Aurones SH1009 and SH9051 possess an identical core of their chemical structure with different functional groups, but they exhibited different half maximal inhibitory concentration (IC₅₀) values against C. albicans (Figures 1A,B). Therefore, the potential of any synergistic interaction of their combination was investigated using a classical checkerboard microdilution assay. The fractional inhibitory concentration index (FICI) value between SH1009 and SH9051 was observed to be an indifferent effect (FICI = 1.25) against C. albicans SC5314 (Figure 1C). The growth inhibition by both compounds in combination was more than 90% at the concentrations of 25 μM for SH009 and 100 μM for SH9051, which is in the zone of indifference, suggesting different modes of toxicity (Figure 1D).

Cytotoxic effects of aurone SH1009 in mammalian cells were reported previously, indicating a selective antifungal toxicity toward the pathogenic C. albicans SC5314 strain (Alqahtani et al., 2019). The cytotoxic effects of aurone SH9051 on the same human cell lines (THP-1, HepG2, and A549) was carried out to examine the selectivity toward fungal cells. Supplementary Table 1 shows the cytotoxic concentrations of SH9051 that caused 50% loss of cell viability for each cell line, with SH9051 showing high CC₅₀ values for each cell line. Since the IC₅₀ of aurone SH9051 against C. albicans was 91.05 μM, low selective index values were obtained for THP-1 and HepG2 cells, but the difference between the CC₅₀ of A549 cells and IC₅₀ of fungal cells was two-fold. Conversely, aurone SH1009 demonstrated highly selective toxicity toward C. albicans cells. The concentration of 25 μM of SH1009 caused significant inhibition of C. albicans growth (>90% inhibition) while the cell viabilities for the human cells remained high at this concentration (Alqahtani et al., 2019). Here, a concentration of 100 μM of SH9051 showed the same inhibitory effects on C. albicans as 25 μM of SH1009, although this concentration of SH9051 decreased cell viabilities for the human cell lines (Supplementary Figure 1). These disparities in inhibitory concentrations and the indifference in the checkerboard results suggest that aurones SH1009 and SH9051 could target different biological pathways. Therefore, genome-wide gene expression analyses of C. albicans cells in response to aurone treatments were performed.

The effect of aurone treatments on C. albicans SC5314 transcriptional responses was investigated after 1 h incubation with concentrations that caused 30% inhibition. A high quality of total RNAs (≥1.5 μg and an A260/A280 ratio ≥ 2.0) with high RNA integrity number (RIN ≥ 9.7) (Supplementary File 1) was recovered for SH1009-treated, SH9051-treated and untreated cells. Genome-wide transcriptional analysis was performed using transcriptome sequencing (RNA-seq). The pre-analysis of RNA-seq computational analyses is a major step to detect any technical biases that may alter the biological differences. Therefore, diagnostic plots showed high-quality RNA-seq data regarding the alignment accuracy, sequencing depth, and filtered and normalized counts distribution (Supplementary Figure 2). The reproducibility among the conditional groups and within technical replicates demonstrated different gene expression profiles with no impact of batch effects. Notably, the two-dimensional plot of principal component analysis showed a distinct separation with greater variability between experimental groups by the first component, indicating that the treatment with aurones was the major source of differences (Figure 2A). In addition, the variability of gene expression profiles of SH1009-treated and SH9051-treated cells was increased to 86% variance when comparing these profiles to each other only, suggesting different modes of action for different aurones and supporting the checkerboard assay results (Figure 2B). The hierarchical clustering with pairwise correlations showed similar clustering that identified three groups of samples based on the treatment with dramatically different patterns of transcriptional profiles (Figure 2C). Treatment with SH1009 and SH9051 resulted in 1249 and 1085 differentially expressed genes between treated and untreated samples (P (FDR) ≤ 0.05 and fold change ≥ 1.5), respectively. A total of 585 genes were up-regulated, and 664 genes were down-regulated upon SH1009 treatment, whereas 689 genes were up-regulated, and 396 genes were down-regulated upon SH9051 treatment (Supplementary Tables 2, 3) compared to their regulation in C. albicans under untreated conditions (Figures 2D,E).

To elucidate the transcriptional response of C. albicans SC5314 during aurone treatment, the induced and repressed genes were examined for clustered analysis of over-represented functionally annotated KEGG pathways and gene ontology (GO) terms in SH1009 and SH9051 transcriptomes. GO enrichment analyses showed that a large subset of down-regulated genes were significantly (P (FDR) values < 0.0001) enriched in response to cellular transporters and regulators associated with metabolic processes (carbon metabolism for SH1009 and sulfur metabolism for SH9051), suggesting that the exposure to chemically different structures of aurones robustly alters the nutrient availability in which the expression of genes that are involved in cellular metabolism and nutrient uptake are distinctively repressed. However, for up-regulated genes that were responsive to aurone treatment, GO enrichment analysis showed (P (FDR) < 0.0001) genes that are associated with RNA processing and ribosomal biogenesis in both aurone profiles in addition to uniquely induced responses for each aurone profile (iron ion homeostasis for SH1009 and arginine biosynthesis for SH9051), suggesting common aspects of detoxification mechanisms for C. albicans with specifically induced routes of adaptations in order to buffer the effects of particular nutrient deprivations (Figure 3).

