further incubated in MM 1% glucose (no xylose) or MM 1% xylose (no glucose) for 4 h at 37°C for xylP repression or induction, respectively. The results were expressed as means 6 SD (n = 12). ****, P # 0.001 (two-way ANOVA and Sidak's post-test). (F) Ergosterol quantification. The strains were grown in liquid MM (1% glucose) supplemented with xylose 1% for 24 h at 30°C to allow growth. Subsequently, the mycelia were washed twice with MM and incubated for 4 h at 30°C in MM (glucose 1%; no xylose) for xylP repression. Heat shock (HS) was induced by transferring the mycelia to fresh preheated MM (glucose 1%; no xylose) for an additional 15, 30, and 60 min of incubation at 48°C. The control was left at 30°C. Ergosterol was extracted and quantified. The results were expressed as means 6 SD (n = 3). *, P # 0.02; **, P # 0.002 (two-way ANOVA and Sidak's post-test relative to wild-type in the same condition). The bars (#) indicate a statistically significant difference: #, P # 0.03; ###, P # 0.0005 (two-way ANOVA and Sidak's post-test).

A. fumigatus Unsaturated Fatty Acid Metabolism and HsfA
Microbiology Spectrum wild-type strain, the same was not observed in the xylP::hsfA or xylP::sdeA mutants. In fact, significant decreases were recorded after 15 or 30 min of HS in the xylP::hsfA and xylP::sdeA strains, respectively. At least for xylp::sdeA mutant strain, these results highlight that ergosterol content reflects the opposite pattern of accumulation recorded for the wild-type strain, which may be a consequence of the altered ratio of unsaturated to saturated fatty acids. Next, to explore how sdeA affects unsaturated FA and phospholipids in A. fumigatus, we again quantified PC, PE, and PS levels in the repressed xylP::sdeA mutant. To this end, the wild-type and xylP::sdeA strains were grown in liquid MM (1% glucose) supplemented with xylose 1% for 24 h at 30°C. Subsequently, the mycelia were washed twice with MM and incubated for 4 h at 30°C in MM (glucose 1%; no xylose) for xylP repression. HS was induced by transferring the mycelia to fresh preheated MM (48°C) for 15, 30, and 60 min of incubation at 48°C. The control was left at 30°C. Our results demonstrated that the depletion of sdeA did not significantly alter the accumulation of these phospholipids compared to the wild-type strain (Fig. 5A). When the phospholipids species were grouped by the degree of saturation, the results confirmed that the abundance of saturated PC, PE, and PS was significantly higher in the xylP::sdeA mutant compared to the wild-type strain. Interestingly, the higher accumulation of saturated PC and PS in the xylP::sdeA mutant was detected at 30°C and after 15 min of HS. For PE, significant increases were observed later, i.e., after 30 and 60 min of HS (asterisks in Fig. 5B to D). The individual quantification of PC, PE, and PS species in the xylP::sdeA strain demonstrated that the main saturated FAs associated with these phospholipids comprise 16:0 to 16:0, 16:0 to 18:0, 18:0 to 16:0, and 18:0 to 18:0 (purple asterisks in Fig. S1A to C) ([https://doi.org/10.6084/m9.figshare.22354603]). This effect is highly consistent with the function of SdeA in converting saturated FA into monounsaturated FA, indicating that the sdeA depletion causes the accumulation of SdeA substrates and, consequently, saturated phospholipids. Accordingly, the xylP::sdeA mutant synthesized less unsaturated PE with one or two unsaturations than the wild-type strain at 30°C (asterisks, Fig. 5C, middle graph). Nevertheless, the levels of unsaturated PC and PS did not significantly change.
Sphingolipids levels are affected by hsfA but not sdeA depletion. The basic structure of sphingolipids harbors two FA molecules, which present an abundant chemical diversity regarding the length of the carbon chain and the number of unsaturation (50). Therefore, the activity of FA desaturases is essential to generate the pool of the different classes of sphingolipids that will be further incorporated into the plasma membrane.
To determine whether hsfA or sdeA affects the metabolism of sphingolipids and chemically defines how saturated and unsaturated FAs decorating these molecules are being altered by the depletion of these genes, the wild-type, xylP::hsfA, and xylP::sdeA strains were submitted to a sphingolipidomic analysis under the same growth conditions used above. We observed that the HS had no significant impact on the accumulation of total sphingolipids in the wild-type and xylP::sdeA strains (  Although our results support the idea that sdeA repression is not as crucial for sphingolipid biosynthesis as for phospholipid biosynthesis, we observed that sdeA repression led to a higher accumulation of some DHC, PCer, and IPC, compared to the wild-type strain (red asterisks; Fig. S7C [https://doi.org/10.6084/m9.figshare.22357045]). It is noteworthy that all these species have in common saturated FA chains both in the LCB and the amidelinked FA, reinforcing that sdeA repression causes not only accumulation of saturated phospholipids (Fig. 4) but also saturated sphingolipid species (Fig. S7C  hsfA depletion compromises SdeA expression. Given that SdeA activity is tightly coordinated with temperature increase to adjust membrane composition and considering that repression of hsfA caused an increase in mono-and polyunsaturated PC and PS during the HS (Fig. 1), we asked whether the expression of SdeA is influenced by HsfA. We initially assessed the sdeA mRNA levels in the xylP::hsfA mutant under repressive conditions (glucose 1%; no xylose). The sdeA gene expression in the control condition was ;50% lower in the conditional mutant in comparison to the wild-type strain (Fig. 6A), suggesting that HsfA has a positive impact on the basal expression of sdeA. Consistent with the activity of a D9-desaturase, the sdeA mRNA abundance significantly decreased during the HS in the wild-type strain, while the same did not occur in the repressed xylP::hsfA under the HS condition.
Next, SdeA protein levels were evaluated under the depletion of hsfA (glucose 1%; no xylose), by submitting the xylP::hsfA sdeA::3Â hemagglutinin (HA) and sdeA::3ÂHA strains to a Western blot analysis. The construction of these mutants is reported in Fig. S4G  indicating that the introduction of 3ÂHA epitope retains significant function of SdeA protein. Consistent with the mRNA accumulation profile, SdeA protein levels were significantly reduced (40%) in the xylP::hsfA strain pre-HS, again suggesting a positive role of HsfA over SdeA expression at basal (30°C) condition (Fig. 6B). Interestingly, SdeA protein levels were maintained at low and sustained levels throughout the HS in the xylP::hsfA strain, whereas SdeA levels in the wild-type strain dropped significantly (20%, 30%, and 60% after 15, 30, and 60 min, respectively). These data further support a positive role for HsfA over SdeA expression only under basal conditions. We also interpret these data to suggest that hsfA repression leverages the level of the desaturase, thus impairing the natural drop in sdeA mRNA and protein levels during HS, which is clearly observed in the wild-type strain. As a result and guided by the impaired hsfA repression during the HS, the xylp::hsfA strain grown under repressive conditions subjected to HS experiences increased accumulation of unsaturated phospholipids and sphingolipids recorded in Fig. 1  Prior research has hinted at the importance of FA metabolism, HSP activation, and HS response in other human fungal pathogens (34,35). Notably, OLE1 sdeA depletion in C. albicans prevented complete activation of Hsf1 HsfA , which downregulated HSP expression upon HS (24). As hsfA and the HSP hsp90 expression are coordinated during the HS (36), we asked whether SdeA influences Hsp90 protein expression, the main transcriptional target of HsfA (36). We observed that under sdeA depletion (glucose 1%; no xylose), Hsp90 protein levels were increased in the xylP::sdeA mutant compared to the wild-type strain at all time points: 80% under control conditions and 178%, 76%, and 15% after 15, 30, and 60 min of HS, respectively (Fig. 6C). Given its broad chaperone activity, Hsp90 has an ever-growing list of client proteins favored by various physiological or stressful conditions to assist the activity of the target protein. To verify whether SdeA is part of the Hsp90 clientele in vivo, a coimmunoprecipitation (co-IP) assay was performed using the GFP-Trap resin and the sdeA::GFP strain. We discovered that SdeA and Hsp90 physically interacted under basal and HS conditions (Fig. 6D). Altogether, our results suggest that the absence of SdeA is sufficient to induce HS response and disclose a relevant connection between SdeA and Hsp90 in A. fumigatus.