To obtain a clearer view of the sequential changes in gene expression patterns of C. albicans cells after SH1009 or SH9051 treatment, a gene regulatory network for the targeted pathway of each transcriptome profile is summarized in Figure 4A for SH1009 and Figure 4B for SH9051. As aforementioned, SH1009 caused a repression in the carbon metabolism pathway which is tightly regulated by the Ras1/cAMP/protein kinase A (PKA) pathway (Sabina and Brown, 2009). The glucose G-protein receptor gene GPR1, for which expression was down-regulated by −1.5-fold, controls the production of cAMP from ATP by activating the down-regulated GTPase gene RAS1 (by −1.48-fold). In addition to Ras1, other key components of the Ras1/cAMP/PKA pathway were differentially expressed by the treatment. The up-regulated genes GTPase RAS2 (by 2.2-fold) and phosphodiesterase PDE2 (by 2.6-fold) primarily reduce the cellular level of cAMP, which is required for inactivation of the inhibitory subunit Bcy1 of PKA. The PDE1 gene, which was down-regulated (by −1.49-fold), acts as the substrate for the PKA complex and provides negative feedback in carbohydrate signaling and regulation. When cAMP binds to the inhibitory subunit of the PKA complex, it releases its catalytic subunits, TPK1, which was up-regulated by 1.6-fold and is involved in stress response, and TPK2, which was down-regulated by −1.56-fold and involved in trehalose synthesis and iron uptake (Robertson et al., 2000; Lin and Chen, 2018). Activated PKA phosphorylates a series of transcription factors that control carbon metabolism, stress adaptation, cell cycle, and other functions, but the terminal point of the Ras1/cAMP-PKA pathway in C. albicans activation is the master transcription factor Efg1p. This transcription factor, for which expression was down-regulated by −2.4-fold, promotes yeast-to-hyphal transition through controlling carbon metabolism by directly regulating expression of the major glycolytic activator Tye7p that was down-regulated by −2.2-fold as well (Robertson et al., 2000).

The transcriptional factors in the down-regulated genes set for the SH1009-treated transcriptome were ranked based on the percentage of transcriptome regulation using the PathoYeastract database (Monteiro et al., 2017). The transcriptional regulator Tye7p was the top transcriptional factor that plays a key role in cellular response to SH1009 by regulating 44% of down-regulated genes. Tye7p regulates the down-regulated genes TPS1 (by −2.3-fold), TPS2 (by −2.5-fold), TSP3 (by −2.4-fold), ATC1 (by −1.5-fold), and NTH1 (by −1.5-fold), all of which are involved in the most significantly enriched biological process in the SH1009 repressed transcriptome, trehalose metabolism (Askew et al., 2009). Moreover, genes that are associated with glucose, glycogen, and galactose biosynthesis were found dysregulated by SH1009 treatment, consequently resulting in a poor supply of carbohydrate, which is the preferred source of energy in C. albicans for glycolysis and the tricarboxylic acid (TCA) cycle. The dysregulation of the TCA cycle attenuates the production of ATP, leading to a low intracellular levels of cAMP, which dramatically affects the activation of the Ras1/cAMP-PKA pathway (Tao et al., 2017), supporting our previous results that suggested the modulatory effect of SH1009 on the ATP-binding cassette (ABC) efflux pumps which drive the fluconazole resistance mechanism (Alqahtani et al., 2019). Significantly, the top down-regulated gene in the SH1009 transcriptomic profile (by −32-fold), HSP21, lies downstream of the Ras1/cAMP-PKA pathway (Gong et al., 2017). The deletion of this small-heat shock protein, Hsp21, weakens C. albicans resistance to oxidative and thermal stresses by accumulating significantly lower intracellular levels of trehalose sugar which is considered an oxidative stress protectant (Mayer et al., 2012). The involvement of Hsp21 in the regulation of trehalose homeostasis and the tight control of trehalose level by G protein receptor (Gpr1) suggest that Hsp21 is a possible link between the Ras1/cAMP-PKA pathway and trehalose biosynthesis (Serneels et al., 2012). Additionally, SH1009 down-regulates the basic carbon metabolism that controls trehalose biosynthesis, the key nutritional and protectant cue for carbohydrate regulation and oxidative resistance in C. albicans.