DISCUSSION
Recently, we have shown that HsfA plays a critical role in the HS response and the cell signaling funneling into the CWI pathway to govern cell wall integrity and thermotolerance (19,36). Here, we demonstrate that HsfA negatively regulates the biosynthesis of The sdeA mRNA levels are decreased in the xylP::hsfA strain during hsfA repression. The strains were grown in liquid minimal medium (MM) (1% glucose) supplemented with xylose 1% for 24 h at 30°C to allow growth. Subsequently, the mycelia were washed twice with MM and incubated for 4 h at 30°C in MM (glucose 1%; no xylose) for xylP repression. Heat shock (HS) was induced by transferring the mycelia to fresh preheated MM (glucose 1%; no xylose) for an additional 15, 30, and 60 min of incubation at 48°C. The control was left at 30°C. Total mRNA was evaluated by RT-qPCR and normalized by the b-tubulin gene. The data are from three experimental and independent replicates (average 6 SD). The fold difference for each condition represents the total mRNA normalized to the wild-type control. The bars indicate statistically significant differences: #, P # 0.01 (two-way ANOVA, with Sidak's post-test). *, P # 0.01, a significant comparison between the wildtype and xylP::hsfA strains in the same time point. (B) SdeA protein expression is repressed in the xylP::hsfA genetic background during hsfA repression. Samples were obtained as described above for panel A. a-HA antibody was used to recognize the SdeA::3ÂHA protein in Western blot analysis of sdeA::3ÂHA and xylP::hsfA sdeA::3ÂHA strains during HS. (C) Hsp90 protein expression is induced in the xylP::sdeA genetic background during sdeA repression. The samples were obtained as described above for panel A. a-Hsp90 antibody was used to recognize the Hsp90 protein in Western blot analysis of wild-type and xylP::sdeA strains during HS. (D) SdeA and Hsp90 physically interact in vivo. Wild-type and sdeA::GFP strains were grown in MM (1% glucose) for 24 h at 30°C and subjected to HS as described for panel A. The interaction between the proteins was verified through coimmunoprecipitation using GFP-Trap-agarose resin and a-GFP, which detects the SdeA:: GFP protein, and a-Hsp90 antibodies. For all Western blot analysis, the expression values are arbitrary and calculated by densitometry using the ImageJ software (85). Coomassie blue staining (CBS) was used as the loading control. HA, hemagglutinin.

A. fumigatus Unsaturated Fatty Acid Metabolism and HsfA
Microbiology Spectrum phospholipids, sphingolipids, and unsaturated FAs that shape these molecules. The main biochemical findings supporting these assumptions are the significant increases in the content of PC and PS during HS in the xylP::hsfA mutant. In addition, unsaturated FA present in PC and PS were significantly increased in the mutant, whereas minor changes in the composition of unsaturated FA were observed for the PE. Although PE is the second most abundant phospholipid in mammalian and fungal cells, it is the only class of phospholipids whose concentration significantly dropped in both strains during the HS, suggesting that unlike PC and PS, PE is not regulated by hsfA. While PC is mostly found in the outer leaflet of the cell, PE and PS are located in the inner (cytoplasmic) leaflet of the lipid bilayer (52). PE is particularly enriched in the mitochondrial inner membranes of mammalian cells, where it is produced in situ via decarboxylation of serine moiety of PS (53). It is possible that the ethanolamine supply during the HS is perturbed, hampering the accumulation of PE catalyzed by the CDP-ethanolamine pathway, the only pathway that synthesizes de novo PE in eukaryotes. Curiously, little is known about the origin of the ethanolamine that feeds this pathway in mammals (54), while virtually no information is available in fungi, although PS and PE can act as modulators of virulence in C. albicans (reviewed in reference 55). The addition of double bonds into the hydrocarbon chains of FA is accomplished by desaturases. A. fumigatus possess eight putative FA desaturases previously identified in silico (56). We originally found sdeA by investigating downregulated genes during hsfA overexpression in this study (Table 1) and previously in data sets associated with hsfA depletion and HS series (36). Given the results presented here, as well as others, we demonstrated that SdeA is the closest homolog of S. cerevisiae and C. albicans OLE1 and A. nidulans sdeA (28,30,31,45). Both genetics and lipidomics analysis and the localization of SdeA at the endoplasmic reticulum suggest that sdeA is the main desaturase of palmitic and stearic FA in A. fumigatus. Thus, repression of xylP:: sdeA increases the amount of saturated FA present in PC and PS and some species of sphingolipids, accompanied by a decrease in unsaturated FA enrichment in such molecules. Consistently, A. fumigatus sdeA overexpression leads to the overall accumulation of unsaturated FA, while the opposite was recorded for saturated FA (45).
We found that sdeA is involved in conidiation. Likewise, deletion of A. nidulans sdeA and depletion of C. albicans OLE1, respectively, reduced conidiation or caused defects in the formation of chlamydospores (30,31). Precedent exists indicating the requirement of FA for bona fide conidiation in aspergilli. For instance, in A. niger, oleic acid represents 31.4% of conidial FA (57), while the conversion of oleic to linoleic acid by the oleate D12-desaturase (repressed 3.5 log-fold in our RNA-seq data set; Table 1) also reduced conidiation and mycelial growth in A. nidulans (58).