Candida albicans cells exhibited a different repressed transcriptomic profile in response to SH9051 (Figure 4B). Organic sulfate compounds may be transported into the cells by the putative sulfate transporter (Seo1) for which expression was down-regulated by −2.06-fold. Sulfate, then, should be reduced through the sulfate assimilation pathway genes, MET3, MET14, MET16, ECM17 (or MET10), which were among the top down-regulated genes (by −10, −11, −12, and −1.7-fold, respectively), to sulfide which would afterward be incorporated into the formation of homocysteine by down-regulated gene MET15 (by −1.9-fold) along with O-acetylhomoserine derived from homoserine by the MET2 gene (down-regulated by −1.7-fold). In addition, expression of genes involved in sulfur amino acid (methionine and cysteine) biosynthesis pathways were down-regulated. Homocysteine is converted to methionine by MET6 (down-regulated by −2-fold) and to S-adenosylmethionine (SAM) in the methyl cycle by SAM2, SAM4, and SAH1 (down-regulated by −1.5, −2, and −1.6-fold, respectively). CYS1 and CYS4, which convert homocysteine to cysteine in the two steps of the transsulfuration process, were down-regulated by −2 and −1.9-fold, respectively. In S. cerevisiae, the levels of methionine, cysteine, and SAM negatively control the activity of transcription factor Met4p (expression down-regulated by −3-fold) that controls the activity of sulfur metabolism pathway by positively regulating the activity of MET genes (Ljungdahl and Daignan-Fornier, 2012). Shrivastava et al. (2021) have also recently confirmed Met4 as a regulator of methionine circuitry in C. albicans, observing up-regulation of many methionine biosynthesis genes, including the sulfur assimilation pathway and the SAM cycle. The findings in this present study indicate that SH9051 down-regulates the sulfur metabolism pathway which dysregulates the methionine and cysteine biosynthesis processes. Taken together, both repressed transcriptomic profiles of SH1009 and SH9051 demonstrated significantly and distinctively major perturbations of different metabolic pathways in C. albicans cells, suggesting different molecular targets.

To detect any potentially synergistic or antagonistic interaction between SH1009 and SH9051 at the molecular level, a Venn intersections diagram was constructed to compare the repressed and induced transcriptomic profiles (Figure 5). Although ∼197 genes were commonly down-regulated in both repressed datasets of SH1009 and SH9051, 44% of these genes were enriched for unknown biological processes while only 11 genes were enriched insignificantly (P (FDR) value < 0.07) for carbohydrate metabolic processes, especially genes that are involved in drug stress response, indicating no synergistic interaction for repressed transcriptomic responses. Comparing the up-regulated genes for SH9051 and the down-regulated genes for SH1009 detected 53 commonly dysregulated genes that are not enriched for any GO term. By way of contrast, a low number of overlapped genes (only 14) for iron ion transport and sulfur amino acid metabolic processes are significantly enriched between the up-regulated genes for SH1009 and the down-regulated genes for SH9051. The up-regulated genes for sulfur amino acid metabolic processes (MET4, MET3, MET15, and CYS1) in the induced transcriptome of SH1009 could be attributed to carbohydrate starvation, which subsequently causes the cells to utilize amino acids as alternative sources of carbon (Miramón and Lorenz, 2017). The down-regulation of genes for iron ion transport (RBT5, FET34, FTR1, and CFL4) in the repressed transcriptome of SH9051-treated cells could be attributed to elevated levels of organic sulfur compounds as a consequence of sulfur amino acid catabolism. This might be concomitant with repression in iron ion import in order to hinder the synthesis of iron–sulfur (Fe-S) proteins (Petti et al., 2012). Nevertheless, this low level of antagonistically differential expression of a small number of genes cannot be interpreted as a fundamentally antagonistic interaction between the aurone modes of action because these antagonistic processes seemed to be secondary effects of the essentially targeted pathways.

However, the induced transcriptome in both SH1009 and SH9051 profiles was remarkably comparable. A large set of genes (∼299) were commonly up-regulated upon SH1009 and SH9051 treatments. These genes were significantly enriched (P (FDR) values < 0.0001) for RNA processing, ribosomal biogenesis, mitochondrial RNA processing, and cleavages involved in rRNA processing. High transcript levels for rRNA processing and ribosomal biogenesis in S. cerevisiae after exposure to high lethal concentration of H₂O₂ have been observed previously, suggesting an urgent need to accelerate the rRNA processing and ribosomal biogenesis that are required for synthesizing oxidative stress response proteins or to replace the rRNA that was damaged by oxidative stress (Shenton et al., 2006). Additionally, nutrient depletion has been associated with ribosomal stress by inducing ribosomal degradation and formation of free ribosomal proteins that can eventually accumulate in the nucleoplasm (Zhou et al., 2015).

In addition to commonly induced genes in response to either SH1009 or SH9051 treatments, uniquely up-regulated genes in response to each aurone have been significantly enriched for prominent biological processes. For SH1009, a set of genes was uniquely induced and enriched for iron ion transport. Most of these genes (C4_03340C_A, 11-fold; CFL2, 9-fold; RBT5, 7.9-fold; CFL4, 7.2-fold; FRE9, 4.5-fold; FTR1, 4.5-fold; FTH1, 3.6-fold; CFL5, 3.4-fold; SIT1, 3-fold; FRE10, 3-fold; FRP1, 2.8-fold; FRE30, 2.7-fold; C7_00430_A, 2.6-fold; FRE7, 2.4-fold; FET34, 1.9-fold; CTR1, 1.67-fold; and CFL1, 1.5-fold) were among the top up-regulated genes. The expression of iron ion acquisition genes in C. albicans is regulated by the induced transcriptional repressor Hap43p (two-fold), which responds to low intracellular levels of iron by inhibiting the transcriptional activator of iron utilization, Sfu1p, which was down-regulated by −1.86-fold. The Sfu1p transcription factor represses genes encoding iron uptake through direct inhibition of a third transcription factor Sef1p, for which expression was up-regulated three-fold, and indirect inhibition of Hap43p expression, since Sef1p regulates the activity of Hap43p (Chen et al., 2011). This strict regulation of iron acquisition and utilization reflects the significance of maintaining ordinary intracellular iron homeostasis to thereby prevent the iron toxicity that leads to severe oxidative stress. Overabundance of iron catalyzes the production of reactive oxygen species (ROS) by the Fenton reaction, resulting in oxidative damage in biomolecules (Missall et al., 2004). A recent study has reported that iron overload is the major determinant of the oxidative stress associated with rRNA cleavage and degradation through promoting stress-strand breaks in rRNA by ribosome-bound iron (Zinskie et al., 2018).