A most unexpected observation in this study was the significant and specific increase in PE harboring FA with three or more unsaturations during the HS, which presumably hinted at the activity of another D9-fatty acid desaturase (Fig. 5C). Intriguingly, when we searched for other desaturase-encoding genes across Af293 and A1163 genomes using sdeA nucleotide sequences as queries, the open reading frame (ORF) encoded by Afu7g05350/ AFUB_090930, annotated as a putative D9-fatty acid desaturase, returned, respectively, in such strains (73% and 69% identity). Given that this gene was not previously identified in the study by Tang et al. (56), we investigated the AFUB_090930 gene by generating a deletion mutant. The AFUB_090930 null mutant showed equivalent growth as the wild-type and reconstituted strains at 30, 37, and 48°C, as well as in the presence of all cell membrane-perturbing agents used in this study (data not shown). These results suggest that AFUB_090930 does not play a role in the A. fumigatus thermotolerant response or membrane stress in A1163 strain. Further inspections in FungiDB (https://fungidb.org/fungidb/ app) indicated that AFUB_090930 is a pseudogene with a coding sequence. An in-depth analysis of reads mapping of published transcriptomics data sets and our RNA-seq data (File S1 [https://doi.org/10.6084/m9.figshare.22354441]) consistently showed that AFUB_090930 is transcribed. However, there is no evidence that this transcript is translated into an active desaturase that could explain our results. In fact, we failed to produce translational fusions of AFUB_090930 with GFP or 3ÂHA epitope (data not shown). Curiously, the ortholog Afu7g05350 identified in the Af293 strain is annotated as a functional gene. Since Afu7g05350 has not been characterized so far, it is unclear whether it encodes a FA desaturase and whether A. fumigatus expresses two functional D9-fatty desaturases during the HS. In A. nidulans, two redundant D9-desaturases (sdeA and sdeB) affect FA metabolism (31). Why the synthesis of PE with three or more unsaturations increases under depletion of sdeA is an open question. While this may likely be necessary to correct the saturated/unsaturated FA balance in xylP::sdeA mutant, further investigation into the function of Afu7g05350/AFUB_090930 in the Af293 and A163 strains will be essential to understand whether such genes are functional in these two clinical isolates and to define the mechanism of regulation associated with potential genetic heterogeneity.
Disbalance in the content of FA caused by genetic manipulation or exogenous supplementation of FA is known to affect growth and drug tolerance in pathogenic yeasts (23,59,60). Conidia of the repressed xylP::sdeA mutant were significantly susceptible to the inhibition of FA synthesis and to AMB, although not to azoles (data not shown). Interestingly, the exogenous supplementation of oleic acid and the PUFAs linoleic acid and linolenic acid increased the resistance of A. fumigatus conidia to itraconazole, but not to voriconazole or CASP (45). Here, we observed that the lack of CASP susceptibility of the xylP::sdeA conidia was not directly transferable to the biofilm state of such strain challenged with this drug. Instead, the marked reduction in the ability of the xylP::sdeA biofilms to reduce XTT following CASP treatment strongly suggests that depletion of endogenous monounsaturated FA, the concomitant accumulation of saturated FA, and the associated alterations in ergosterol accumulation enhanced the sensitivity to b-glucan inhibition (Fig. 4). Consistently, CASP significantly increased the SdeA apical tip localization, reinforcing previous observations that reallocation of ER proteins such as SdeA, Cyp51A, and Cyp51B at the hyphal tip after exposure to antifungals may represent a general A. fumigatus response to stresses that compromise the apical growth (48). Additional experimentation is needed to reveal the connections between saturated/unsaturated FA balance, ergosterol content, and cell wall integrity perturbation in A. fumigatus biofilm. Aside from some likely alterations in the cell permeability, which may account for different levels of drug uptake, our results point out a potential synergy between echinocandins and inhibition of monounsaturated FA synthesis, which specifically increases the efficacy of this antifungal against the biofilm state found in established infection of most filamentous fungi. Examples of molecules and determinants of virulence that sustain this concept are increasingly relevant in A. fumigatus (61,62). We also demonstrated that inhibiting FA synthase strikingly altered the localization pattern of SdeA, since the granular structures following trans-chalcone treatment are specific to SdeA and significantly large to be considered stress granules observed in aspergilli (63). We observed an increased accumulation of saturated FA upon sdeA depletion, which presumably caused an increase in palmitic and stearic acid content. In A. nidulans, stearic acid accumulation induced by sdeA deletion stimulates fatty acid synthesis via the upregulation of the fasA gene that encodes the a-chain of FA synthase (31). It is possible that the chemical imbibition of FA synthase by transchalcone deteriorates the pool of saturated FA in the sdeA-depleted cells, and one of the consequences is the SdeA reallocation from the ER.