For SH9051, a set of genes that are involved in the arginine biosynthetic process was uniquely induced (DUR1,2, 8.4-fold; GDH1, 5.9-fold; AFP99, 3.6-fold; ARG11, 2-fold; ARG8, 2-fold; ARG1, 1.89-fold; ECM42, 1.9-fold; CAR1, 1.75-fold; CAP2, 1.65-fold; ARG5, 1.6-fold; and CAP1, 1.56-fold) in response to the aurone. The induction of the arginine biosynthetic pathway has been reported previously as a transcriptional response in C. albicans to oxidative stress resistance (Issi et al., 2017). Additionally, as a defense mechanism against host-phagocytosed cells, C. albicans induces the arginine biosynthetic pathway to resist macrophage-derived antimicrobial ROS (Jiménez-López et al., 2013). These findings indicate that both aurones could induce cellular oxidative stress; however, uniquely up-regulated mechanisms also suggest key roles for the functional groups in the chemical structure of each aurone.

By linking the transcriptional responses of aurones SH1009 and SH9051 to their core chemical structures and functional groups, three critical hypotheses can be proposed. First, the hydroxyl groups and/or methoxy group in aurone SH1009 can react with transition metal ions (iron), which produces high levels of toxic ROS that would result in rRNA cleavage and degradation along with ribosome stress, resulting in a potent pro-oxidant activity. Additionally, these functional groups could uniquely affect the accumulation level of the oxidative protectant, trehalose (Gencer et al., 2018), which in turn down-regulates the Ras1/cAMP-PKA pathway and negatively impacts carbon metabolism and its regulation (Perfect et al., 2017; Figure 4A). Second, since C. albicans utilizes amino acids as a nitrogen source, the nitrogen atom in aurone SH9051 could interfere with the negative regulation of transcription factor Met4p, simulating a high level of methionine and cysteine supply resulting in inactivation of MET genes, thus suppressing the sulfate assimilation pathway (Hébert et al., 2013) and inducing sulfur amino acid catabolism (Figure 4B). Third, the C ring structure of both aurones could contribute to generating more ROS, resulting in the common up-regulation processes (rRNA processing and ribosomal biogenesis). This would provide SH9051 a mild pro-oxidant activity, but it is lower than that of SH1009 which can be observed from the depth of the red color of the heatmap in the SH1009 gene expression profile compared to that for SH9051 (Figure 5). Therefore, a reverse genetics approach and quantitative biochemical assays were applied to validate the transcriptional signatures of aurones SH1009 and SH9051 and to establish the structure-activity relationship.

The enrichment analysis of differentially regulated genes upon aurone treatment indicated carbon and sulfur metabolism as the most significantly down-regulated pathways. Accordingly, engineering of C. albicans deletion strains was attempted using the CRISPR-Cas9 gene editing technique in order to validate that the trehalose and sulfur amino acid biosynthetic metabolic pathways are the primary targets of aurones SH1009 and SH9051, respectively (Figures 6A,B). We reasoned that if the transcriptional factors of carbon metabolism, Tye7p, and sulfur metabolism, Met4p, are absent, then the aurones would not be active against the mutants. Supplementary Figure 3A depicts the genetic modification of a TYE7Δ mutant from a C. albicans SC5314 strain background by introducing double-strand breaks in which two gRNAs guide Cas9 nuclease to catalyze the cleavages at the 5′ and 3′ ends of the open reading frame (ORF) of the targeted gene TYE7 and replace the gene sequence with the gene drive cassette (Halder et al., 2019). Although the gene drive system harbors two single gRNA sequences and SNR52 RNA polymerase III promoter, allowing the gene drive to transcript and propagate to any wild-type locus for deletion, the platform failed to generate a MET4Δ mutant even with multiple rounds of transformations using three different gene drive systems containing three different dual gRNAs which target different sites on the MET4 gene sequence. Therefore, a MET4Δ deletion mutant was obtained from the Fungal Genetics Stock Center (FGSC) (University of Missouri, Kansas City, MO, United States), and the absence of the MET4 gene sequence was verified by colony PCR (Supplementary Figure 3B).

The TYE7Δ mutant was documented previously for higher storage carbohydrate levels (trehalose and glycogen) as a result of severely reduced glycolytic flux due to the repressed expression of the genes encoding the glycolytic-committing enzyme phosphofructokinase (PFK1 and PFK2). This gives Tye7p the role to regulate the flux between carbohydrate storage and production at the glucose-6-phosphate branch point in order to avoid futile glycolysis cycling (Figure 4A; Askew et al., 2009). Consequently, while the growth of C. albicans SC5314 was inhibited by ∼4-fold during the exponential phase of the growth curve under the IC₅₀ concentration of SH1009 (16 μM), the TYE7Δ mutant showed evident SH1009-resistant growth (Figure 6C), suggesting that the greater storage of oxidative stress protectant (trehalose) could rescue TYE7Δ from the proposed pro-oxidant activity of SH1009. Additionally, it has been reported that exposing C. albicans to ∼30 μM extracellular iron led to accumulation of more ROS, which could cause a synergistically inhibitory activity with SH1009 (Kaba et al., 2013). However, supplying growth media with different concentrations of extracellular iron did not increase the inhibitory activity of SH1009 (Supplementary Figures 4A–C). In fact, it marginally reversed the inhibitory activity which could support the claim that the uniquely up-regulated genes for iron ion uptake signifies the low intracellular level of iron ions upon SH1009 treatment. Because C. albicans possess a sophisticated system of regulation of iron ion transport due to the potential toxicity of iron by generating ROS, it could be speculated that the impact of SH1009 on the level of iron ions was endogenous and it cannot be affected exogenously.

Treatment of the MET41 mutant with SH9051 resulted in a more susceptible response compared to the wild type indicating a potential secondary effect or pleiotropic effects of the mutation (Figure 6D). However, unlike the recent findings of methionine auxotrophic phenotype in C. albicans MET4Δ (Shrivastava et al., 2021), our C. albicans MET4Δ strain is not auxotrophic for sulfur amino acids (methionine or cysteine) (Supplementary Figure 4D) which could be attributed to different genetic backgrounds. Also, the additional supplements of cysteine or methionine were not able to restore the fitness of MET4Δ without any stress. In addition to the SH9051 stress, addition of sulfur amino acids further inhibited the growth of MET4Δ (Figures 6E,F). This could support the hypothesis from RNA-seq data where SH9051 treatment stimulated a toxic supply of sulfur amino acids, suggesting a synergistically toxic effect between the SH9051 and these amino acids through sulfur amino acid catabolism. For this reason, most of the methionine and cysteine taken up by MET4Δ cells might be degraded thus reducing SH9051 toxicity even in the lack of de novo biosynthesis of these amino acids. Therefore, in order to further investigate the cellular responses of the wild type and relevant mutants during aurone SH1009 and SH9051 treatments, quantitative biochemical assays were conducted.

Comparisons between the repressed and induced transcriptomic profiles of C. albicans cells treated with aurones SH1009 and SH9051 indicated that a common feature in the aurone chemical structures could contribute to elevating intracellular ROS levels. Additionally, distinct functional groups in each aurone indicated uniquely up-regulated and down-regulated pathways (Figure 5). Accordingly, the core chemical structure of an aurone without any functional group (phenyl aurone, Figure 7A) was included in cellular metabolism assays to quantitively measure the structure-activity relationships of the aurone core and functional groups. The cell viability assay for the phenyl aurone showed that inhibition of C. albicans SC5314 at the highest concentration of 200 μM was approximately 28%, suggesting that approximately one third of any aurone inhibitory activity was proportionally derived from this basic chemical structure (Figure 7A).

Since the core chemical structure was the only common feature between aurones SH1009 and SH9051, we hypothesized that the C ring structures, comprised of the oxygen double bound, oxygen atom, and/or the extended conjugation that connects the benzofuranone heterocyclic ring with the phenyl group (Figure 7A), could contribute to generating ROS. Therefore, the exponential growth of C. albicans SC5314 wild-type, TYE7Δ, and MET4Δ cells was perturbed by 200 μM of each aurone for 1 h, after which yeast cells were stained with the oxidative stress indicator, CM-H₂DCFDA, that is oxidized by ROS to a fluorescent molecule (Hassani and Dupuy, 2013). Intracellular ROS accumulation relative to fluorescence intensity illustrated that the highest oxidative stress was caused by SH1009 with a 77.4% significant increase (P value < 0.001) in ROS levels compared to the untreated sample, confirming the pro-oxidant activity of this aurone against C. albicans cells. In comparison, the pro-oxidant activity of the phenyl aurone raised the ROS level significantly by 45.4% (P value < 0.01) relative to the untreated sample. It was statistically significant that the basic structure was responsible for 24.2 and 48.1% of ROS generation in SH1009 and SH9051, respectively, relative to untreated sample (Figure 7B).

Because aurone SH1009 uniquely induced expression of iron ion transport pathways, it was speculated that treated cells would have low intracellular iron ion levels due to the catalyzing of these stress ions by the SH1009 structure, promoting the pro-oxidant activity. There are two intracellular states of transition metal iron, the oxidized form of iron, ferric (Fe³⁺) ions, which react with oxygen to form oxides with relatively low solubility, and the reduced form of iron, ferrous (Fe²⁺) ions, which react with oxygen to form ROS through Fenton reactions (Fourie et al., 2018). Typically, the reduction of ferric iron(III) to ferrous iron(II) by reductases increases the availability and solubility of iron. However, compromising the efficiency of the network controlling iron homeostasis or changing the availability of ferrous ions are highly related to toxic ROS production (Gammella et al., 2016). Therefore, the levels of Fe³⁺ and Fe²⁺ iron ions were determined using an iron probe assay (iron chromogen) after treating C. albicans wild type and TYE7Δ cells with aurones.