Previous experiments indicated that OLE1 depletion prevents the complete activation and phosphorylation of Hsf1 in C. albicans, which decreases mRNA levels of HSPs such as HSP104 and HSP21, while cells retained constant levels of Hsp90 protein expression (24). Although HsfA activation was not studied in our work (Fig. 7), we discovered that Hsp90 protein levels were significantly induced under depletion of sdeA. Interestingly, neither hsfA nor hsp90 were modulated in the RNA-seq analysis upon overexpression of A. fumigatus sdeA (45). Thus, our findings suggest a different regulatory mechanism observed in C. albicans. It has been found in several models that the HSPs levels can be modulated as a result of changes in the plasma membrane, without the occurrence of HS or protein denaturation (34, 35, 64-67). Different HSPs are associated with the plasma membrane in specific lipid microdomains, which may be a strategy for organisms to compartmentalize HSPs close to other signaling proteins that respond to cell surface receptors (68,69). For instance, a direct strong interaction between Hsp90 and different compositions of lipids was reported. Hsp90 preferentially binds to more unsaturated phospholipid species, and the affinity was higher with negatively charged lipids than with zwitterionic lipids (70). We propose that the predicted accumulation of saturated FA or the deficiency of some unsaturated fatty acid that build up phospholipids and sphingolipids caused by sdeA depletion possibly requires the Hsp90 chaperone activity at higher levels to cope with membrane stress caused by the unbalanced saturated/unsaturated FA ratio in organelles. It is tempting to speculate that the thermophilic nature of A. fumigatus supports this observation; however, the definition of HsfA activation status in the sdeA-depleted cells is crucial to validate this concept. Indeed, we observed that depletion of HsfA decreases SdeA levels at non-HS conditions, suggesting that HsfA is necessary to sustain the production of unsaturated FA. In summary, our results highlight the complex interaction between HS and thermophily in A. fumigatus with effector proteins such as HsfA and SdeA and the consequences to the unsaturated/saturated FA in complex lipids such as phospholipids and sphingolipids.
To induce heat stress in liquid cultures, 1 Â 10 8 conidia of the wild-type, xylP::sdeA, xylP::hsfA, sdeA::3ÂHA, and xylP::hsfA sdeA::3ÂHA strains were incubated in 50 mL of liquid MM (1% glucose) supplemented with xylose 1% for 24 h at 30°C to allow growth as described previously for other xylP conditional mutants (72). Subsequently, the mycelia were washed twice with prewarmed MM and incubated for 4 h at 30°C in MM xylose 1% (no glucose) or MM (1% glucose; no xylose) for induction or repression of the xylP promoter, respectively (36,72). HS was induced by transferring the mycelia to fresh preheated MM (1% glucose) for an additional 15, 30, and 60 min of incubation at 48°C. The control was left at 30°C. Mycelia from each time point, from both pre-and post-HS exposure, were collected via vacuum filtration, flash-frozen in liquid nitrogen, and stored at 280°C until used for either RNA or protein extractions.
Strain construction. All DNA cassettes were generated using a PCR-based strategy and in vivo recombination in yeast (71). For the xylP::sdeA cassette construction, two fragments from the 59-untranslated region (59-UTR) region of sdeA (AFUB_091500) and the sdeA gene were PCR-amplified from genomic DNA of the CEA17 strain according to Fig. S4A (37) and the pyrG gene as a prototrophy marker and was constructed as described previously (36). The cassette was transformed into protoplasts of the A. fumigatus DKU80 pyrG1 according to previously described procedures (71). Transformants were carefully tested by PCR with primers IM-319 and IM-440 ( To construct the double-tagged strain SdeA::GFP Cyp51A::monomeric red fluorescent protein (mRFP), the cyp51A (AFUB_063960) sequence without the stop codon was amplified using primers IM-700 and IM-655 from A1163 strain and cloned in-frame after the gpdA promoter amplified by the primers set IM-702 and IM-699 from the genomic DNA of A. nidulans GR5 strain. The mRFP gene was amplified from the AGB655 strain (kindly provided by Gustavo H. Goldman) (77). Finally, the 3ÂHA prtA cassette was again amplified from pUC 3ÂHA prtA using the primer IM-701 and IM-703 and cloned downstream the mRFP fragment. The cassette was amplified with the outermost primers IM-702 and IM-703 and transformed into the sdeA::GFP strain: The transformants were selected for pyrithiamine resistance.