Figure 7C depicts the concentrations of Fe³⁺ and Fe²⁺ in treated and untreated cells where the ratio of Fe²⁺/Fe³⁺ in wild type C. albicans SC5314 cells treated with SH1009 was the highest (3.7) compared to treatments with SH9051 (2.2) and phenyl aurone (2.03) in wild type cells. This validates the transcriptional signature of SH1009 and supports the proposed pro-oxidant activity and the results of the ROS generation assay. Although the Fe²⁺/Fe³⁺ ratio of SH1009 in TYE7Δ mutant was 6.0, this strongly underscores the role of trehalose in protecting the cells since the TYE7Δ mutant showed an intrinsically resistant phenotype against SH1009 (Figure 6C). SH9051 treatment resulted in a 2.2 higher Fe²⁺/Fe³⁺ ratio in the wild type compared with a ratio of 2.03 in response to the phenyl aurone, suggesting that most of the ferric ion reduction trait in SH9051 (∼89.3%) could be the impact of the C ring structure. This impact contributes to (∼54%) the ability of SH1009 in catalyzing the reaction with metal ion stress (iron) which poses a question about the role of the hydroxyl groups and/or methoxy group in the trehalose level.

Intracellular trehalose levels were determined enzymatically based on hydrolysis of trehalose by trehalase into two glucose monomers after eliminating the original glucose from the samples. The glucose concentration, which is directly proportional to the trehalose concentration, was detected fluorometrically by coupling the oxidation of glucose to peroxidation of the fluorogenic substrate, Amplex Red (Carillo et al., 2013). This method overcomes the previous limitation of colorimetric detection which did not have the sensitivity to detect the low trehalose levels in exponentially growing yeast cells (Askew et al., 2009). We observed significantly increased levels of trehalose in the wild type and TYE7Δ cells under all treatments except the SH1009-treated wild type sample (Figure 7D). As expected and documented previously (Askew et al., 2009), the trehalose level was two times higher (P < 0.001) in untreated TYE7Δ mutant cells compared to untreated wild type cells and provides the TYE7Δ cells protection from the SH1009-pro-oxidant effect. SH1009 treatment considerably decreased the trehalose level in wild type cells 7.7 times (P < 0.001) relative to the untreated-wild type sample and 10.6 times (P < 0.0005) relative to the SH1009-treated TYE7Δ sample, demonstrating SH1009 resistance via higher trehalose storage in the TYE7Δ mutant. Additionally, a significant observation was the failure of the phenyl aurone to suppress trehalose synthesis with insignificant differences in trehalose concentrations compared with untreated samples either in wild type or mutant cells. This suggests that suppression of trehalose synthesis could be almost entirely attributed to the functional groups in SH1009 aurone, hydroxyl groups and/or methoxy group.

The impact of the aurone SH9051 nitrogen functional group was not only inactivation of sulfur amino acid biosynthesis, but also activation of degradation of these amino acids, simulating a high exogenous supply of methionine or cysteine (Hennicke et al., 2013; Xu et al., 2018; Figure 4B). According to previous studies, the end products of methionine catabolism are volatile organic sulfur-containing compounds (VOSCs) such as thioesters, thioethers, dimethyl trisulfide (DMTS), and dimethyl disulfide (DMDS) (Landaud et al., 2008). In C. albicans, large quantities of sulfite have been reported as a consequence of cysteine catabolism (Hennicke et al., 2013). To determine the role of the aurones in sulfur amino acid catabolism, total sulfite was measured based on the sulfite-triggered cleavage of the disulfide bond in Ellman’s reagent, which also reacts with compounds containing free thiols, such as VOSCs (Coulomb et al., 2017). SH9051 was the only aurone treatment that significantly (P < 0.0001) elevated the sulfite levels (Figure 7E) while the values produced by SH1009 and phenyl aurone treatment of cells were nearly zero, similar to untreated cells. The overproduction of sulfite associated with SH9051 treatment demonstrated the impact of the nitrogen atom on uniquely transcriptional responses to SH9051. In addition, combining SH9051 with methionine or cysteine demonstrated their synergistic effects on C. albicans growth (Figures 6E,F) by considerable overproduction of sulfite and VOSCs which were more significantly produced (P < 0.0001) in MET4Δ mutant than wild type cells, demonstrating the hypersensitivity of MET4Δ toward SH9051 (Figure 6D). These observations indicate that Met4p activity in C. albicans might not only be involved in sulfur metabolism but also participates in defense against several stresses, including oxidative stresses.

Candida albicans SC5314 strain was purchased from the American Type Culture Collection (Manassas, VA, United States). The C. albicans MET4Δ mutant (SFY39) was prepared by Aaron Mitchell and obtained from the FGSC (University of Missouri, Kansas City, MO, United States). YPD agar and broth, 3-(N-morpholino) propanesulfonic acid (MOPS) buffer, dimethyl sulfoxide (DMSO), BeadBug^TM pre-filled tubes, dithiothreitol (DTT), nourseothricin sulfate (NAT), and ampicillin (AMP) were all purchased form MilliporeSigma (St. Louis, MO, United States). RPMI-1640 medium was purchased from Corning Incorporated (Corning, NY, United States). PrestoBlue, TRIzol^TM Reagent, phosphate buffered saline (PBS), salmon sperm DNA, lithium acetate (LiAc), polyethylene glycol (PEG) 3350, RiboPure^TM-DNase I Treatment, CM-H2DCFDA (Oxidative Stress Indicator), and Amplex^TM Red Glucose/Glucose Oxidase Assay Kit were purchased from Life Technologies Corporation (Carlsbad, CA, United States). CutSmart Buffer, NgoMIV, PacI, Xho1 restriction enzymes, Phusion high-fidelity (HF) DNA polymerase, 2 × Gibson assembly mix, and Chemically competent DH5α Escherichia coli cells were purchased from New England BioLabs (Ipswich, MA, United States). The RNase-Free DNase Kit, Miniprep plasmid extraction kit, Gel extraction kit, and PCR purification kit were purchased from QIAGEN (Germantown, MD, United States). The E. coli bacterial culture containing the CaCAS9 plasmid (reference number 89576) was purchased from Addgene Headquarters (Watertown, MA, United States). The trehalose assay kit and total sulfite assay kit were purchased from Megazyme Inc. (Chicago, IL, United States). The iron assay kit (colorimetric) was purchased from Abcam (Cambridge, MA, United States). Drop-out mix synthetic growth medium, minus methionine and cysteine was purchased from United States Biological Life Sciences (Salem, MA, United States).

Stock solutions of aurone SH9051 and SH1009 were prepared in DMSO at a high concentration of 40 mM. Aurone interactions were carried out in RPMI 1640 medium (adjusted to pH 7.0 with 0.165 M of MOPS buffer) in 96-well flat-bottomed microtitration plates using broth microdilution checkerboard assays as described previously (Shrestha et al., 2015). Two-fold serial dilutions were initially prepared at four times the desired final concentrations (3.125–200 μM) of aurones. Aliquots of 50 μL of each concentration of SH9051 were added to columns 1–7, and then 50 μL of SH1009 was added to rows A to G. Row H and column 8 contained the two-fold serial dilutions of SH9051 and SH1009 alone, respectively. Column 9 was the drug-free well that served as negative control. The inoculum of C. albicans strain SC5314 was prepared as described previously (Alqahtani et al., 2019). The final concentrations of SH9051 and SH1009 after adding 100 μL of the prepared inoculum (0.5–2.5 × 10³ CFU/mL) ranged from 200 to 3.125 μM. The microtiter plates were incubated at 35°C for 24 h. PrestoBlue reagent was used to detect the percentage of fluorescent-viable cells in each well using a SpectraMax M5e spectrophotometer (Molecular Devices, LLC, United States). The minimum inhibitory concentration (MIC) was defined as the concentration that reduced the growth by 90%. The assay was performed in triplicate and the IC₅₀ values were calculated using GraphPad Prism (GraphPad Software, United States). Drug interaction was estimated by FICI model. This model was determined according to the formula: FIC SH1009 = MIC SH1009 in combination/MIC SH1009 alone. FIC SH9051 = MIC SH9051 in combination/MIC SH9051 alone. FIC index = FIC SH1009 + FIC SH9051. FICI value was interpreted to characterize aurone compounds interactions as; FIC index ≤ 0.5 = synergy, FIC index 0.5 ≤ 1.0 = additivity, FIC index 1.0 ≤ 4.0 = indifference, FIC index > 4.0 = antagonism.

Three human cell lines (A549 human lung carcinoma epithelial, human monocytic THP-1, and HepG2 human liver carcinoma epithelial cells) were used to assay the toxicity of aurone SH9051. Human cell lines were grown, maintained and seeded into 96-well microtiter plates at a density of 1.0 × 10⁵ viable cells/well as described previously (Alqahtani et al., 2019). After seeding, the cells were treated with two-fold serial dilutions (3.125–200 μM) of aurone SH9051. After incubation for 24 h at 37°C with 5% CO₂ in a humidified incubator, the cell viability was evaluated using 20 μL of PrestoBlue. The assay for each cell line was performed in triplicate. The 50% cytotoxicity concentration values were calculated using GraphPad Prism (GraphPad Software, United States).