Biomass quantification in biofilms. The quantification of biomass in the mature biofilms was performed as described by Gravelat et al. (78). In brief, 1 Â 10 5 conidia/mL were inoculated into 200 mL of MM (1% glucose) supplemented with xylose 1% in U-bottomed 96-well plates and allowed to grow for 20 h at 37°C. Next, the biofilm was washed three times with MM and incubated in MM (1% glucose) or MM supplemented with 1% xylose for four h at 37°C for xylP repression and induction, respectively. Following the incubation, the medium was removed, and the adhered mycelia were washed three times with sterile phosphate-buffered saline (PBS). A total of 150 mL of a 0.5% (wt/vol) crystal violet solution was added to each well for 5 min to stain the residual mycelia. Excess stain was gently removed by washing once with sterile water. The residual biofilm was destained with 200 mL of 95% ethanol per well for 16 h at room temperature. The biofilm density was measured by determining the absorbance of the destaining solution at 570 nm. The results were normalized by the fluorescence values of reduced resazurin (Sigma), which was used at a concentration of 10% in a separate and independent experiment to evaluate the growth of the strains and expressed as percentage of biomass. The experiments were performed in 12 technical replicates with at least three biological replicates. Statistical analysis was performed using two-way analysis of variance (ANOVA) with Sidak's post hoc test (P # 0.05).
XTT cell viability assay. The XTT assay was performed as previously reported to measure the metabolic activity of mature A. fumigatus biofilms (79). A total of 1 Â 10 5 conidia/mL of the wild-type, xylP:: sdeA, and xylP::hsfA strains were inoculated into 24-well plates containing 500 mL of MM (1% glucose) supplemented with 1% xylose and incubated at 37°C for 18 h. After growth, the culture medium was removed, and the wells were washed three times with MM. The biofilm formed was then incubated with MM for additional 4 h at 37°C. The control plates were kept at this condition, while cell wall stress was induced by adding 0.5 mg/mL of CASP for 15, 30, or 60 min. The culture medium was removed, and 300 mL of a 0.5 mg/mL XTT solution containing 50 mg/mL of vitamin K were added. The plates were left at 37°C for 1 h, and the absorbance of the resulting reduced XTT was read at 490 nm using the SpectraMax i3 microplate reader (Molecular Devices). The experiments were performed with three technical triplicates, and the data presented represent three biological replicates. Calculations were performed based on the basal values obtained from untreated biofilms for each strain. Statistical analysis was performed using two-way ANOVA with Sidak's post hoc test (P # 0.05).
Fluorescent microscopy. Conidia of sdeA::GFP cyp51A::mRFP strain were inoculated in glass-bottomed dishes (MatTek Corporation) containing 2 mL of MM and incubated for 8 h at 37°C. Filter sets 38HE and 63HE (Carl Zeiss) were used to detect GFP and mRFP, respectively. Germlings were analyzed under a 100Â magnification oil immersion objective (NA 1.4). The nuclei were visualized by staining with Hoechst (20 mg/mL). Images were captured with an AxioCam MRm camera coupled to a Zeiss Observer D.1 microscope and processed using ZEN software. For the SdeA::GFP relocation assays in the presence of antifungals, confocal laser scanning microscopy was performed using a Carl Zeiss LSM 800 confocal microscope with a Plan Apochromat 63Â/1.40 Oil objective. The detection parameters for SdeA::GFP experiments were fixed as follows: laser, 488 nm; 35% 1.89 AU/84 mm detection wavelength, 488 to 574 nm; and detection gain, 850 V. The z-stack increments were 0.3 mm. The images were analyzed using ZEN Blue 2.3.
Phenotypic assays. The radial growth of the wild-type and xylP::sdeA strains at different temperatures was analyzed by spotting 1 Â 10 4 conidia of each strain in the center of 90-mm petri dishes containing solid MM (1% glucose) supplemented with various concentrations of xylose (0.25% to 5%). The plates were incubated for 72 h at 30, 37, or 48°C and analyzed. The same procedures were applied to investigate the radial growth of the strains in the presence of different FA. An aqueous solution containing 1% Igepal CA 630 was used for the FA solubilization in the MM, as previously described (30). The plates were incubated for 96 h at 37°C and photographed. Alternatively, 1 Â 10 4 conidia of each strain were inoculated in U-bottomed 96-well plates containing 200 mL of solid MM (1% glucose) supplemented with 2.5% xylose and different concentrations of aureobasidin A, cerulenin, trans-chalcone, myriocin, lovastatin, voriconazole, fluconazole, and AMB. The same procedures were followed for tests with the xylP::hsfA strain but using 0.06% xylose. Such xylose concentration allows the growth of the conditional lethal mutants and can evidence phenotypic differences among the strains grown in repressive conditions (1% glucose) in the presence of different stressing agents. All of the 96-well plates were incubated for 72 h at 37°C and analyzed.
RNA extraction and RT-qPCR. Flash-frozen mycelia from the HS cultures were disrupted by grinding in liquid nitrogen. The total RNA was extracted with TRIzol reagent (Thermo Scientific) according to the manufacturer's protocol. RNA was purified, quantified, treated with Turbo DNase I (Thermo Scientific), and reversetranscribed with a high-capacity cDNA reverse transcription kit (Thermo Scientific) as described previously (80). Quantitative reverse transcription PCR (RT-qPCR) was conducted with Power Sybr green PCR Master Mix (Thermo Scientific). The primers for the individual genes were designed using Primer Express 3.0 software (Life Technologies) and are listed in Table S3 (https://doi.org/10.6084/m9.figshare.22354573). RT-qPCR was performed in duplicate from three independent biological samples in a StepOne Plus real-time PCR system (Thermo Scientific). The fold change in mRNA abundance was calculated using 2 2DDCt (81), and all the values were normalized to the expression of the A. fumigatus b-tubulin (tubA). Statistical analysis was performed using two-way ANOVA with Sidak's post hoc test (P # 0.05).
RNA sequencing. To induce hsfA overexpression, 1 Â 10 8 conidia from wild-type and xylP::hsfA strains were incubated in 50 mL of liquid MM (1% glucose) supplemented with 1% xylose for 24 h at 30°C to allow growth. Subsequently, the mycelia were washed twice with MM 1% xylose and incubated for 4 h at 30°C in the same medium (xylose 1%; no glucose) for hsfA overexpression (36). Mycelium from each strain was collected via vacuum filtration, frozen in liquid nitrogen, and stored at 280°C. Total RNA was extracted using TRIzol reagent, treated with DNase I (Qiagen), and purified using the RNAeasy kit (Qiagen), according to the manufacturer's instructions. Sample preparation, library construction, and data analysis were performed as described previously (36). Short reads were submitted to the NCBI's Short Read Archive under Bioproject PRJNA690780. Gene ontology (GO) enrichment analysis was performed using the KOBAS tool (kobas.cbi.pku.edu.cn) (82).
Protein extraction, immunoblotting, and co-IP procedures. To assess the SdeA::3ÂHA expression, the mycelia obtained upon HS, according to the description above, were disrupted by grinding them in liquid nitrogen. Protein extraction and Western blotting were performed as before (80). The detection of SdeA::3ÂHA protein was achieved by using a-HA (H3663; Sigma) antibody according to the manufacturer's instructions, as previously reported (83). To perform the co-IP assay, the sdeA::GFP strain and the GFP-Trap agarose resin (ChromoTek; GTA-20) were used as described elsewhere (84). The detection of SdeA::GFP protein was achieved by using a-GFP antibody (sc-9996; Santa Cruz Biotechnology). A. fumigatus Hsp90 was detected using a custom polyclonal a-Hsp90 antibody raised in rabbit (19). For a-HA and a-GFP detection, a-mouse IgG horseradish peroxidase (HRP) antibody (A4416; Sigma) was used, while a-rabbit IgG-HRP (Sigma; A0545) was the secondary antibody for a-Hsp90 detection. Chemoluminescent detection was performed by using an ECL Prime Western Blot detection kit (Cytiva). The images were generated by exposing the polyvinylidene difluoride (PVDF) membranes to the ChemiDoc XRS gel imaging system (Bio-Rad). The images were subjected to densitometric analysis in ImageJ software (85).