The CRISPR/Cas9 gene editing was used to delete the TYE7 (C2_04060C) allele based on the gene drive platform as described previously with minor modifications (Halder et al., 2019). Briefly, sequences of gene drive constructs, which contain synthetic guide RNAs (sgRNAs), repair templates, and SNR52 RNA polymerase III promoter, and their verification primers are listed in Supplementary Table 4. The CaCAS9 plasmid contains a cassette of the Candida codon-optimized CAS9 endonuclease gene, NAT, and ampicillin (AMP) cassettes (selective markers), gene drive cloning site, and NEUT5 locus sequences for genomic integration. After digesting the CaCas9 plasmid with NgoMIV, ∼1000 bp Oligo pairs of TYE7-gene drive were cloned separately to CaCas9-digested plasmid using Gibson assembly master mix. Then, 1 μL of ligation reactions were introduced into chemically competent DH5α E. coli cells according to the manufacturer’s transformation protocol. Transformants were selected on LB agar + ampicillin. Plasmids with corrected-orientation inserts were verified by amplifying the insert region with universal primers (PAC6725_F and PAC6389_R) and with plasmid fingerprinting using Xho1. After digesting with PacI to linearize the plasmids, ∼1.5 μg were transformed into C. albicans cells (SC5314 background) with the LiAc/ssDNA/PEG/DTT method. Fungal cells were recovered in YPD broth for 4 h and then plated on YPD agar supplemented with 250 μg/mL of NAT. After incubation for 4 days at 30°C, NAT^R transformants were plated again onto fresh YPD agar + NAT plates. NAT^R mutants were validated for their desired gene deletions using colony PCR and gel electrophoresis.

The inocula of C. albicans SC5314 wild type, TYE7Δ, and MET4Δ cells were prepared in YPD broth media at working concentrations of 1–5 × 10³ CFU/mL, and 100 μL were added into the wells of a 100-well Bioscreen honeycomb plate. Then, 100 μL of SH1009 or SH9051 were added to respective wells to reach the final concentrations of the IC₅₀, 16 and 90 μM, respectively, with or without the supplements of methionine, cysteine, and iron(II) chloride. Plates were placed into the Bioscreen C instrument with Bioscreen software (Growth Curves USA, Piscataway, NJ, United States) at a temperature of 30°C with continuous shaking for 24 h. The optical density reads (OD₅₃₀) were measured at 30 min intervals to monitor the growth curves.

Candida albicans, TYE7Δ, and MET4Δ cells from an overnight culture were diluted in fresh YPD to an OD₅₃₀ of 0.12 and allowed to grow to the early exponential phase (OD₅₃₀ = 0.15). Cells were pelleted in 96-wells plate at concentration of (2.5–5 × 10⁵ CFU/mL) and treated with 200 μM of SH1009, SH9051, and phenyl aurone and incubated for 1 h at 30°C. ROS levels were assessed using the fluorescent dye chloromethyl-dichlorodihydrofluorescein diacetate (CM-H2DCFDA) at a final concentration of 20 μM. Cells were incubated at 37°C for 1 h. The fluorescence intensity was measured with a 485 nm excitation and 535 nm emission using SpectraMax M5e spectrophotometer (Molecular Devices, LLC, United States). ROS accumulation was calculated with respect to background fluorescence.

Cultures of C. albicans strains SC5314 and TYE7Δ (MET4Δ mutant was excluded because of the limited resources) were grown until reaching the exponential phase (as descried in measurement of ROS levels experiment) and plated in 12-well plates at a concentration of 1 × 10⁷ CFU/mL and treated with 200 μM of SH1009, SH9051, or phenyl aurone, and incubated for 1 h at 30°C. After treatment, cells were washed once with water, and resuspended in water. Afterward, cells were lysed mechanically with BeadBug^TM pre-filled tubes, containing glass beads. A 50 μL volume of each sample was plated in a 96-well plate for intracellular ferrous iron (Fe²⁺) and ferric iron (Fe³⁺) ions determination according to the instructions of Abcam iron assay kit (colorimetric).

For trehalose determination, yeast cell lysates were prepared by incubation at 95°C for 30 min, after which 200 μL of supernatants were used for enzymatic analysis (Megazyme Trehalose Assay Kit) following the manufacturer’s procedure with modifications. Briefly, 10 μL of solution 2 (NADP+ and adenosine 5′-triphosphate), 20 μL of solution 1 (buffer), and 2 μL of suspension 3 (hexokinase and glucose-6-phosphate dehydrogenase enzymes) were added to the wells to convert the glucose to gluconate-6-phosphate. After incubation with constant shaking for 60 min at 30°C to remove glucose, the reaction plate was sealed and heated at 90°C for 15 min to inactivate the enzymes. After cooling, 2 μL of suspension 4 (trehalase enzyme) was added to wells, and the plate was incubated at room temperature for 8 min. The released glucose monomers were assayed fluorometrically using the Amplex^TM Red Glucose/Glucose Oxidase Assay Kit. Fluorescence readings were converted to absolute concentrations of trehalose based on an internal calibration curve of nmol trehalose standards and normalized for the cell protein concentration (mg) using the BCA protein assay.

A commercially available total sulfite detection kit (Megazyme) was used for sulfite identification in C. albicans and MET4Δ cultures. In brief, cultures of strains SC5314 and MET4Δ were grown in YPD media (as described in the measurement of ROS levels). After reaching exponential phase, cells were treated with 200 μM of each aurone, 20 mM methionine, and 10 mM cysteine for 1 h at 30°C. Afterward, sulfite concentrations were measured according to the instructions of the manufacturer. All sample measurements were normalized to the respective background blank media.