Phospholipid and sphingolipid extraction. For sphingolipid extraction, 300 mg of mycelia obtained upon HS according to the description above were suspended in Mandala buffer (86). Lipid extraction was carried out as described previously (87) with a few modifications. To disrupt the mycelia, the samples were vortexed and sonicated for 2 min in the presence of 0.2 mL of glass beads. The supernatant was collected, dried, and submitted to Bligh and Dyer extraction (88). One-third of each sample obtained after Bligh and Dyer extraction was reserved for inorganic phosphate (P i ) determination, while the remaining was subjected to the alkaline hydrolysis of phospholipids (89). The same protocol was followed for phospholipid extraction but without the alkaline hydrolysis step.
Sphingolipid mass spectrometry analysis. Sphingolipid mass spectrometry analysis was carried out as previously reported (90). Briefly, a Thermo Accela high-performance liquid chromatography (HPLC) system (San Jose, CA) was used to separate the dried extracts dissolved in 150 mL of ammonium formate (1 mM) with 0.2% of formic acid in methanol. A Peeke Scientific Spectra C8 (Redwood City, CA) HPLC column (150 Â 3 mm) was used, into which 10 mL of samples were injected. The HPLC was coupled to the HESI source of a Thermo TSQ Quantum Ultra triple quadrupole mass spectrometer (San Jose, CA). The sphingolipid profile was performed using positive ion mode, with the high voltage set to 3.5 kV, vaporizer temperature at 400°C, sheath gas pressure at 60, auxiliary gas pressure at 15, and a capillary temperature of 300°C. The collision cell was operated at 1.5 mTorr of argon. For the duration of the run, transitions for each lipid species were monitored at 100-or 50-ms dwell time. A total of 20 lipid standards for our profile from Avanti (Alabaster, AL) were used to develop calibration curves, and these curves were then used for lipid species to be monitored. Processing of the samples was done using Thermo Xcalibur 2.2 Quan browser software and exported to Excel for reporting results. Sphingolipid concentration determined by mass spectrometry was further normalized by the P i abundance (91).
Phospholipid mass spectrometry analysis. Phospholipid mass spectrometry analysis was carried out as previously mentioned (92). The dried residue was reconstituted in 0.5 mL of the starting mobile phase solvent for LC-MS/MS analysis. Phospholipid classes were separated by reverse-phase LC using a Supelco 2.1 (inner diameter) Â 150 mm Ascends Express C18 column (Sigma, St. Louis, MO) and a binary solvent system at a flow rate of 0.4 mL/min with a column oven set to 45°C. PC, PE, and PS were each quantified during each run by MRM analysis, and in addition, the structure was confirmed via a MS2 scan within the linear ion trap during the elution of each peak. For liquid chromatography-tandem mass spectrometry (LC-MS/MS) analyses, a Shimadzu Nexera LC-30 CE binary pump system coupled to a SIL-30 AC autoinjector and DGU20A5R degasser coupled to an AB Sciex 5500 quadrupole/linear ion trap (QTrap) (SCIEX Framingham, MA) operating in a triple quadrupole mode was used. Quarters 1 and 3 were set to pass molecularly distinctive precursor and product ions (or a scan across multiple m/z in quarter 1 or 3), using N 2 to collisionally induce dissociations in quarter 2 (which was offset from quarter 1 by 30 to 120 eV); the ion source temperature was set to 500°C. The internal standards were l,2-diheptadecanoyl-sn-glycero-3-phosphocholine (17:0/ 17:0 PC), l,2-diheptadecanoyl-sn-glycero-3-phosphoethanolamine (17:0/17:0 PE) and l,2-diheptadecanoyl-sn-glycero-3-phospho-L-serine (17:0/17:0 PS). Phospholipids concentration determined by mass spectrometry was further normalized by the P i abundance (91).
Ergosterol extraction and quantification. For sterol extraction, the same protocol used for sphingolipid extraction described above was followed, including the alkaline hydrolysis and the P i estimation but without the addition of internal standards. Quantification was performed by thin layer chromatography (TLC). The ergosterol standard (Cayman Chemical) and the dried lipid samples were resuspended in 50 mL of chloroform/methanol (2:1 vol/vol), and 10 mL of each sample were spotted onto a silica gel plate (EMD Millipore) and run in a 10-foot Â 10-foot glass TLC tank containing chloroform/methanol/ water (65:25:4, vol/vol/vol) as a mobile phase prepared the day before. After 90 min of separation, the plates were dried at room temperature and developed with iodine crystals. Ergosterol was visualized by UV light (254 nm) for image acquisition. The images were subjected to densitometric analysis using the ImageJ software (85) and normalized to the P i quantification.
Data availability. All the data herein described are included in the article.